GPTZero is the pioneer in AI detection software. Founders Edward Tian and Alex Cui launched the product soon after the launch of ChatGPT, as a way to help assess whether text is AI-generated or human-written. Their vision evolved over the past three and a half years, from a mission to preserve what’s human on the internet to preserving critical thinking in humans, and building the authenticity suite of products to tackle this.

I’ve talked to so many people in the last few years — educators and students, marketers, hiring managers, trust and safety leaders at platforms with a lot of user-generated content, who have benefitted from the GPTZero product. By joining forces with Superhuman, GPTZero will come to benefit many more people, across the 40 million daily active users of Superhuman’s products — Go, Grammarly, Coda, and of course Superhuman Mail.

Footwork led GPTZero’s Series A round two years ago, and served as the company’s largest investor and only external board member. At a personal level, I’ve known Edward for four and a half years, since he was a junior at Princeton. To see him go from a college student with a big idea, to growing a business that’s tens of millions in revenue and users, profitability, dozens of talented employees, a category-definer and leader, and now the type of exit achieved by very few companies, is remarkable. This is a business that never spent a dollar that it raised, and ended up being lifetime profitable to the tune of millions of dollars, a testament to the market pull and the talent level of the team.

Though it is bittersweet to no longer work together day-to-day, I’m incredibly proud of what Edward, Alex, and the GPTZero team have built. A big shout out to the hard work of Jonathan White, the company’s operations lead, for steering the company through the acquisition process as well. Thank you all for having us on the journey, on behalf of Mike, me, and Footwork’s limited partners, and our terrific co-investors Uncork Capital, Reach Capital, Neo, and more.

The future is bright, as GPTZero is in great hands. Superhuman CEO Shishir Mehrotra has a clear strategy of bundling applications to make the whole suite greater than the sum of its parts. These are two companies that have a shared belief in the importance of human authorship and trust in this age of AI.

GPTZero will soon be available in Superhuman Go, the AI assistant that works in 1 million apps and websites. So it feels appropriate to say: Go Superhuman! We can’t wait to see what Shishir, Edward, Alex, and this combined team cook up together in the years ahead, as they build the authenticity layer that travels with you, wherever you read, write, and create.

I’m flying today from SFO to EWR to head to Princeton for my 15th reunion. Reunions are a big deal at Princeton. It’s one of the largest alumni gatherings in the world every year, with more than 25,000 people typically in attendance, because alumni are encouraged to come back even if isn’t a “major reunion” year (aka every 5 years).

I haven’t been back since 2017; living in San Francisco makes it tricky, our 10th was canceled due to COVID, and we’ve had weddings and kids that have gotten in the way of making the trip. So even though the weather looks awful, I’m incredibly excited to reunite with my classmates.

It is hard to believe that it has been nearly 20 years since our class started college. As I reflect on that time, I have so many special memories. Roommates who have become lifelong friends. Friends who I’ve only seen a few times since graduation. Classmates who I barely knew then but who I’ve gotten to know well since. Professors who I looked up to as role models, and who helped me become a better thinker and writer.

I studied molecular biology at Princeton. I wanted to concentrate in a subject that I felt I could only learn deeply if I studied it in college. And I also felt that innovations in molecular biology would be more likely to define my life than those in most other fields. While MOL hasn’t defined my career, I have no regrets over this decision, and the intersection of biology and computing continues to interest me as an investor.

My real learning came outside of the classroom — in the startups I worked on during college (one of which managed to keep going), and in the clubs and activities around entrepreneurship that I spent much of my time working on throughout undergrad.

Looking at what the Class of 2011 has accomplished since graduation, it is hard not to infer that there was something very special in the water on campus during our years there. We were students from 2007 to 2011: social-media-native having started to use Facebook during high school, beginning college a few months after the iPhone had launched, students during the launch of the App Store, and blissfully protected from the job market during the financial crisis of 2008-2009. We graduated and experienced a decade+ bull market, with technology spurred on by the platform shift to cloud services, and with mobile as the key new surface area for opportunity for both consumers and enterprises.

So it is perhaps no surprise that many of our classmates became entrepreneurs. But even so, the density of entrepreneurial talent in this class of just 1206 graduates is impressive. Here are 15 of them to highlight for our 15th:

• Patrick Wendell, Co-Founder & VP Engineering at Databricks. Patrick dropped out of his PhD at Cal two years post graduation to co-found Databricks, the data intelligence platform for enterprises. The company is now worth $134 billion, with $6 billion in revenue run rate.

• Jack Altman, General Partner at Benchmark, Founder at Lattice. Jack and I met at the Princeton Junction train station back as undergrads. I ended up investing in Lattice when my prior firm led the Series B. We ended up being LPs in each other’s funds. Then he turned into a podcaster and tech celebrity.

• Oguzhan Atay, David Tsao, and Sukrit Silas co-founded BillionToOne, which recently went public on NASDAQ. We all had MOL classes together, so particularly special to see them go on to build a business at the forefront of disease detection. is close to $4 billion in market cap.

• Trevor Martin, Co-Founder & CEO at Mammoth Biosciences. Another MOL major. If you’ve driven up and down 101, it’s hard not to notice the Mammoth building in South San Francisco. The company’s valued at $1.35 billion, working on CRISPR-based therapeutics.

• Nell Diamond, Founder & CEO at Hill House Home. 15 years ago, Nell and I were on stage together as co-chairs of our Class Day graduation event. She went on to create a lifestyle brand, with products such as the iconic Nap Dress.

• Seth Priebatsch, Founder/CEO at LevelUp. Seth and I worked on a startup together freshman year, but then he came up with a new idea, dropped out, launched SCVNGR, turned it into LevelUp, and sold it to GrubHub for $390M.

• Julio Fredes, Co-Founder/CFO at Say Technologies. Acquired by Robinhood for $140M.

• Chetan Narain, Co-Founder at Pepper. Chetan and I met before Princeton — his father and mine were college classmates! We then ended up living in the same dorm freshman year. He went on to found Pepper, which has scaled beyond the Series C stage.

• Caroline Clark, Founder/CEO at Arcade. Caroline and I were high school and college classmates, and it’s been awesome to see her soar with Arcade.

• Sarah Hoffman, Co-Founder/CEO at Maker Wine. Like Caroline, Sarah and I went to both high school and college together. Maker is served on the United flight I’m writing this post from!

• Stephen Lamb, Co-Founder/COO at ZBiotics. We all need to have Stephen’s product in heavy supply for the next several days ;).

• Genevieve Ryan Bellaire, Founder/CEO at Realworld. Genevieve has been on the entrepreneurial journey for many years, helping people figure out how to navigate adulthood.

• Kate Adamson, Founder/CEO at Folio, building an AI-native company at the intersection of hospitality and fintech.

• Brian Jeong, Founder/CEO at Hawthorne. We recently messaged after many years, and it’s been awesome to see Brian’s personal care brand out in the wild.

• Sam Dorison, Co-Founder/CEO at ReflexAI. I can’t resist but mention Sam, whose company Footwork led the Seed round of a few years ago. ReflexAI is the leader in AI-enabled training and QA for high stakes call center operations.

This is by no means a complete list. Classmates like Addie Lerner, Alex Rosen, Dan May, Kyle O’Donovan, and myself have founded investment firms. Breandan Beneschott, Brooks Barron, George Xing, and Kofi Frimpong are some of the entrepreneurs who I’ve jammed with on their ideas in the years since graduation. We have classmates like Max Branzburg who leads all of consumer and business products at Coinbase, and Adam Fry, who is the core product lead for ChatGPT at OpenAI — not technically founders, but helping shape the future of some of the world’s most used products. I would be remiss to not mention that Dario Amodei, co-founder/CEO at Anthropic, was a graduate student who also graduated from Princeton in 2011 with us. And still more who I of course have no idea about, or who I’m forgetting (my apologies!).

Entrepreneurship was not a path particularly encouraged by Princeton in our era — there was only one real class for example, High-Tech Entrepreneurship, taught by legendary professor Ed Zschau (an inspiration for many of us!). It’s never been cool to be working on a startup at Princeton like it is at Stanford or MIT.

And so most of the group above wasn’t thinking about starting a company while in school. We all got to critical junctures in our careers, however, where we had a choice to make. Do we take the traditional path? Or one with a lot more risk? One that is unproven. One that many will call us crazy for pursuing. But one that could be way more rewarding. Something that, if it works, could be our life’s work.

I’m proud to celebrate my entrepreneurial classmates from the Great Class of 2011, and very excited to see many of them back at Old Nassau over the next few days.

‘11 ‘11' ‘11!

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

This week, I had the privilege of dining at The French Laundry (TFL). The restaurant recently celebrated its 32nd anniversary under Chef Thomas Keller, who took it over in 1994. Since the first Michelin Guide of San Francisco in 2007, TFL has received 3 Michelin stars every single year, and is one of the best restaurants in America.

As part of the evening, we were fortunate to get to spend a bit of time in the kitchen with Chef. While the food and wine were exceptional throughout, seeing and hearing from Thomas in this setting was my favorite moment. Looking around, there were a few inescapable details, none more so than a clock placed in the center of a blank wall facing the kitchen, with a plaque below it that reads Sense of Urgency.

Thomas spoke to this phrase in his remarks to our group, and on his website he states the following:

In each of our restaurants and bakeries, we have a clock accompanied by a sign that reads “Sense of Urgency.” We have to begin our day with a sense of urgency and continue on through the rest of it with that same drive. It is a reminder for us to push ourselves—for the need to be organized and focused so that we are able to complete our job and the ones following it successfully. Located within sight of our chefs as they cook and of our dining room staff as they proceed to and from serving our guests, this sign serves as a daily reminder.

I find this fascinating, and extremely relevant to our work at Footwork, for a few reasons.

The actual dishes at TFL are meticulously prepared and the meals last for hours given the number of courses. Sense of Urgency isn’t a phrase that feels top of mind when experiencing TFL as a customer. Yet, there it is in the kitchen (and at the bottom of each side of the menu). It’s an input, not an output. And it’s a value that drives the team to be able to deliver an exceptional customer experience for guests.

Four hours before being in the kitchen at TFL, I was chatting with one of the founders we’ve invested in. My main feedback on the call was to communicate how important moving with pace is (on product, on go-to-market, etc.) in this moment, where the ground beneath us is shifting so quickly given advancements in AI. I referenced the urgency with which several of our other portfolio companies are moving, and how I didn’t feel that urgency in his company at quite the same level when I had recently spent time with the broader team in-person. You can imagine why, four hours later, I was taken aback by seeing Sense of Urgency plastered on the wall in a very different setting.

A couple of months ago, I wrote a post about levels of greatness. I think the greats across industries — tennis players like Roger Federer, chefs like Thomas Keller, entrepreneurs like Jensen Huang — share many of the same traits. They work extraordinarily hard. They are obsessed with their craft. They surround themselves with people more talented than they are on certain dimensions. They are constantly learning and evolving and raising the bar.

They have a sense of urgency about all of the above. These are inputs that enable them to perform at the very highest level, for a very long period of time. You can’t build the next big thing, or be it, without them.

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

A few weeks ago, I spent the day with the team at Canva, a company I’ve been fortunate to be an investor in since their Seed round in 2014. At their 2026 Canva Create, they showcased their latest work in AI, as well as long-requested features such as offline mode.

Subscribe now

But it was in a private conversation with investors that I saw something that I’ve kept thinking about since the event. Founder/CEO Melanie Perkins showed a screenshot of the original deck created in 2011 for the concept that was then-called Canvas Chef.

It’s simple. Type into a box what you want to chef up - e.g. a brochure, invitation, photo book, lab report. And then the design starts to get generated. Does this UI look and sound familiar? 😉

Let’s go back to this moment in 2011. Mel and Cliff were in Perth, Australia, working on Fusion Books, a publishing system for the school yearbook market. They had the idea that the software they were working on could appeal to a broader market than yearbooks. The iPad had launched in 2010 with the App Store, so creating this software as an iPad app made sense to Mel and Cliff. The initial name for the concept was Canvas Chef, and they created this deck, well before they had raised any funding and pre-Canva having a name (Mel adds “It didn’t make it to the finals of a pitch competition 🤣”).

I find the prescience of this deck astonishing. More than a decade before generative AI became a thing, Mel and Cliff imagined a product experience where a user could type in what they want to design, and it starts to get generated. Prompt the base, layout, toppings, photos, and along the way options come up to select. Something that today makes such intuitive sense as a UI, but back then was ahead of its time.

Another thing I love about this deck is that it showcases different potential designs to chef up (brochures, invitations, photo books, lab reports), something that defined Canva from the beginning. When I met Mel and Cliff in 2014, Canva had launched six months prior, and was being used for so many different use cases - Facebook posts and social media graphics, posters, presentations and slide decks, photo collages. Many investors passed because it was hard to put Canva “in a box.”

But that breadth of usage was a feature, not a bug. It has meant that Canva has always been used by regular people for lots of regular use cases, versus being defined by a particular user interface such as chat. I believe this is critical to their success in this AI era — it is the first real AI experience for many of its users, who still are unsure of exactly what to use ChatGPT for. Canva today is the third most used AI product in the world, behind ChatGPT and Gemini, but I suspect the UI of Canva leads to deeper and broader usage today than those other products.

The tagline of the Canvas Chef product in the deck is “Design, Collaborate, and Share beautiful documents in just a few minutes.” 15 years later, that vision is a reality in Canva’s product, and through the magic of AI. Mel and Cliff called their shot in 2011, and they’ve gone on to hit it, to the tune of over $4 billion in annualized revenue and over a quarter of a billion monthly active users. What a story.

A huge thank you to Mel and Cliff for sharing the full Canvas Chef deck, to enable us to share it with the world.

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

I’ve had one of those moments in 2026. Last month, in March, I arrived at an airport, finished a phone call, and rushed through TSA to get to the next one. I got on the phone with George Kurdin, co-founder and CEO at Monk. For the first five minutes, George asked me questions, and we ended up having a thoughtful discussion about the portfolio situation that had made me late to the call.

And then George launched into the story of Monk. It was simple, concise, precise. I could tell he was unwavering in his conviction on why the business is working and why it will be big. I was hooked. My partner Mike met him in NYC the following day, we all met in SF a few days later, and then at the end of the week, we met George and his co-founder Joe Zhou in-person again to give them an offer to invest.

Today we’re announcing Monk’s $25M Series A, co-led by Footwork and Acrew Capital (a delight to work with Lauren Kolodny, who I’ve grown up alongside in venture for the past decade+), with participation from Better Tomorrow Ventures (huge thanks to our friend Sheel Mohnot for the connection!). More from Ryan Lawler in Axios here, and directly from the founders here.

Why did we get so excited to partner with Monk?

1. The team has built a magical customer experience. Monk is the only AI-native accounts receivable platform in market today. The product automates the invoice-to-cash lifecycle: invoicing, intelligent collections, cash application, and reporting. This is a manual process for most finance teams today — sitting between emails, PDFs, spreadsheets, payment rails, and banks. Monk automates it, saving finance teams time, and enabling businesses to get paid accurately and more quickly.

2. It’s loved by several forward-thinking AI-native companies. For customers like ElevenLabs, Profound, and Siro, Monk is making the lives of the finance teams, and those around them, so much easier. It simply gets them paid, faster. On average, customers see 40% lower DSO (days sales outstanding), 24% higher collections response rate, and 25+ hours/month saved. The platform typically pays for itself in month one. These are demanding teams operating at the frontier of AI. Delighting them is hard, but in doing so Monk is building a product that can one day be used by every enterprise.

3. The vision starts with AR but is much bigger. Stripe started with payment processing. Ramp started with corporate cards. Like those two companies, Monk has the opportunity to move beyond its initial wedge to building a broader B2B finance and revenue platform.

4. It really does all come back to the team. George and Joe have built a small in-person team in NYC with an exceedingly high bar for talent. They are intense. They are mission-driven. They hold nothing sacred as an AI-native company, and realize they must evolve quickly as the capabilities of the models evolve.

If the above resonates, come join us at Monk — we’re hiring across engineering and go-to-market. And if your business can benefit from automating AR, please reach out.

We started Footwork five years ago, and have written annual posts about the journey: at inception, year 1, year 2, year 3, year 4, and now this one to mark year 5.

Year 5 felt like Footwork’s most important year yet, by a wide margin.

We made more investments than in any other year - seven (compared to four in each of the past two years) - taking our total portfolio to 24 active companies. We made our final two investments out of our first fund, bringing the initial investment period to a close after over four years, and activated our second fund. Our increased pace was downstream from the highest quality set of potential investments we’ve seen in any year — many more companies that we went deep on as a result of our excitement for the founders, market opportunities, and business momentum.

What are we getting excited to invest in? The themes we outlined at the start of the year still feel relevant. There are really two classes of companies we’re spending time:

1. AI-native small teams building products that leverage the frontier of the models’ capabilities, or are enabling the model companies to be successful, and hopefully aren’t in the kill zones of the model companies themselves 😏 - for example, Anything, Confido, Fuse, Protege, and Waldo

2. Teams that are leveraging AI but don’t have AI as the key “why now” for their business - instead, there’s another market or secular tailwind that has to chance to drive extraordinary growth - for example, Benable, Honeydew, Snout, and a few unannounced investments from the past year in areas like consumer healthcare and marketplaces

When we started Footwork in 2021, we were thinking about AI given our involvement prior in companies like Canva and Stitch Fix that were early pioneers in leveraging ML to improve their product experiences. But we weren’t all-in on AI by any means. Five years in, every portfolio company we work with touches AI in a meaningful way. Some are leveraging AI to automate the entire company. Others are playing in markets that didn’t exist until the past couple of years of AI progress. Morphing ourselves to be a more AI-native investment firm has been a creative and generative experience, and something we are very excited about for the year ahead as we automate Footwork.

What gets us most excited as we reflect on the past year is the fundamental progress and revenue momentum we’re seeing across our portfolio. Some examples:

• Elicit and Power won some of the biggest pharmaceutical companies in the world as customers

• Felt and Heard infused AI deeper into their products and organizations, increasing growth and capital efficiency

• GPTZero continues to build the category-leading product and brand in AI detection, and publicly disclosed that it has grown quickly and profitably to tens of millions in ARR

• Organic shoutouts from customers like Sam Altman was increasingly the norm for Cradlewise as it delighted parents and babies with its products

• ReflexAI and Table22 compounded their businesses while refreshing their brands

• Tracksuit publicly disclosed more than 1000 customers and 240% annual growth as it announced its Series B

• WindBorne is growing very quickly in revenue while building such a unique, N of 1 company — collecting its own proprietary data through its weather balloons, building its own increasingly accurate weather forecasting models, and rewiring their organization with AI (founder/CEO John Dean’s Substack is worth following)

We want Footwork’s brand to be built based on the success of our portfolio, not in advance of it. This takes time, especially when you invest like we do at the early-stage with just a handful of new investments every year. If year 5 is any indication, we are now well on our way. We were also grateful this year for recognition such as Nikhil’s naming to the Forbes Midas Seed List and Mike’s to the Forbes BLK Masters 50 list. We both did a rare joint podcast recently, with Turner Novak on The Peel, in case of interest to hear more from us. We won’t rest until Footwork is a firm that many of the best founders think about going to as they raise their Seed and Series A rounds.

Onward to Year 6!

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

Last month I wrote a post inspired by one of our portfolio companies’ goals for the year — “Automate the entire company.”

At Footwork, we’re of course thinking about this too.

Subscribe now

Our approach has been to deconstruct the venture capital flowchart — find, decide, win, help, exit — and to think about how AI can improve inputs and outputs in each, such that it ultimately drives superior returns for our funds.

In addition to these five areas, there’s the opportunity to leverage AI in our internal operations to automate previously manual processes. We call this bucket firm building — for example, our reporting and communication to limited partners, internal finance function, managing our office, and other activities that may not perfectly reside in the core flowchart.

We have a very small team of 5 people — 2 co-founders that are the entire investment team, and 3 operations folks who give us leverage and focus on firm building. We’ve each been going deep with the models to run our own experiments of where AI can give us leverage, but our goal is to centralize these efforts, and to push towards the frontiers of the models’ capabilities.

We see a few paths to explore more deeply — one is what our friends at USV wrote about publicly last week in their Meet the Agents post. In just a couple of months, they rebuilt their systems to enable the firm to have AI co-workers in several areas. The system is centralized by a web application server, and connects into applications such as GDrive/Cal/Mail, Granola, and Notion to gain context. Another is to centralize with Claude or ChatGPT, connecting all our applications to it, and to spin up agents from there.

We’ve been experimenting with both approaches, but we feel like there is so much more to explore — for example, in setting up the context graph correctly across our team and for each of us individually, with the right security underpinning it, and in doing more novel work than just automating existing manual processes. We are open to other approaches as well — with Claude Code and Codex’s rapid improvements, and new agentic platforms launching daily, we suspect we will be building and rebuilding these systems not infrequently.

One thing on our mind as we automate Footwork is how to make sure we preserve what makes us human and what makes us unique. This week, one of our founders who has been meeting new investors remarked to us:

95%+ of the questions I have been asked are exactly the same in meetings. And I thought that was strange, and then [investor, name redacted] admitted on our call just then he was just asking what Claude told him to ask.

While we believe that AI can improve both our processes and our returns, we also believe that human judgment is what actually drives investing out-performance, perhaps even more so in the years ahead than in the years past.

If you’re reading this and intrigued, consider this a job description. We’d love to find someone who can help us in these efforts. You’ll work closely with me, based in our San Francisco office. You’ll learn about venture capital, have a tremendous amount of agency, a meaningful token budget to experiment aggressively, post publicly about your work, and help build an AI-native early-stage focused venture firm of the future with us. Tell us more about yourself here.

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

Traditional financial institutions — think credit unions, automotive lenders, regional banks — are amongst the laggards. At the heart of consumer and commercial lending is the Loan Origination System (LOS), the platform that financial institutions use to manage the entire lending lifecycle, from application to underwriting to funding. The LOS is a system-of-record and system-of-action that is notoriously hard to get up-and-running, taking more than a year, and thus the industry trend has been to hang on to an existing LOS for 5-10 years before considering switching it. Yet, what we’re seeing in AI across several verticals at the moment is that agents can enable a system to get up-and-running quickly, demonstrate a short time-to-value, and displace an incumbent platform.

Enter Fuse, which launches today out of stealth, an AI-native Loan Origination System and Account Opening Platform.

We first met Fuse founders Andres Klaric and Marc Escapa back in 2022, when they were working on an automotive lending platform, Upfront. As they were originating loans, they became increasingly frustrated with the legacy software they were using, and realized an opportunity to reinvent the stack, particularly as they started to work with LLMs in 2023.

They re-launched as Fuse, which today serves a number of forward-thinking financial institutions. Navigant, for example, is a credit union that has never been able to get a system up-and-running as quickly as Fuse. Canopy is another credit union that is able to automate workflows that were previously prohibitively expensive before they made the switch to Fuse. Fuse enables these FIs to drive more revenue through handling more loan volume and decrease time-to-close, as well as lowering costs via reducing manual workloads, all through an agentic platform.

Today, the company is announcing a $5M Rescue Fund for Credit Unions, enabling the next 50 qualifying credit unions who join Fuse to receive full use of the Fuse platform free of charge until their existing LOS contract expires.

Fuse is also announcing $25M in Series A funding led by Footwork, Primary Venture Partners, NextView Ventures, Commerce Ventures, Clocktower, and FJ Labs, along with early investors and operators of companies such as Chime and OpenAI. More from Marina Temkin here in TechCrunch.

It’s been a delight to partner with Andres, Marc, and the team at Fuse. Just as our other vertical AI portfolio companies such as Confido in CPG, Elicit in scientific research, Waldo in brands and creative agencies, and WindBorne in weather are changing the status quo of their industries, Fuse is automating away previously manual operations and revolutionizing financial institutions. We couldn’t be more excited about the opportunity ahead, and we’re hiring across engineering, operations, and sales, in NYC and remote — come join us.

At the end of Elicit’s board meeting last week, founders Andreas Stuhlmüller and Jungwon Byun said:

One of our goals for the year is to automate the entire company. If we all go on vacation at the end of the year, the company should, in many ways, keep running without us.

What does this mean? Engineering keeps writing code and merging pull requests. Sales keeps prioritizing leads and drafting outreach. Growth keeps running experiments, both paid and product-led, that drive top-of-funnel signups. Finance makes sure vendors and employees are paid, and receivables are collected. The routine workflows within and across every function keep running on their own, without human involvement.

A couple of months ago, I would have thought this goal to be delusional. But in the last couple of weeks, it suddenly feels realistic. That’s emblematic of the rate of progress in AI at the moment, where end-to-end agentic tasks in software engineering, thanks to recent advances in the models, are making those outside of code generation also feel plausible.

Another portfolio company of ours, WindBorne, has AI agents deeply embedded in their work and company communication. Co-founder and CEO John Dean published a post this weekend highlighting some examples: the Growth team runs case studies on their forecasting model by pinging an evaluator agent, the Deep Learning team has agents starting and monitoring all training runs, with bots posting as they debug in Zulip (an open-source Slack equivalent), and the CEO is coding, as he walks through the grocery store, using Zulip as well. One of WindBorne’s senior engineers, Davy, is now fully-focused on building agentic systems for the company.

Subscribe now

In the past few weeks, examples such as the ones from our portfolio abound.

Spotify said on their Q4 earnings call that their best developers have not written a single line of code since December, as a result of their integrations between Slack and Claude Code and an internal system called “Honk” which speeds up the writing and deployment of code.

Shopify’s CEO is suddenly shipping a lot more code, and states:

Running a company is just context engineering internally. Now that skill has even more value in the agentic world.

Agents, not engineers, now do the work at a portfolio company of Michael Bloch’s, which has rebuilt how it operates over the past month.

At Speak, every function is abstracting from doing the work directly to directing AI that does the work, and functional boundaries are blurring.

And at Anything, the team is of course building the Anything platform with… Anything ;).

As the capabilities improve, it is inevitable that every company will have the ability to automate most functions. For now, there are levels, and a set of companies are pushing the frontier of what’s possible. Human taste and decision-making will be more important than ever as AI does more of the grunt work. It’s going to be fun to see how far Elicit, WindBorne, Anything, and others in the Footwork portfolio can go this year.

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

I’m constantly reminding myself that there are many levels to greatness.

Take the sport of tennis. You can be terrific at tennis in high school — #1 on the team, a state champion. You’re objectively great at tennis. But you’d lose in straight sets to the next level of greatness, someone who’s #1 on a D1 college team and vying for the national championship. That #1 college player is great. But they’d lose in the first round of a Grand Slam to a seeded player, someone in the top 32 in the world. And that top 32 player? They are exceptional. They’ve made it to almost the pinnacle of their sport. And yet, they have no chance of winning a Grand Slam at the moment, because they’d be beaten comfortably by Alcaraz or Sinner or Djokovic.

Since the start of 2017, only two men not named Federer, Nadal, Djokovic, Alcaraz, or Sinner have won a Grand Slam (Dominic Thiem and Daniil Medvedev, if you’re curious) out of 36 tournaments played. So there’s the level of greatness that those five players represent, the very highest level, and there are several levels of greatness (four? six?) below them, depending on your frame of reference. The gap between that highest level and even one level below is so stark that barely anyone has been able to touch their level for the past decade. That highest level on their best day can beat almost every other tennis player in the world, whatever level of relatively great, 6-0, 6-0.

What I’m describing is a version of the power law. It’s impossible to ignore as a venture capitalist, where a few investments and companies drive the majority of returns. It’s hard to internalize this sort of exponential function as a human (at least, it has been for me). But it’s out there, in all sorts of places. A few stories that command the most attention. A few cities in which most of GDP is concentrated. A few people that have the most wealth. Power, in the hands of a few.

What keeps me up is that I believe AI is already exacerbating this dynamic, and will only continue to, as any exponential function is wont to do. Since my post over the weekend about agentic engineering and how code generation has become much more autonomous since the start of the year, I’ve heard from a number of founders and companies, most of whom think that they are in the “bucket 2” I describe: “Those who are learning what’s possible by already generating a lot of code.” They may be, but the gap between them and the small set of companies in bucket 1 is huge.

AI builder and investor Matt Shumer, in a post that went viral this week, Something Big Is Happening, stated the following:

AI is now building the next AI

There’s one more thing happening that I think is the most important development and the least understood.

On February 5th, OpenAI released GPT-5.3 Codex. In the technical documentation, they included this:

“GPT-5.3-Codex is our first model that was instrumental in creating itself. The Codex team used early versions to debug its own training, manage its own deployment, and diagnose test results and evaluations.”

Read that again. The AI helped build itself.

This isn’t a prediction about what might happen someday. This is OpenAI telling you, right now, that the AI they just released was used to create itself. One of the main things that makes AI better is intelligence applied to AI development. And AI is now intelligent enough to meaningfully contribute to its own improvement.

Matt captured more eloquently than I did in my weekend post what is happening in code generation in these model companies: the models are now building themselves. And so, they are just going to get better and better, at an increasing rate, particularly when it comes to code generation.

It’s hard to understand exactly where this recursive self-improvement loop will take us in AI writ-large, beyond code generation, but one thing is for sure: a small handful of companies — Anthropic for sure, and likely OpenAI, Google, Cursor, Cognition too — are at the highest level of greatness when it comes to their ability to engineer software. And no one else is close.

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

In a portfolio company board meeting last week with co-investor and friend Sarah Tavel at Benchmark, she said the following:

I was talking to another founder who said they suddenly have no feature backlog. The amount of output that an engineer can have has grown so much just in the last month that, all of a sudden, the feature backlog has gone poof.

What’s Sarah referring to?

There seems to be broad consensus in Silicon Valley that something really changed in the last few months. The coding models from Anthropic (Opus), OpenAI (Codex), Gemini, Cursor, Cognition (Devin), and more got really good. Not just at generating code, but at accomplishing coding tasks that require longer time periods and context windows. Andrej Karpathy posted about this on X on January 26th, describing his own experience with Claude Code and positing that the models crossed some kind of threshold of coherence around December 2025:

And then he followed it up by defining the term agentic engineering as this method of programming via LLM agents that is increasingly becoming the default workflow for professionals, to differentiate it from vibe coding, a term he coined a year ago:

Engineers who have been leveraging these models to code are able to do more and more and more, with multiple agents running 24/7 to build features and products — and hence why the feature backlog on the product roadmap may no longer be a backlog for a set of companies. At Footwork we’re floored on a weekly basis by the amount of product the tiny teams (1-5 people) we’re meeting with are able to build in a short period of time, and anecdotally it feels like there are more and more companies that fit this bill. SemiAnalysis published this week in Claude Code is the Inflection Point that 4% of GitHub public commits are being authored by Claude Code right now, and that will be 20% by year-end. In case you’re wondering what agentic engineering looks like IRL, OpenClaw founder Peter Steinberger posted a photo of his setup which features roughly 15 concurrent Codex sessions:

What does this mean?

There are four types of companies:

1. Those already living in the reality of seeing immense productivity gains from agentic engineering. The companies building the coding models themselves such as Anthropic, Cursor, and OpenAI are almost certainly the furthest along here.

2. Those who are learning what’s possible by already generating a lot of code. Many of them think they are agentic engineering, but they’re not anywhere close to the level of the above group of companies at the bleeding edge.

3. Those who are only just waking up to the coding models and their possibilities, by experimenting with them.

4. Those who have no idea about any of this.

Every company should be in the 2nd bucket after reading this post. The models are already showing productivity gains, and are only going to get better. Having your engineering team be as fluent as possible in these tools is table stakes. If you’re not leveraging the power of agents overnight, someone else competing with you is, and your product will be left behind. If you are in bucket 3, a post I found helpful from this week is by Mitchell Hashimoto, founder of HashiCorp: My AI Adoption Journey.

The good news for you as a startup is that many larger enterprises are in buckets 3 and 4. I’ve always believed the greatest advantages of a startup are speed and focus. AI is providing us with new capabilities on a daily basis to make speed an even greater superpower for those who are able to wield AI. Where to focus is still all about human judgment in most markets, though this is likely to become more agentic over time too.

Some have said that the broad selloff in public software companies in the past week - another “SaaSacre” - is effectively the public markets waking up to the power of Claude Code (as a proxy for agentic engineering); that is, how much software revenue may be at risk because software is becoming so easily developable, and thus commodified.

I think one of the keys for every public SaaS company this year is to show the market how capable they are of agentic engineering. Case studies of how quickly they are shipping new features and products leveraging code generation. Those new features and products driving revenue acceleration. More profitability by doing more with less. Enterprise AI ROI.

And for every company hoping to be public this year, wow does this feel important. Claude Code may be killing the feature backlog. Let’s hope it doesn’t kill the IPO backlog too.

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

Though I drafted this post in the first few days of January, the month went by so quickly that I did not have a chance to refine it and hit “publish.” Oh well… better late than never.

Each year, it sharpens my thinking to try to identity what I believe will be the key trends that will define the year ahead. I graded last year’s predictions here and did alright (3 As, 3 Bs, 1 C). It would be great to hear what you think of this year’s.

Subscribe now

Agentic AI: 2026 is the year that agentic AI moves beyond coding to other areas (while code generation of course keeps getting better and more automated, enabling agents to be more powerful). I think we’ll realize how difficult it is for AI to actually complete many tasks end-to-end (i.e. expectations are too high and we will be underwhelmed), but there is so much investment here that we’re bound to see some really great production use cases of agents. We will be wowed by some demos and use cases, and we will also be scared by some of what the agents are able to go off and do.

Enterprise AI ROI, Consumer AI Fun: if I could use one acronym to describe what I think will happen in Enterprise AI in 2026 it’s ROI. The onus is now on AI to deliver real results: revenue growth, cost savings, more profit. There’ll be consolidation after a period of high adoption at companies that were early to experiment and now have a better sense for what is really working. And for later adopters, AI platforms that get adopted will be the ones that are too obvious in terms of ROI to ignore. In Consumer IA, the one word that I think will describe 2026 is Fun. We humans love to be entertained. While AI has started off as a productivity tool, I think we’ll see several AI products capture our attention in 2026 that are entertaining and whimsical, and were just not possible until the advances of the past couple years.

AI-Enabled Entrepreneurship: one of the mega-trends that I am most excited about in this AI era is that the barrier to starting a business has been lowered by AI, and with certain jobs being eliminated, many people will actually be forced to think of entrepreneurial paths for their careers. With apps and digital products being easier than ever to build, there will several inspiring stories of unlikely entrepreneurial success in 2026. And I expect that the venture-backed kind of technology entrepreneurship will flourish as well.

Energy: It is hard to ignore the amount of attention and investment going in to one of the key substrates for AI: energy. With power increasingly seen as the bottleneck to compute, and the projected data center build out enormous, this is an all-hands-on-deck effort in the U.S., as it has been in China for some time. The energy bottleneck will be at the forefront of the AI conversation in 2026.

Consumerization of Healthcare: the secular trend we’ve seen over the past five years that has led to enormous enterprise value creation has been the consumerization of healthcare. Since COVID, people have increasingly taken healthcare into their own hands. Couple that with innovations such as GLP-1s, and an increasing awareness of the importance of exercise, nutrition, and sleep to one’s health, and it is no wonder that companies such as Hims and Ro, Oura and WHOOP, Function and Rythm, and many others have experienced rapid growth. At Footwork, 1/3 of our portfolio touches healthcare in some way, and we believe that 2026 will be another banner year for this theme.

A Banner Year for Exits: speaking of banner year, as I predicted last year, I think 2026 should be an awesome one for both M&A and IPOs. There are so many private companies of public company scale. I hope to see many of them embrace retail investors this year.

The World Cup: the event I’m most excited about in the U.S. this year is the World Cup. In a country with immigrants from every nation, I expect there to be phenomenal crowds and atmospheres at most of the matches, in spite of the high ticket prices. And with talent spread across so many teams, it is anyone’s tournament to win. While I don’t see the beautiful game becoming much more popular in the U.S. as a result of the World Cup, it will provide lasting memories for those fortunate enough to experience it live, or follow along from home.

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

We spent time meeting every pet-related company we could find, to no avail, until earlier this year, when we were introduced to Emily Dong, founder and CEO of Snout, by fellow founder and exec coach, Nick Moryl.

Emily explained that, while the mega trend of the humanization of pets continues to hold strong, the veterinarian market is going through several shifts. Vet prices are at all-time highs, thanks to growing demand and the consolidation of vet clinics by private equity. This is leading to a lot of pet parents choosing simply to avoid the vet, and consequently, 2025 saw vet visits actually go down year-over-year. A set of pets are not getting the routine care they need which does not bode well for their health. And these are the “why now”s that are catalyzing the growth of Emily’s company, Snout, which has scaled to millions in annual revenue, growing 5X year-over-year.

Snout’s B2B2C platform enables vet clinics to offer wellness plans to their customers. By becoming a Snout member, a pet parent can have peace of mind that all their routine care is covered, such as vaccines, monthly medications, and parasite screening. They get unlimited telemedicine and office visits. Vet clinics get a loyal customer, predictable revenue, and fewer uncomfortable conversations about paying bills. Most importantly, the pet’s health is prioritized instead of being neglected.

As a result of this win-win-win, Snout’s business is flourishing, with over 250 veterinary providers already in-network, including at 8 of the top 15 largest national clinic groups. As we spent more time with Emily, we got more and more excited about Snout’s signs of product-market fit, the enormous market it’s going after in pet health and wellness , and the team she’s built, including COO David Nietzke and CTO Kyle Cartmell. It also became clear that Snout is Emily’s life’s work, having been building in the pet space for more than a decade, and seeing first hand the challenges of navigating the pet healthcare system with her own dog, Bowser.

Today, we’re delighted to announce Snout’s $10M Series A, led by Footwork, with participation from existing investors Pear VC, Bread and Butter Ventures, and Griffon Partners. The company’s also raised $100M in a non-dilutive capital line from Clear Haven to fund payments to thousands of clinics and support pet parents nationwide. Read more about the journey so far and path ahead in this piece from Allie Garfinkle at Fortune that was published in Term Sheet this morning, and from Emily here.

Snout is another Footwork portfolio company that is both consumer and enterprise in its nature. Their platform is loved by both pet parents and vet clinics, for it enables the wellness plans that work well for both sides. It also plays into themes we’ve written about and invested in for years, including consumer subscriptions and democratizing access.

It’s a joy to partner with Emily, David, Kyle, and the Snout team on their mission to ensure no one ever has to make a healthcare decision for their pet based on the money in their bank account. If that resonates with you, we’re hiring, across the board (customer success, engineering, finance, marketing, operations, recruiting, and sales!), particularly in Michigan and New York City. Come join us.

I very much enjoy the process of pinging leaders whose thinking I respect to gather predictions for the new year. If you’re interested in looking back, here’s the archive of these year-end posts from 2025, 2024, 2023, 2022, and 2021.

This year’s contributions are from the founders of companies like Applied Intuition, Instacart, and NVIDIA, leaders at companies including Canva and OpenAI, and venture capitalists at firms such as a16z, Emergence, and USV.

AI once again dominates the themes, with predictions that reasoning models will yield meaningful advances for applications, AI will show up in bottom line ROI for businesses, and that memory and proactivity will lead to better AI experiences. There is optimism for consumer technology having a renaissance, for AI moving further into the physical world, and for AI yielding scientific discoveries. And in our world of startups and venture capital investing, there is the expectation of a robust capital markets environment, with much-awaited IPOs and significant M&A in 2026.

Happy New Year!

Subscribe now

… systems of record vs. AI agents.

The next big thing in 2026 will be an escalation in the battle between Systems of Record companies and AI agent companies. Slack fired a shot across the bow in 2025 by blocking Glean, but we’ll see this ratcheted up as Systems of Record start blocking 3rd party AI agent access to data, start charging hefty fees, and most importantly, start building AI agents themselves and bundling them for free or cheaply with their core offering. On the other hand, AI agents will start bypassing APIs (and associated fees) through computer use automation, pulling in data - traditionally associated with systems of record - into their own systems, and putting their own interfaces around this data. Legal and Healthcare are the two areas where we’ll see this battle play out the strongest in 2026.

-Gokul Rajaram, Founding Partner at Marathon Management

… end-to-end, agentic AI.

The next big thing in 2026 will be autonomous computing — AI that doesn’t just respond but operates. Every major AI lab will ship consumer-facing computer use, and at least one company built for non-developers will hit a $1B+ valuation doing it.

-John Milinovich, Head of Product, GenAI at Canva

The next big thing in 2026 will be a shift from copilots to pilots. Behavior + comfort + capabilities will usher in a shift from AI as copilots (AI assistants selling into and helping the doctor, lawyer at the law firm, accounting firm) to pilots that take over the work and turn the markets towards the end user with full stack approach. The best will give away lots of value for free, transforming the structures of these markets, with monetization coming through the handoff to humans when they need to step in.

-Rebecca Kaden, General Partner at USV

The next big thing in 2026 will be AI systems that replace entire workflows, not individual tasks. Not smarter chat. Not better demos. End-to-end agents that own outcomes, not prompts.

-ChatGPT 5.2, prompted based on everything I learned in 2025

The next big thing in 2026 will be AI agents that own end-to-end business processes. Not just chatbots or single-function AI tools, but autonomous agents that can execute complete workflows. We’re moving from “AI that helps” to “AI that does” - agents that can reason across long horizons, coordinate with other agents, and actually replace entire job functions rather than just augmenting them. The infrastructure is finally there, the economic pressure is real (companies need efficiency gains), and your portfolio companies are proving the market demand exists. The companies winning in 2026 won’t be selling AI features - they’ll be selling business outcomes delivered autonomously by AI agents.

-Granola, prompted based on everything I learned in 2025

… reasoning.

The next big thing in 2026 will be reasoning systems translating directly into AIs that are more versatile and robust. Reasoning will impact not just large language models, but every single industry, from biology to self-driving cars to robotics.

-Jensen Huang, Co-Founder & CEO at NVIDIA (via No Priors)

The next big thing in 2026 will be AI application companies that use reasoning as a core primitive. To date, reasoning models have primarily been used to enhance the capabilities of existing products (e.g., in legal and support). In 2026, we’ll begin to see companies whose core product was fundamentally unlocked by the advances in reasoning models.

-Kimberly Tan, Partner at a16z

… business gains from AI.

The next big thing in 2026 will be AI showing up in the bottom line, not just the top line. As capabilities expand into knowledge work tasks outside of coding, large companies will see tangible profitability and operational improvements from AI adoption.

-Aashay Sanghvi, Partner at Haystack

The next big thing in 2026 will be the deployment inflection: autonomous agents operate independently across knowledge work, adding measurable points to GDP, and expanding to long duration, complex tasks. Meanwhile, AI starts to hit the physical world: robots begin to move out the lab and ship the first production-ready, end-to-end systems into biopharma labs, manufacturing lines, and logistics warehouses. Not prototypes. Not pilots. Revenue-generating deployments that deliver ROI.

-Talia Goldberg, Partner at Bessemer Venture Partners

… limitations of AI.

The next big thing in 2026 will be a reckoning for AI-enabled services.

The category soaked up venture dollars on a bold promise: unlock TAMs an order of magnitude larger than SaaS while preserving software-like gross margins. The catch? Unlike SaaS, rapid revenue growth and strong logo retention don’t prove product–market fit.

You only truly have it when AI is doing a material share of the work at high gross margin. Otherwise, you’ve built a good services firm financed with the wrong kind of capital.

That distinction stops being theoretical in 2026, when these companies return to market for growth rounds. Investors will stop underwriting the promise of AI-enabled delivery and start underwriting proof in the P&L that machines, not armies of humans, are carrying the load.

Some of today’s fastest-growing, best-branded AI services companies will stall, take painful terms, or quietly sell when it becomes clear they’re still mostly human-powered.

The shakeout will be uncomfortable but healthy. It will force founders to shift from bespoke customer work to internal AI leverage, choosing the important over the merely urgent. And it will distinguish AI that compounds from labor that merely scales.

That’s a feature, not a bug. AI-enabled services remain one of the most promising new business models in decades, and 2026 is when the winners will finally earn their multiples.

-Jake Saper, General Partner at Emergence Capital

The next big thing in 2026 will be AI as intelligent as the smartest PhD in any domain, but still not knowing your business. Structured and unstructured context will enable AI to transform business intelligence.

-Jamie Davidson, Co-Founder at Omni

The next big thing in 2026 will be deployment as AI’s bottleneck. Next year the challenge in AI will move from capabilities to deployment. By now Improving the models is much less important than deploying/implementing the technology across the real economy and 2026 will be the year that deployment emerges as the major bottleneck to true acceleration.

-Yoni Rechtman, Partner at Slow Ventures

… proactive AI.

The next big thing in 2026 will be AI that watches how we work and makes it better without being asked (which is both amazing and terrifying).

-Katie Jacobs Stanton, Founder at Moxxie Ventures

The next big thing in 2026 will be proactive AI agents. They’ll have good ideas before you do—and you’ll start to trust them. Over time, you’ll rely on AI not just to organize your inbox and calendar, but to decide what’s worth your attention at all, quietly shaping everything from daily priorities to major purchases.

-Max Mullen, Co-Founder at Instacart

… memory.

The next big thing in 2026 will be memory.

AI applications to date have been interesting, but the real breakout hits will come once product designers figure out how to use memory in original ways. I wrote about this in “Memory Is the New Moat,” but the core idea is that consumers are increasingly willing to give developers deep access to their digital lives in exchange for products that are more personalized, predictive, and useful.

When the mobile platform shift happened, it took years for developers to understand GPS, location, and “on-the-go” use cases. Then it clicked and we got Uber, DoorDash, Tinder, and entire categories that only exist because phones know where you are and live in your pocket 24/7.

We’re in the same place today with AI. Models are powerful, but we still haven’t learned how to use memory to build genuinely new products. In 2026, the best apps will remember conversations, emails, documents, preferences, context, history, and intent.

What a time to be alive if you love building products.

-Jeff Morris Jr., Founder at Chapter One

The next big thing in 2026 will be Memory becoming the new moat in AI. In 2026, companies will split into two camps: those who treat memory as a strategic asset and others who don’t. Winners will figure out their memory strategy and use it to increase engagement & retention. A/B Tests with and without memory activated will show large differences in bounded retention and session length.

-Mercedes Bent, Co-Founder at Premise

The next big thing in 2026 will be AI companions that actually know you. Not chatbots that forget yesterday’s conversation. Not assistants that need constant reminding. I’m talking about AI that remembers your mom’s birthday is coming up, knows you’re stressed about the project due next week, recalls that joke you made three months ago, and understands the subtle ways your preferences have evolved over time. We’ve spent 2024-2025 figuring out how to make AI smart. 2026 is when we make it present. When memory becomes persistent and personalized, AI stops being a tool you use and starts being a presence in your life. Think less “search engine” and more “that friend who somehow always remembers the important stuff.” The future isn’t AI that can do everything. It’s AI that remembers everything about you.

-Claude Sonnet 4.5

… consumer tech.

The next big thing in 2026 will be a set of startups that radically re-imagine existing consumer categories like travel, finance, healthcare, and maybe even real estate. The other shift will be from search-driven experiences to AI systems that proactively observe, suggest, and act on a user’s behalf. Instead of asking for answers, consumers will increasingly delegate intent—granting AI permission to notice opportunities, prepare decisions, and execute with approval. The winners will feel less like apps and more like trusted agents.

-Saar Gur, General Partner at CRV

The next big thing in 2026 will be a huge increase in AI toys and AI digital and physical products for kids & teens. Some already have been hugely successful like Stickerbox. But I think we’ll see a lot more of this type of thing (the younger someone is the least un-natural it feels to have AI entertainment, an AI companion, AI best friend, etc).

-Tony Staehelin, Founder & CEO at Benable

The next big thing in 2026 will be a consumer startup renaissance. We will see more tech founders get started than ever before the previous 10 years. More importantly for the consumer startup founders who are 1-2 years into the journey we shall see a handful of them get to $100m+ revenue run rate.

-Niko Bonatsos, Managing Director at General Catalyst

The next big thing in 2026 will be a fresh boom of consumer platforms. Boredom with older platforms + curiosity at what new tools can create + continued desire for connection and the new will usher this in. After a long stretch of it being too hard to interrupt incumbents, the cracks are showing and the newer, the weirder the better as we crave new consumer digital experiences and connection.

-Rebecca Kaden, General Partner at USV

.. on-device AI leading to better experiences.

The next big thing in 2026 will be taste graphs. For 20 years, consumer tech was built around social graphs. Who you know. What’s been missing is a real model of what you value. Taste shows up in patterns over time, not in prompts. AI agents won’t work without that layer. On-device AI finally makes it possible to learn taste privately and continuously. 2026 is when on-device capability, agent adoption, privacy expectations finally converge. Social graphs shaped attention. Taste graphs will shape action.

-Kirsten Green, Founder & Managing Partner at Forerunner

The next big thing in 2026 will be real experimentation and adoption of local inference. This is the inevitable result of demand for longer inference-time compute runs, open-source reasoning models reaching SOTA status, and these frontier models becoming increasingly accessible on consumer-grade hardware.

-Sumeet Singh, Founder & Managing Partner at Worldbuild

… generative media.

The next big thing in 2026 will be generative media, with image, video, audio, and 3D models (finally) ready to underpin an explosion in applications. Companies like Fal have become critical infrastructure players, but we’re now ready to see the app layer take center stage.

-Rex Woodbury, Founder & Managing Partner at Daybreak

The next big thing in 2026 will be AI video ads. They may overtake non-AI video ads on social networks in 2026.

-Kanyi Maqubela, Co-Founder & Managing Partner at Kindred Ventures

… generated worlds.

The next big thing in 2026 will be Generated Worlds. World models are the most slept-on development in AI right now.

Entertainment hasn’t changed in decades. Whether you’re in Tokyo or Toledo, you watch identical pixels, follow identical stories. Same heroes, same quests, same endings. We spend more time than ever staring at screens, yet every screen shows the same thing.

That’s about to break. World Labs launched Marble, turning prompts into explorable 3D environments. Google’s Genie 3 generates playable worlds that respond as you move through them. Last week, Runway released GWM-1, their own world model with explorable environments and conversational avatars. Meta shipped consumer glasses with displays this fall. Each piece exists. None talk to each other yet.

In 2026, I expect them to finally break through. We’ll see our first entertainment that generates uniquely for each viewer. Not recommendations of existing content, but stories, characters, and worlds that exist only for you. The economics will make sense: Netflix spends $17 billion annually chasing a median viewer who doesn’t exist. Personalized generation will cost less than streaming pre-made content.

We’ll see a glimpse of why our kids will never understand why movies weren’t made just for them.

-Nikunj Kothari, Partner at FPV Ventures

… AI hardware.

The next big thing in 2026 will be AI Wearables - Jony Ive’s new launch, Meta RayBan + other products will reach peak fever. There’s a feeling that the AI revolution constitutes the start of a tech leap that’s just a bit....unfinished. We have all the makings of new intelligence, but it’s stuck in our mid 2000 era phones.

-Mercedes Bent, Co-Founder at Premise

The next big thing in 2026 will be AI hardware starting to make sense. My bet is on the Daylight being the first consumer device to create something sincerely magical here.

-Mario Gabriele, Founder at The Generalist

… verification of AI / non-AI.

The next big thing in 2026 will be verification and provenance becoming core product features on most platforms. As synthetic media explodes products and platforms start really emphasizing ‘this came from a real human’ (eg verified non-AI badge).

-Tony Staehelin, Founder & CEO at Benable

… in AI pricing.

The next big thing in 2026 will be aligning AI pricing with the nature of the work. We are moving past the blunt instrument of seats, but we must be careful not to blindly rush into outcomes. Outcome-based pricing is the ultimate goal, but it is a privilege earned through High Autonomy, meaning true auto pilot behavior, and High Attribution, meaning clearly proven value. For many companies, the strategic bridge will be credit-based models, which offer the predictability of SaaS with the fairness of usage. The winners will not be those with the smartest models, but those with the smartest monetization architectures, mapping pricing to autonomy versus attribution and charging for tasks where they must and outcomes where they can.

-Madhavan Ramanujam, Co-Founder & General Partner at 49 Palms Ventures

… AI app stores.

The next big thing in 2026 will be the OpenAI App Store. Fidji Simo’s hire by OpenAI was all the indication the tech crowd needed and their 2025 fall announcements show with 7 brand partners doubled down. The main thing I’m watching is how larger incumbents reach to sense they’re being disintermediated by OpenAI - how do brands contend with ChatGPT’s embedded app experience taking away native primacy?

-Mercedes Bent, Co-Founder at Premise

… AEO... or not.

The next big thing in 2026 will be an AEO boom and confusion. The conversation going on right now in every retail and local business strategy room (from Jeni’s Ice Cream to Lululemon) is how to rank better in the new AI platforms. Budgets will be big here, results hard to track, and not nearly as predictable as SEO.

-Tony Staehelin, Founder & CEO at Benable

… the neo-labs.

The next big thing in 2026 will be the rise of the neo-lab. A new crop of foundation model companies will spring up, as leading researchers from the existing large labs (Deepmind, Anthropic, OpenAI, etc) branch out to start their own model companies, all with some hypothesis on a new type of research breakthrough they want to double down on

-Jamin Ball, Partner at Altimeter Capital

… physical AI.

The next big thing in 2026 will be Physical AI. This is taking artificial intelligence beyond text and into the real world. It’s about bringing intelligence to the machines that move us and protect us, and that make and deliver the goods we rely on. That’s cars. That’s trucks. That’s fighter jets. That’s drones. That’s robots. The impact of this will change everything. Intelligent machines, for example, will transform the global supply chain, coordinating together into a single intelligent system that’s more sustainable and efficient. It will transform cities as autonomous vehicles become the norm. AI has already made a huge impact in our daily lives. But that impact is only going to be magnified as intelligence makes it into the physical world.

-Qasar Younis, Founder & CEO at Applied Intuition

The next big thing in 2026 will be the Waymoification of the real world. You’ll see AI autonomously pilot more, and increasingly larger, things around you. You’ll pass a construction site and see an excavator moving dirt with no one in the driver’s cab…

-Jake Saper, General Partner at Emergence Capital

… AI x Science (and Cyber).

The next big thing in 2026 will be AI and Science! I think what happened with software engineering last year will happen with science this year and it’s going to be super exciting.

-Kevin Weil, VP Science at OpenAI

The next big thing in 2026 will be the first credible, end-to-end scientific result produced by an autonomous system, from hypothesis generation through iterative experimentation to peer-reviewed publication, alongside the first truly destabilizing AI-enabled cyber or information operation that forces international escalation. Of note, is is likely that they won’t be unrelated events. The same capabilities that enable agents reason, plan, and act across complex systems will prove productive in science and dangerous in security.

-Nathan Benaich, Founder at Air Street Capital

… in healthcare.

The next big thing in 2026 will be a significant shift in both functionality, accessibility and regulatory ability for AI to prescribe directly, allowing swaths of full healthcare experiences to be delivered without human touch. Over time, this will change the cost structure and accessibility of care.

-Rebecca Kaden, General Partner at USV

The next big thing in 2026 will be in healthcare. Silicon Valley has had lots of false starts where it believed technology would disrupt healthcare -- but fundamentally it’s still an industry using fax machines. AI is different. The first big wins won’t be flashy: in 2026, AI for back-office automation plays (that move the needle financially for large healthcare incumbents) will start to break out. We’ll also start to see signals of a few bigger disruptions -- AI-based diagnostics and consumer-centric healthcare companies -- that will make a bigger shift in the industry long-term.

-Travis May, Founder & CEO at Shaper Capital

… in energy.

The next big thing in 2026 will be in energy. The need for abundant, cheaper, and, ideally, cleaner energy only continues to explode given the rapid growth of AI leading to increased innovation in each piece of the programmable energy market—new forms of energy, how it’s orchestrated and delivered. We see a rapid rise in all buildings needing batteries, creating massive opportunity for new solutions to deliver and install them and also build on top of that network.

-Rebecca Kaden, General Partner at USV

… in organizational development.

The next big thing in 2026 will be defying the burden of change management proving to be a top moat in the AI era. When the game is changed - in terms of new platform shifting, a new playbook for marketing in your industry, or modern organizational practices - the realizations themselves are surprisingly evenly distributed. People are more networked than ever before, journalists tout and merchandise the breakthroughs, and consultants capitalize on moments like these. However, switching up the gears of execution takes many years. Change management is human rewiring, not technical. Change management never happens naturally. On the contrary, the ancestral lizard brain in all of us recoils from change by default. As a result, teams without the burden of change management gain the most advantage during generational platform shifts. As a larger and established company, the only way to transcend this outcome is the willingness and might to hack the organization from the top down - the proverbial “founder mode” as Brian Chesky calls it. Transplants of strategy, mission, and practice are just that: transplants. The healthier the organization, the more they require blunting the immune system in the form of a painfully clear narrative, hacked reward systems, and relentless execution. The companies that win in the AI era will be either brand new (and void of change management) or able to undergo serious and fast change within - often with the help of a founder or refounder.

- Scott Belsky, Founder of Behance, Author/Investor, Partner at A24

The next big thing in 2026 will be the end of the “two-pizza team.” The atomic unit of execution inside companies collapses to a one-person squad; a ‘full-stack operator.’ As AI reduces the cost of execution to near zero, the right person can out-build a 50-person org. The bottleneck becomes finding full-stack operators with curiosity, taste, and judgment.

-Sam Chaudhary, Co-Founder & CEO at ClassDojo

The next big thing in 2026 will be promotions. The early adopters of AI are breaking out as top performers at their company. We’re seeing across the portfolio in-house legal professionals, marketers, and all types of workers reduce outside spend by 40% or more. Top performers aren’t just doing good work, they’re cutting back on spend and generating more revenue than ever before. They have meaningful leverage going into annual performance reviews.

-Brianne Kimmel, Founder at Worklife Ventures

… startup venture activity.

The next big thing in 2026 will be a meaningful increase in venture activity. The “public AI trade” will roar back, GDP growth will keep coming in strong, inflation will come in lower, interest rates will drop - all leading to another big spike in venture market activity. The “hottest” companies will raise 3+ rounds in 12 month periods (and increasingly two rounds at once in tranched structures), and valuations will soar across everything from venture to growth stage investments. VC funds will race to raise larger and larger pools of capital, and by the end of the year we’ll all be saying “how could we do this again, 2021 wasn’t that long ago.” That being said - despite all of this, some of the best venture investments of all time will be made in 2026.

-Jamin Ball, Partner at Altimeter Capital

… from incumbents.

The next big thing in 2026 will be the empire striking back. If the last two years have been dominated by discourse about the foundation model players and AI-native companies growing like crazy, in 2026 the pendulum will swing back to the incumbents who will layer in strategic M&A with their existing distribution.

-Shravan Narayen, Partner at IVP

… M&A.

The next big thing in 2026 will be acquisitions by the large AI labs being the main way that venture-backed companies will achieve large exits. Startups have a window of opportunity to gather industry-specific data and context by building and distributing vertical-specific applications. Over time, the AI labs will want to suck up this context and data themselves and will aggressively acquire these businesses.

-Damir Becirovic, Founding Partner at Relentless

… IPOs.

The next big thing in 2026 will be IPOs and liquidity. OpenAI has no choice but to go public, Anthropic has plans, Databricks, for the first time said “they don’t rule out going public by the end of 2026.” And lots of mid-cap businesses will be emboldened to go out once the “generals” go. This is good for the overall US economy, and for limited partners.

-Ram Parameswaran, Founder at Octahedron Capital

The next big thing in 2026 will be the return of banner tech IPOs and related IPO hijinks, from mascots to VCs and former execs jostling for a listing day invite. (Looking at you, SpaceX, Databricks, Anthropic etc.) This will have 3 main consequences: 1) the latest round of hand-wringing and arguing about IPO ‘pops’ and lots of people tagging Bill Gurley on X; 2) a massive round of VC victory laps releasing years of pent-up, illiquid frustration; 3) newly-minted employee billionaires launch more startups, angel invest in each other again, and pay for subscriptions to upstartsmedia.com.

-Alex Konrad, Founder & Editor at Upstarts Media

… financial shenanigans.

The next big thing in 2026 will be the world relearning about Financial Shenanigans and why accounting is actually important.

-Shravan Narayen, Partner at IVP

… entrepreneurship.

The next big thing in 2026 will be regular people shipping large software businesses ($1m+ revenue) with AI.

-Dhruv Amin, Co-Founder at Anything

The next big thing in 2026 will be the enablement through AI of a new era of creators, in content and in code. We’ll see the first independent content creator go viral through primarily AI-generated content. And we’ll see new organizational structures emerge because more people can now ship products end to end.

-Chetan Narain, Co-Founder at Pepper

… kindness.

The next big thing in 2026 will be Kindness. Apolitical, nondenominational, online and offline kindness. Mainstream tech population is exhausted by the escalations of the last decade. Begins investing time and energy in more community, civics, and charity. Leaves the most toxic social spaces to trolls and bots. And this is aided by AI, because when chatting, its default is more polite - which subtly influences the auto-completes and auto-replies we send, as well as retrains us to be a little nicer generally. The etiquette dividend of the AI boom!

-Hunter Walk, Co-Founder & Partner at Homebrew

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

A year ago, I wrote What I’m preparing for in 2025. Today I’ll grade those 7 predictions, as well as assess others’ submissions for what The next big thing in 2025 will be…

Subscribe now

AI-Native Applications

What we’ve seen over the past year is a number of companies fitting a certain profile — a small team, building an AI-native product that is growing quickly in users and revenue, often serving both consumers and enterprises, and doing so very capital efficiently. For example, one of Footwork’s portfolio companies is about to surpass $10 million in annual recurring revenue, having launched two years ago, with a team of less than 20 people today, and with more capital on the balance sheet than they’ve raised in total venture funding (i.e. they have been profitable over the lifetime of building the business). A few years ago, this would have been truly extraordinary, but today there are dozens of companies that are writing a similar story. And by the end of 2025, I think there’ll be hundreds of companies like this, because it is uniquely possible in this era of AI. Software is faster and cheaper to build, and products can spread by word-of-mouth very quickly. ChatGPT is of course the canonical example here (though not in capital efficiency given OpenAI’s level of burn), having reached more than 300 million weekly active users since launching just over two years ago. A truly mind-boggling growth rate.

Grade: A. In our world of tech startups, this felt like the key theme of 2025. There are so many AI application companies that have grown quickly and efficiently to tens of millions of dollars in ARR — we’ve almost become numb to how incredible these stories are!

AI Workers

Related to the above, some of the applications that will grow the fastest this year will be AI workers — coders, salespeople, customer support agents, debt collection agents, clones, and more, that are AI team members instead of human ones. Though the “agent” term has become widespread, I really love this framing from Elicit co-founder Jungwon Byun: “The next big thing in 2025 will be AI moving from assistants to collaborators.” I think of the compelling and productive AI workers more as collaborators than as agents, and suspect that this framing will resonate with end users and software buyers as the year goes on.

Grade: B. While AI became more agentic in 2025, the promise of AI workers completing end-to-end tasks fell short of reality. It feels obvious that this is coming, perhaps as the key theme of 2026.

Non-AGI Breakthrough

While all eyes are on AGI, which we will clearly make progress towards in 2025, if not “achieve” by many definitions, I believe we’ll see a breakthrough in another area of technology, such as biology, robotics, energy/climate, healthcare, that will blow us away and be a ChatGPT-in-2022-like moment in 2025. It is likely to be related to all the advancements in AI but I think it will be distinct and open up a new set of “platform” opportunities for early-stage startups in the way that LLMs have.

Grade: B. While there wasn’t an obvious “ChatGPT moment” in 2025, there were several breakthroughs - for example, Claude’s models that made code generation, vibe coding, and application building things that AI is extremely capable of - that I think we’ll look back on as fundamental. And the progress made in other areas such as robotics and scientific research, while yet to bear a groundbreaking moment, felt very meaningful.

Crypto <> AI

The intersection of crypto and AI has felt like a fertile ground for compelling new products for a while, and I hope we see some awesome user experiences uniquely enabled by this intersection in 2025. I am no expert here, but when I think about the adoption of stablecoins over the past few years, crypto-native products like Bridge, Farcaster and Sling growing in usage, the magic of many AI-native user experiences, the usability issues with many crypto-native experiences, the data trust issues with many AI-native products… it just feels like there is much that can marry the best of each area while eliminating the worst of each. As with small teams building fast-growing, capital efficient, AI-native application companies, per the first theme in this post, I’d love to meet anyone building something that users are loving at this intersection of AI and crypto.

Grade: B. Interestingly, it feels like 2025 was very significant for crypto applications, but just not those that intersect directly with AI! It was stablecoins and prediction markets that both experienced serious inflections in market pull this year.

Hardware! (not Apple)

Over this past year, I got into our Model Y in San Francisco, reversed out of our garage, turned on full-self-driving, and did dozens of drives where the only times I took over from FSD were upon reaching our destination to park (a few lane changes too, but with FSD still engaged). It is remarkable how far Tesla and Waymo (which I also took dozens of times in SF last year) have come, and autonomous driving will have an exciting year of expansion to more cities in 2025. In consumer hardware, I can’t believe I’m typing these words (and this is not investment advice!), but I am long Meta, short Apple. I became a very active (weekly, often daily) user of Meta Ray-Bans last year, which are especially great to use with young kids (hands are often full!). I’m excited for Orion. And I’m really bummed about the latest iPhones, (the lack of) Apple Intelligence thus far, and Vision Pro.

Grade: C. I don’t think we got a revolutionary piece of new hardware in 2025. OpenAI did buy Jony Ive’s io and it will be exciting to see what they launch together in late 2026 or 2027. But given my two straight years of optimism and failed hardware predictions, I’m not holding my breath for what unfolds here.

Exits! (though 2026 will be better)

IPOs and M&As will be back this year. I thought that last year would be a readiness year for many companies to be able to go public this year. And they will, although I think 2026-27 will be more exciting IPO years than 2025, given the lack of much urgency to go public for leading prospects.

Grade: A. The markets were liquid in 2025 with some massive M&A such as Google buying Wiz for $32B and ServiceNow acquiring Arnis for close to $8B just this week. Global M&A across all sectors was close to $5 trillion, one of the highest years on record (second only to 2021). I stand by my prediction that 2026 will be better, especially with a few long-awaited IPOs that should finally arrive, such as SpaceX.

Government (In)Efficiency

If you spend anytime online, particularly on X, it will be hard to avoid following the efforts of DOGE (the department of government efficiency) this year. I think there’ll be a lot of attention on where they think they’re finding “efficiency” but in reality it will all be more inefficient than the folks working on this want it to be, because of the chaotic way in which they go about trying to enact change. I am optimistic that some of the change will be positive progress, but I suspect there will be more talk than action that actually moves the needle on efficiency (particularly given the % of federal spending on entitlements, defense, and debt, that is near-impossible to touch).

Grade: A. I’m particularly proud of this take, because as the year draws to a close, DOGE is nowhere to be seen, the federal government from the outside-looking-in seems to have been barely changed for it, and all the noise about efficiency that we had at the end of last year seems to (sadly) be dead.

There were many great submissions by 50 of technology’s top thinkers to last year’s The next big thing in 2025 will be… piece. Here are some of the ones that feel particularly on-the-nose as the year draws to a close (look for the ⭐s!).

… AI applications.

The next big thing in 2025 will continue to be the evolution of AI at the application layer. AI applications will continue to advance and be able to create significant value across industries, domains, and modalities

-Chetan Puttagunta, General Partner at Benchmark

The next big thing in 2025 will be verticalized product offerings with AI at their heart that take on boring old industries with enormous TAM. What SaaS was to On-Prem, this new business archetype will be to SaaS!

-Anand Chandrasekaran, Founder & President at Crescendo and Managing Partner at Celesta Capital

The next big thing in 2025 will be in the world of artificial intelligence, where it’s going to be cool to be a GPT wrapper again. Consumers and businesses both need help figuring out how best to use already very powerful language models so we’re going to see a bunch of applications crop up that help people get the most out of these models. Startups will build sales motions, integrations, customer know-how, etc and won’t need some huge technological moat to justify their existence. Of course no one will want to call their company a “wrapper.” Think more “GPT refiner.” These companies will get data particular to their use-case, do some human reinforcement learning, and deliver an easy-to-use product that leads the horse to water instead of an intimidating empty text box.

-Eric Newcomer, Founder/Writer at Newcomer

The themes called out here by Chetan, Anand, and Eric, of AI applications continuing to evolve and grow, including those in verticals, and those leveraging LLMs to create easy-to-use products, were key in 2025.

… software development.

The next big thing in 2025 will be the “YouTube-ification” of software. Building software will become the next frontier of creative expression. Just as video production once seemed daunting, AI coding assistants are turning us all into software developers. Why only do a podcast when building a high quality, monetizable software product will be as simple as making a YouTube video?

-Jeff Morris Jr., Founder & Managing Partner at Chapter One

The next big thing in 2025 will be lots of new products coming from the falling barriers to software engineering. So far, most AI products were imagined by AI engineers and researchers - we’ve yet to fully see the creative output of non-AI builders using this general purpose technology. I think a blockbuster app will top iOS or similar charts that’s built by someone who can’t actually code.

-Nathan Benaich, Founder & General Partner at Air Street Capital

One of the fastest growing areas of applied AI in 2025 was in code generation and software development. Jeff and Nathan nailed this trend, as exhibited by the growth of companies like Lovable, Replit, Anything, and others this year.

… small teams scaling growing companies.

The next big thing in 2025 will be startups of 10 people getting to $10M ARR in less than 18 months with less than $5M raised.

-Garry Tan, President & CEO at Y Combinator

Anecdotally, I’m amazed by the number of small teams that have quickly built a business exceeding $10M in ARR (though “ARR” means something different in different companies — definitely not always “annual recurring revenue,” but some sort of annualized revenue run rate). The Lean AI Company Leaderboard also grew in number of companies, and in the scale of those companies, during 2025, as one proxy for this trend that Garry called out.

… San Francisco.

The next big thing in 2025 will be San Francisco. SF is BACK BABY! Promising local election outcomes, a Mayor who wants to work with the private sector and tech again, IRL matters for communities and startups, abundance YIMBY mindset to kickstart building, YC/Ron Conway never gave up on the city, and so on. Yes, there are challenges - some of which will take several years to fully move on - but 2025 is the year the narrative starts turning. I used to worry it would take a kinetic crisis (eg earthquake) to unite us but a non-kinetic one (post COVID ‘doom loop’) inspired and united enough locals to fight again for the city we love. LFG.

-Hunter Walk, Co-Founder & Partner at Homebrew

As we exit 2025, San Francisco clearly feels like the center of the technology world. The companies at the bleeding edge of AI are here, and the city feels better than it has felt in a long time as a place to live and work. Hunter captured this prediction perfectly as the narrative on SF very much turned this year.

… data centers.

The next big thing in 2025 will be Data Centers. Expect to see increased investments in data centers, cooling technology for data centers, and energy to run data centers. Eg. under-water data centers, diamond-cooling, nuclear renaissance.

-Sriram Krishnan, Co-Founder & General Partner at Kearny Jackson

There was a consistent undercurrent of discussion and work in 2025 on the energy and data center layers, where build out is happening at an unprecedented level, to try to keep up with the demand for AI. Just look at all the OpenAI data center announcements as one proxy. Sriram called this accurately.

… IPOs, M&As, and liquidity.

The next big thing in 2025 will be both the tech M&A and IPO markets booming, as companies take advantage of a more receptive stock market and a more permissive FTC to deliver liquidity and exits to shareholders.

-Gokul Rajaram, Co-Founder & General Partner at Marathon

The next big thing in 2025 will be liquidity. With the new government being far more pro deregulation, we’ll see the IPO market and M&A come roaring back.

-Nikunj Kothari, Venture Partner at Khosla Ventures

The next big thing in 2025 will be the continued maturation of financial products for late stage, seemingly pre-IPO companies. “Will they go public or not?” will continue to be a debate, but firms will offer secondaries, private debt, partial buyouts and more. And with an ironic twist, you’ll start to see names of those firms like General Catalyst or A16Z floated more frequently to IPO themselves.

-Aashay Sanghvi, Partner at Haystack

Consistent with my exits prediction and reflection, 2025 was a strong year for M&A and liquidity as Gokul and Nikunj thought it would be, and there was a lot of late stage private activity in advance of companies going public, as Aashay predicted —see OpenAI’s $60B+ raised, plus their unique deals with hyperscalers, and Databricks’ $4B Series L (!), as examples.

… big tech earnings.

The next big thing in 2025 will be the acceleration of EPS and free cash flow/share at big tech (Google, Amazon, Meta and Microsoft).

If 2024 was a stabilization of revenues as customer confidence came back, 2025 and beyond could result in optimism for both US corporates and consumers. As consumer confidence returns with the prospect of a stable jobs market and increasing wages, consumer spending should be robust. Software spend should accelerate as corporates feel better about their prospects and spend budgets on areas outside of preparing their data estates for AI applications.

This should result in accelerating top lines are the largest mega caps - for example, Google search revenue should accelerate. This thinking is simply not captured in consensus.

My contrarian (and potentially controversial) prediction is that GPU/AI-driven capex will remain high BUT will come in lower than expectations - resulting in accelerating free cash flows at the mega caps. Combined with cost discipline, FCF / Share and EPS will likely accelerate over 2025 (and potentially 2026).

-Ram Parameswaran, Founder at Octahedron Capital

The next big thing in 2025 will be mainstream adoption of agents in large enterprise leading to increased earnings and quarterly beats, resulting in continued stock ATHs among hyperscalers.

-Chad Byers, Co-Founder & General Partner at Susa Ventures

The prediction award of 2025 goes to Ram, who got several of his specific calls right, such as EPS acceleration at most of the big tech companies, and Google’s search revenue accelerating (12% year-over-year growth, beating expectations). Chad also accurately predicted all time highs amongst several of the hyperscalers, and quarterly earnings beats for most of this group.

That’s a wrap on 2025 — what a year. I’d love to hear your reflections on these predictions, and what the key developments were in the last 12 months from your perspective.

Next up: 2026, and a post on what many of tech’s best thinkers believe the next big thing will be next year (send me a note or leave a comment if you have a contribution!), as well as what I’m personally preparing for.

Merry Christmas and Happy Holidays!

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

I’m currently on a Friday evening flight back to San Francisco after spending 4 days in NYC. The WiFi barely works, I’m exhausted, but I’m in a reflective mood, and thus using this opportunity to write.

It’s been a year of many new investment announcement posts on this newsletter, but few of my thoughts beyond them. The investments are, of course, expressions of what I and Footwork have been excited about. But the story behind the story is that I’ve been extraordinarily busy — we’ve made more investments in 2025 than in any other year. I haven’t had enough time for deep work beyond these investing sprints, or for proper catch-ups with many founders outside of the regular check-ins and board meetings with those we currently work with.

This week in NYC brought me back to my roots and reminded me why I love this job so much. And the words are flowing off the keyboard as a result.

Subscribe now

NYC has been good to me

While moving to Silicon Valley from the UK as a teenager is what changed my life, NYC is where I got my start. From 2007 to 2011, I went to college an hour south, and gradually became more familiar with the NYC tech scene, as it really started to emerge in that era.

We started working on Artsy on Princeton’s campus in Fall 2008, won the rookie disruptor award at the very first TechCrunch Disrupt in NYC, and our first office ended up being at General Assembly in Flatiron.

While founder Carter Cleveland took Artsy forward, I decided to try my hand at investing, with my first real job beyond working on startups: joining Insight Partners in Summer 2010. I loved the job and life in the city, and knew I wanted to come back to Insight after graduation. During my senior year, I worked to expand the surface area of the Princeton Entrepreneurship Club, leading our first NYC TigerTrek, where a group of students met the Etsy and ZocDoc founders, as well as Fred Wilson (photo below!).

As a side note, I was so inspired by Fred - both from meeting him that day at the USV office and from reading AVC - that I almost started my first blog in 2011, once I was back at Insight full-time. I thought it could be titled A JVC in NYC. I wish I’d had the confidence to do it, or perhaps that the concept of anon accounts existed back then.

When I moved back west to join Shasta Ventures in 2012, I was excited to steep myself in classic early-stage Silicon Valley investing. The first few new investments I worked on, such as ClassDojo, were based in San Francisco, which increasingly felt like the center of the tech universe back then (vs. Palo Alto and other places in the SF Bay Area). But I shadowed existing portfolio company board meetings in NYC from the get-go, and then, in 2014, I found myself getting excited about more and more companies outside of the Bay Area — Canva in Sydney, and Hinge in NYC were two such investments that year.

My former partner Tod Francis called me by the nickname “New York” in this phase — I was constantly out there meeting founders, making more than 10 trips per year. And I ended up finding and leading our investments in Frame.io in 2016, The Farmer’s Dog in 2017, and Brigit in 2019.

These 5 companies — Artsy, Brigit, Frame.io, Hinge and The Farmer’s Dog — all have become category-definers, 3 of them awesome acquisitions, and collectively represent $2.5B+ of annual revenue and $10B+ of enterprise value today. And I got to be on their journeys, at inception in one case, and at Series A in the rest.

This week in NYC, 2 of those 5 founders and their spouses invited me over to their beautiful homes for dinner, and we got to reflect on our years of partnership, plus what has unfolded since. It was very special. I also had several unplanned run-ins and catch-ups (a uniquely NYC experience!) with founder and investor friends, and college classmates (two of whom, Oguzhan Atay and David Tsao, took their company BillionToOne public on NASDAQ yesterday - simply incredible!).

Now, in the Footwork era, 7 out of 23 of our portfolio companies’ CEOs are based in NYC, including Confido, GPTZero, Honeydew, ReflexAI, Protege, and 2 that we have yet to announce. I think that, when all is said and done, the NYC companies I’ve been fortunate to work with will represent a very compelling body of work, and high hit rate, in terms of both returns and impact.

Empire State of Company-Building

Why is NYC a great place to invest in and build companies? As I reflect on my own experiences with NYC companies, the through-line is their focus on the fundamentals. These companies aren’t (or weren’t, in the case of the ones that have exited) exciting because of hype driven by media or by massive fundraising rounds. They worked, or are working, because they look like great businesses. The teams are executing such that the P&L tells the story. Take the Footwork companies — 3 of the 7 are growing at incredible rates while also being free cash flow positive. They have more cash on their balance sheets than they’ve ever raised in equity capital. The rest are growing quickly, and all are already millions or tens of millions in revenue.

I do believe that San Francisco and Silicon Valley continue to have the highest concentration of the most ambitious founders, and likely always will. But NYC has plenty of founders with world class ambition, and most let their businesses speak for themselves, which feels different from a lot of what we see in SF.

Perhaps one of the keys is that NYC has many people, not just startup founders with world class ambition — arguably the highest concentration of any city anywhere. You may be a great founder, but you’re around great artists, bankers, chefs, lawyers, and more. When everyone around you is reaching for the top, it can’t help but raise the collective ceiling. Everyone honing their craft, finding their footwork in every field, in the city that never sleeps. Just as in San Francisco, I don’t believe elections affect this spirit; there’ll be ups and downs but NYC will always attract many of the best.

Take the ingredients of talented humans, a focus on the fundamentals, ambition in every domain, a 24/7 energy unlike any other city, and great things happen. Perhaps I have also personally benefitted from the lack of as much intense competition for venture investments in NYC as in Silicon Valley. And from being an outsider in NYC, just as certain NYCers have benefitted from being outsiders in Silicon Valley.

So here I am, raising a glass to New York City from 36,000 feet. Thank you for the memories this week, and for the past 15 years. Here’s to building more companies that become the next big thing in this great city. I hope I get to be part of a handful of them.

Thank you to Damir Becirovic — our walk on the High Line this morning crystallized much of my thinking to be able to write this in one (airplane) sitting — and to everyone else I saw this week, especially J and Z for the lovely dinners.

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.

Why did we get so excited?

Word of mouth drives one quarter of the global economy. Billions of dollars change hands every day because someone says “you have to try this!”— whether it be a restaurant or hotel, a set of products to buy, or a service provider to contract. Offline, recommendations have always been the most powerful driver of commerce. But figuring out how to bring those human recommendations online has remained stubbornly unsolved.

That’s where Benable comes in. Benable is the horizontal recommendation platform that makes it easy for anyone to compile and share what they love — products, places, services, and more. Today, over 500,000 users across 150 countries have shared more than 10 million recommendations on Benable, from local coffee shops to travel itineraries, from trusted skincare to go-to realtors. Someone adds a new recommendation every two seconds. And the growth the product has experienced recently has been remarkable: 50% month-over-month, almost entirely from… word-of-mouth 😉.

How did Benable get here?

Benable’s founder, Tony Staehelin, spent years in the startup idea maze reflecting on what to build next after the last company he co-founded, Teespring, went through a real startup rollercoaster. Once upon a time, Teespring was a Silicon Valley darling. It grew from $5M to $150M in GMV and 70 to 200 employees in a year, raised capital at a $650M valuation, and then it ran into headwinds, finding it very difficult to get off the drug of chasing its own growth curve and course-correct. Tony witnessed the downsides of hyper-growth without an enduring product, and was determined to do things differently as he embarked on the Benable journey. He spent two years outside the spotlight, interviewing people about their biggest problems - strangers he would seek out on Nextdoor and Indeed, need-finding in a way we’ve never seen. He became fascinated by how great recommendations travelled via word of mouth and the role that plays in daily life. Tony studied why previous attempts at recommendation networks had failed, and methodically crafted a product that felt magical from the first interaction.

We were fortunate to meet Tony through our good friend, co-investor, and Tony’s former Teespring colleague, Jack Altman, and got very excited about his DNA as a founder, the magnitude of the opportunity Benable’s going after, and the early signs of product-market fit. This week the company is opening up and doing its first-ever press, already at significant scale — more here from Ann Gehan in The Information. Footwork led Benable’s $11M Seed round, and the company’s investors also include Jack, venture firms Third Prime and PJC, and angels like Scott Belsky, Walker Williams, and Cat Lee.

The opportunity in AI, the creator economy, and beyond

In the age of AI, it feels like the experience that makes us human is at an all-time premium. While people are turning to AI to get help, the validation and authenticity humans bring from their lived experiences seems even more valuable to capture. And the desire for this is showing up in Benable’s usage and growth. Built for the creator era, Benable is tapping into the worldwide surge in the number of people sharing content, and brands funneling spend into the creator economy. On Benable, anyone can create and share recommendations, not just creators with a massive following. And all users can monetize their content along the way, via the more than 35,000 brands that partner with Benable today, like Etsy, Ulta Beauty, Target, VRBO, Nike, and TripAdvisor. Brands pay creators whenever people shop, and use Benable as a way to find talented authentic voices passionate about their products.

To get started with Benable yourself, make your first list here, and check out some of my favorites: feeling crafty, baby essentials, sustainable restaurants, and London with kids (by me!).

Benable’s going after a massive opportunity — the ubiquitous social network for recommendations. A place where the $6 trillion of annual word-of-mouth commerce globally is organized, searchable, discoverable, and monetized . A generational company on the scale of Pinterest — but broader, deeper, and more authentic. We couldn’t be more excited to work with Tony and the team to realize this vision.

One of the magical experiences that AI has made possible is that of building apps without needing to know how to write code. It is a sensational feeling to see your words and ideas quickly turn into a prototype of a website, a mobile app, or whatever else you are dreaming up. There are plenty of companies going after this idea, with several products growing very quickly in customers and raising a lot of capital. But most of these companies fall short of enabling users to launch production software.

That’s where Anything comes in. Anything is built with reliability at its core on its own infrastructure stack — an AI agent that lets you go from text to an app that can actually be launched for paying users to use. There are three core elements to the product, beyond reliability, that set Anything apart from others:

1. Native mobile apps and web apps. You can ship apps to the App Store even if you’re non-technical!

2. Designs that pop. When you’re building something that you plan to launch, it needs to be well-designed, and Anything’s agent makes sure of that.

3. Everything built-in. From databases to storage, payments to AI models, everything you need for your app to be production-ready is built in to Anything.

Anything launched last month, and, as they posted about recently, the business hit a $2M revenue run-rate in 2 weeks. What is most exciting is seeing all the ways in which real builders are using the product — launching mobile apps to paying subscribers, internal tools for training colleagues, products that would otherwise have cost tens of thousands of dollars for a development shop to produce. Users like real estate agents, film producers, healthcare workers, people solving problems they know first hand, who are creating products through Anything that become real businesses.

Today we’re delighted to announce Footwork is leading an $11M round in Anything. We are so excited to partner with co-founders Dhruv Amin and Marcus Lowe, and the small but mighty 8-person team building this magical product, alongside investors Uncork, Bessemer, M13, and the founders of Shopify, Zapier, Intercom, Turing, Front, Docsend, and Shippo. More on the story here directly from the company, and from Marina Temkin in TechCrunch.

Why did Footwork invest?

Thanks to an introduction from Andy McLoughlin at Uncork, we first met Dhruv and Marcus just over a year ago. But we didn’t get there to invest at the time — thousands of users were using their then-product, Create, but it didn’t seem that there was deep customer love yet.

To their credit, the team realized this too. They went back to the drawing board to rebuild their infrastructure, and as they watched users create more and more complex apps, they rearchitected the AI around a new agent that could keep up with this complexity. They renamed and relaunched the product, and thus Anything was born.

Four characteristics — (1) seeing the above game film on Dhruv and Marcus over the past year, (2) a breakthrough product with compelling signs of product-market fit, (3) a thesis area of deep interest to Footwork, and (4) an expansive vision to build a massive company — made us rethink our decision from a year ago to quickly lead this round.

Anything has built a jaw-dropping customer experience. Type in a few words, create a mobile app or website or tool, tweak it, add payments or whatever else you need to have it be usable, and launch it. The product-market fit has shown up in the rapid growth to millions in annualized revenue, but more importantly, in the 120% net revenue retention in their first month post-launch, due to power users who consistently upgrade and deepen their usage with Anything. On Friday they launched Anything Max, which is an autonomous agent that runs your app in a real browser, identifies issues, and ships fixes and features without hand-holding. The product just keeps getting better, and quickly, for power users.

Last year, we wrote a piece titled Consumer and Enterprise in the AI era, and Anything is perhaps the canonical example of an AI-native company at the intersection of both consumer and enterprise. Its users range from consumers to prosumers to professionals who are becoming entrepreneurs through the Anything platform, akin to Shopify, but where the business they’re creating is the app vs. physical goods to sell online. It reminds us of course of Canva, which has democratized access to design for non-designers. Anything is democratizing access to software creation for non-technical users.

Dhruv and Marcus have an expansive vision, one where much more money is created by the Anything platform than taken by Anything the company. In their words: “The next wave of software will be built by people who never called themselves founders or engineers. That feels inevitable now. Anything exists to accelerate it. Anything will be the way most people make software.” If that excites you, we’re hiring in San Francisco across AI, infra, product, and GTM. Come join us.

Waldo is building the strategy engine for agencies and brands: an AI-powered platform that turns the messy, multi-week process of research, ideation, translation into creative, and more, into minutes. At its core, Waldo combines proprietary data integrations (from social platforms, ad libraries, syndicated research, and more) with over 50 structured workflows built for strategists. The result: better briefs, faster pitches, sharper insights. It’s not just another AI wrapper — it’s strategy-grade AI, designed with the workflows and judgment calls that agency strategists live by.

We first got to know Justin Wohlstadter, Waldo’s founder and CEO, over a decade ago when he was building Wonder, a research company that became a trusted partner to firms like McKinsey. That journey taught Justin a critical lesson: pick a customer, and build obsessively for them. With Waldo, he chose the creative strategist as the ICP — the lynchpin role inside every agency. His product intuition and persistence, combined with a tight-knit team he’s worked with for years, including Jesse Rogers, Head of Engineering, shine through in Waldo’s velocity and customer love. Waldo has caught the attention of the biggest agencies in the world, with most of them, such as Havas and IPG, as well as brands like Conair and Kettle & Fire, already customers. Check out the video below to learn more about the product and story from Justin himself.

Our conviction in Waldo rests on more than just the quality of the team, product, and customer base. The company sits at the intersection of two powerful shifts: (1) marketers and brand teams under pressure to do more with less, and (2) AI helping to tease out better, deeper, sharper insights that are proactive and delivered in a scalable way. The opportunity is massive—agencies represent over $200B of global spend, and Waldo's brand-side pull only expands the TAM. Add in proprietary datasets, data network effects, customer workflow IP, and an easy-to-use UI—built specifically for marketers, strategists and creatives—and Waldo has the ingredients to become the operating system for all GTM teams.

We could not be more excited to partner with Justin, Jesse, and the rest of the Waldo team. Agencies are using Waldo to pitch better and win more business, wow their clients, grow existing accounts, and speed up day-to-day research and insights work. If you realize you should be one of them — or more broadly, if you’re a strategist, marketer, or brand-builder looking to supercharge your creative process — check out waldo.fyi or please reach out.

Dermatology is one of the most striking examples: more than 100 million Americans live with skin conditions like acne, eczema, and psoriasis, yet only a fraction ever get the care they need. Consumers spend billions on treatments that don’t actually work, while their conditions get worse and scarring sets in. There aren’t nearly enough dermatologists to service patient demand (in fact, we only graduate 500 per year in the US, and more than half end up doing cosmetics), getting an appointment thus takes months, and so 9/10 patients never see a dermatologist. 1/3 of the patients who even get to see a derm and who are prescribed a treatment for their skin condition don’t end up picking up the prescription because of issues with insurance reimbursement and the cost of these medications. Patients end up suffering for years with skin conditions that are entirely treatable.

When David Futoran experienced this pain firsthand, through his own skin journey with chronic acne, he knew he had to do something about it. He called up his dermatologist, the renowned Joel Spitz (just ask a dermatologist friend if they know of him!), who had changed his life through prescribing Accutane to solve his acne. Together they co-founded Honeydew, which has reimagined the experience of treating skin conditions, and is on a mission to ensure that no one in the world has to suffer from a treatable skin disease because of an access issue ever again.

As a Honeydew member, you can quickly connect with a licensed dermatology provider, get a comprehensive treatment plan (that may include prescription medication), and get ongoing support from their care team that tracks your progress and makes sure your symptoms improve. And it works — Honeydew has helped over 20,000 patients, with awesome outcomes, very high customer NPS, and rave reviews.

At Footwork, we’ve been thinking about the consumerization of healthcare since we started the firm. We love businesses that democratize access to services that were previously inaccessible, and have a strong affinity for consumer subscription business models. So when we met David and Joel last year, through an intro from co-investor Rex Woodbury, we were instantly enthralled with Honeydew’s business. Our excitement has been backed up by our experience working with the team since — they are maniacally focused on delivering a great patient experience, meeting patients where they are already looking for help, enabling clinicians to enjoy and see impact in their work, and leveraging AI to improve the service for all.

Partnering with Honeydew has felt a lot like partnering with consumer companies like The Farmer’s Dog and Stitch Fix from the early days. For one, the metrics are too compelling to ignore — check out the active subscribers growth chart above, which has happened with very little marketing spend. But the positive impact on real people (okay, and dogs in TFD’s case!) is what gets us most excited. Like those companies, Honeydew has the chance to be a membership for many years for many of its customers, which we see already through flat-lining retention in older cohorts. The skincare journey, just like the dog-food-eating and personal styling journey, is lifelong.

Honeydew’s ambition doesn’t stop at dermatology. The team imagines a world where (1) clinically-backed expertise is readily available (and eventually free), (2) treatment is attainable, and (3) results are down to a science. They’re building that future, first with life-changing prescription treatments and ultimately for the entire skin industry. If you’re as excited as we are about this vision, learn more about Honeydew’s story in the video below, and come join us.

I started next big thing to share unfiltered thoughts. I’d love your feedback, questions, and comments!

Leave a comment

👇🏽 please hit the ♥️ button below if you enjoyed this post.