February, 2026 - Cup of Zhou

February 25, 2026February 24, 2026

AI-Native VC Firms

There’s been this term that’s been thrown around as of late. By GPs. Funnily enough, not by a single LP so far. The “AI-native” venture firm.

Somehow everyone seems to think they’re the only ones doing so. But they’re not. While not everyone approaches it in the same way—and the definitions do get muddled a little—there is a growing audience of GPs building a firm leveraging agentic tools.

So, inspired by a recent conversation with my man Arkady, here’s me offering my definition of what an AI-native VC firm is. It’s a set of workflows that even the large incumbents cannot replicate by adding more people to the problem, despite the fees they draw. It’s where technology is at the core of the firm, as opposed to humans. It’s where experience actually becomes an inhibitor to investment innovation. Call it the legacy tax. Or experience tax. Or as Bret Taylor puts it, the strategy tax.

In a recent Uncapped episode, Sierra‘s Bret Taylor put it this way when speaking of SaaS business models. “In these moments of big platform shifts, what were your strengths can become weaknesses. […] You start to say that ‘Well, I don’t want to just start from scratch. We’ve got all these assets. So how do we do it in a way that takes advantage of all of our assets.’ And so all of a sudden, you’re like, ‘Let’s not just build a great product. Let’s transition from this product to that product.’ […] ‘That’s our strength, we should play to our strength.’ And you start to basically make all these decisions that sound sound very clever because you’re playing to your strengths. And in practice, if the technology wave is bigger than the category, which I think the web was as an example, you end up chipping away at doing a pure play value proposition.

“It can also work with business models though. In that time, you have perpetual license software and moving to software-as-a-service, that’s a huge change for a business to make. For your customers, that goes from being CapEx to OpEx. For you as a company, it changes ratable revenue. Shantanu did this at Adobe. Very few companies can make that transition. You have to sell it differently. You have to compensate salespeople differently. Revenue recognition is different. So you have the product strategy tax. You have the business model strategy tax. You have even the incentives of sales people—there’s a strategy tax. […] All the advantages that you had, all of a sudden, become anchors that are holding you back from doing the right thing.”

But I’m probably not the only one that can see the transposition of venture models on this analogy of SaaS models. I’ve also been in a few rooms where LPs are starting to slow to halt their pacing into investing in AI companies and funds. If so many of these products can be built overnight with Opus/Cowork, Codex, Cursor, Replit, Base44, Emergent, OpenClaw (and all of its clones in the last 2 weeks) and the list goes on, how many of these application layer AI companies will be stripped of their value almost immediately. And likely, in one month’s time from writing this piece (if not within the next few weeks), the list may already be obsolete.

In fact, I was catching with my LP friend (who’s not technical) last week and he used one of the above tools to build a portfolio management tool in two hours that has more functionality than what I’ve seen from companies trying to solve the exact same problem who’ve raised up to 9-figures. And I wish I could say I was joking. He told me, “Tools for us have historically been limited by the marriage of domain expertise and technical expertise. Most of us didn’t have both. But with these tools, they solve the technical expertise part, which empowers domain experts to build their own tools. So why bother paying for any other tool that doesn’t have the same depth of data that I’ve accumulated across decades?”

Henry Ford has that line many of you are probably aware of, “If I had asked people what they wanted, they would have said faster horses.” Most investors—big and small, multi-stage and emerging, generalist and specialist, solo GP or partnerships—are building faster horses.

So an AI-native firm has to reimagine the way the venture business is done. That isn’t an AI-written memo. That’s not just an agent strapped onto a data scraper. Assume everyone at some point can discover and find every company out there. Today’s firms who claim being AI-native (in my anecdotal experience) are highly focused on sourcing. What pools of data aren’t actively being collected now that software can relentless dig into. Today’s firms aided by AI are leveraging tools to make more informed investment decisions. The picking. It still requires, for the most part, a human in the mix to make the final call. There are very, very few, arguably no one, truly leveraging AI nativity to win deals.

In venture, there used to be the classic question VCs would often ask founders: “What’s stopping Google from what you’re doing?” Now it’s “Why can’t OpenAI or Anthropic just do what you do?” Analogously, why can’t a16z or GC do what you do? Historically, the answer, if there is a good one, is tied down to:

Business model and portfolio construction, which is still a valid point if you plan to stay as a first check purist.
Or, it’s come down to the person. You have lived experience that no one else does in cybersecurity or leading healthcare systems. You have privileged relationships no one else has. You’ve built communities that are centered around you as the central node. It’s usually been a “people” answer.

But with AI-native firms, the answer must be technological. Even if large firms were empowered with the idea, why will they still fail in the technological implementation of it? Why can’t these large multi-stage funds:

… “see” the volume you do? Are you pulling data from non-obvious hubs or non-public datasets? How defensible is it in the longer term?
underwrite deals at an earlier stage like you do? Are you doing diligence at scale? How do you get the “truth” from sources that other firms powered primarily by human capital cannot?
provide the portfolio value (at scale) like you do? Do you give your portfolio companies access to the same systems you’ve built? Can they leverage your tools to empower their customers and/or talent?
provide the same liquidity opportunities to LPs at a predictable pace like you can? Not sure if agentic software will be good enough here for now, given how much humans themselves are still figuring this out. But a goalpost for some visible future.
provide their co-investors with the same value as you can? This one might be hard. There’s always the balance of what is unique to you versus unique to others. The more you share, potentially, the less of an edge you’ll have. But my stance has always been, if access to certain analyses or information will be inevitable at some point, it’s better to be the rising tide that raises all ships than the last drop of rainwater in the ocean.
provide LPs with the same depth and/or breadth of value like you do? This last one is probably not something any AI-native firm I’ve talked to is focused on, but given that all venture firms are marketplaces at the end of the day, answering this question would be extremely powerful.

I’m not saying that’s the only way to do business. That’s not the only way to invest. I’m confident that even when GPT 15.0 and “Opus” 9.4 comes out, there will still be firms that operate primarily around a core set of individuals as opposed to technologies. We have mass-produced cars, but there is still a great demand for hand-crafted luxury vehicles. The same is true for fashion and accessories. So yes, there will still be a world even if they’re not the largest players by market cap where people prefer the luxury of the human touch.

Photo by James A. Molnar on Unsplash

Stay up to date with the weekly cup of cognitive adventures inside venture capital and startups, as well as cataloging the history of tomorrow through the bookmarks of yesterday!

The views expressed on this blogpost are for informational purposes only. None of the views expressed herein constitute legal, investment, business, or tax advice. Any allusions or references to funds or companies are for illustrative purposes only, and should not be relied upon as investment recommendations. Consult a professional investment advisor prior to making any investment decisions.

February 20, 2026February 19, 2026

Are you going to start an advisory?

October last year, I was having dinner with an aspiring GP who, for whatever reason, thought I should do more advisory work. A comment made after I shared that I do very little advisory work. So she asked why.

And I said something to the effect of: My goal in life is to spend time with the most curious, the most ambitious, and the highest performing individuals. In the early days, when advisory is opportunistic and assuming you don’t need the capital, you can choose to work with some incredible people. Over time, as you build a sustainable business, and as this becomes your primary source of income, to pay the bills, you may end up working with folks that you may not have chosen to work if you had the choice. You end of building a fantasy portfolio of, in my case, GPs, that you’ve allocated your time to. The one resource you can never get back. And because I’m someone who likes optimizing different parts of my life, I may very well fall victim to my own optimization of being an advisor. I would rather not see myself inevitably choosing those circumstances. What scares me is not the work but the person I will end up becoming.

Just earlier this week, I had another conversation around the same topic with a GP I deeply respect and have chosen to work with. So I thought it seems to be time I share this publicly.

Many of my contemporaries have built robust businesses for themselves being advisors to GPs and LPs. And I think it’s a beautiful thing. The world needs more great advisors. We always seemed to be starved of them. The world needs more people who are willing to pay it forward. To share their lived experiences with those who have yet to live. But I don’t think I could ever do what they do.

To me, this blog and my podcast are what I need as outlets to help the world. Two things that will always stay free. Although for my podcast, many a time I have resisted the temptation to create a paid product to keep the podcast’s lights on. I hope good and useful knowledge continues to stay that way. Free. Through that, the frameworks and lessons I’ve come across and/or use.

But the reason I don’t think I could ever do what some of my advisor friends do is because I think a lot about optimization. And in the theme of optimization, I will take more opportunities than I would like. But I’m also really bad at breaking up. And so to not put myself in that situation, it is better to not begin.

Have I advised folks? Yes. Will I continue to? Probably. Opportunistically. Will I still say no to most advisory opportunities even if there is money to be made? Yes.

Does that mean I build an advisory practice? No.

If you’ve been a long-time reader of this blog, you’ll know I’m a deeply flawed individual. I don’t claim to be perfect. And I’m definitely not a profit maximalist. Although I do wonder what kind of person my alter ego would be. But there are a specific set of choices that I believe I can make so that I live my most fulfilling life. And one of those choices is choosing the people I get to spend time with. So if I were to do any advisory, it’d only be with people I deeply respect AND can learn something from them as much as they from me.

Photo by Thea on Unsplash

Stay up to date with the weekly cup of cognitive adventures inside venture capital and startups, as well as cataloging the history of tomorrow through the bookmarks of yesterday!

February 13, 2026February 12, 2026

On 2026 Liquidity

“This is going to be exciting year for IPOs.”

I’ve heard this statement more than a few times since the beginning of this year and even late last year from GPs. I’m no public market or IPO expert so I’ll refrain from commenting on it as if I am. But as a layman, speaking to other LPs who are less lay in this regard, we have about $3 trillion of pre-IPO tech companies that will really make a splash. Plus or minus half a trillion. Figma’s IPO was held in bated breath. 6 months out, post-lock-up, they were down from their IPO price of $33. A great outcome for early stage investors. TBD for people who came in right before the IPO. Whether that’s going to move and open markets or not, that’ll be for people smarter than me to opine on. But the truth is we’re going to see some realization this year.

And the conclusion drawn from public markets creating liquidity is that there will be more capital moving to emerging managers, which naturally, emerging managers are excited about. Yes and no.

Yes, in the sense that early employees will have more liquidity to invest in friends. High net worth individuals who come in at a Fund I, maybe Fund II. So yes, if you’re an emerging manager with friends at these soon-to-be public companies, great. If not, you’ll have to hustle just like everyone else.

Yes, also in the sense that any institutions that are holding equity directly or indirectly in these companies will have a payday. But when they reallocate that capital, it’ll go back into their existing portfolio construction. Most of which isn’t for venture. Many more institutions are not at the growth phases of their portfolio construction, so allocating to high growth, high risk assets are not at the forefront of their mind. Last year, I heard more conversations on who to re-up on in their existing portfolio, as well as chatter on LBOs/PE funds and credit funds than I have on venture. So maybe, in the optimistic case, they do 1-3 more net new checks to managers this year, maybe next. But the uncertainty with the “AI bubble” has many allocators hedging their AI bets in large AI companies with other asset classes.

And so, in that sense, no.

Assuming we have a number of large companies go public this year, we still won’t have the exuberance of 2020 and 2021 in emerging managers. The dollars that are made liquid will just go into the funds that had those outliers in their portfolio, not new managers. And even if they do find the budget to allocate to net new, I don’t think it means they hire a new team member specifically to focus on emerging managers where they write smaller but more checks. The diligence process is still largely the same regardless of check size. And unfortunately, that also means more taxes, K1s and reporting to do if they have more in the portfolio. Speaking anecdotally from institutional allocators I’ve been lucky enough to pick their brain on in the last few years, emerging manager complexity is just not worth it for many allocators. Even through a fund-of-funds vehicle. If you’re reading this blog post, you’re probably well aware of the fundraising environment for fund-of-funds, both standalone ones as well as those part of larger venture or secondary organizations.

That said, I’m personally still bullish on emerging managers, which is where I choose to spend my time and energy. This blogpost is in parts a sanity check, but also a reminder to myself to not believe the bubble in which I live in is what the world thinks.

As the great Richard Feynman once said, “The first principle is that you must not fool yourself and you are the easiest person to fool.”

Photo by Austin Neill on Unsplash

Stay up to date with the weekly cup of cognitive adventures inside venture capital and startups, as well as cataloging the history of tomorrow through the bookmarks of yesterday!

February 6, 2026February 6, 2026

Will AI Take Over LP Investing?

I recently watched Brandon Sanderson’s keynote on whether AI is art or not.

It’s a great talk. And I highly recommend you check it out even if you don’t work in the creative industries.

We’ve seen writer strikes in Hollywood as well as a proliferation of AI use cases in creative industries. James Cameron joined the board of Stability AI. The Russo brothers behind Marvel’s superhero sequel prowess have created their own AI studio. Pouya Shahbazian as well, using AI over the next four years to create 30 AI-generated films. The list goes on. As such, the question of “Is AI art?” is an interesting one to answer. And admittedly, harkens to a series of conversations I’ve had with allocators on leveraging AI in investing practices.

From the VC/GP side, there are folks like Yohei, Sarah, and Ben and Matt, just to name a few who’ve all been building and incorporating AI recently into their workflows and deal flow pipelines. Yes, I know I’m missing a lot more names. But you get the point. From the LP side, progress is still slower, but many younger LPs are quickly adopting AI as well. The conversations I’ve had come from senior allocators on whether it makes sense to use AI. And if so, how much?

Which begs the question: If AI can do your job, do you still have a job?

I was at a dinner last year where the CIO of a large endowment shared that the reason she knows what to look for in managers today, how to underwrite funds, and how to build a venture portfolio was due to the fact she made a plethora of mistakes on her way up the allocator ladder. Small mistakes, like mixing up a decimal on the spreadsheet which led to a venture fund needing a $10B outcome instead of a $1B outcome. Or like Jamie Rhode once said on Superclusters, that she failed to check before she made a commitment to a fund if the fund actually had the commitments that the GP advertised, leading to her check being a larger proportion of the final fund size than she anticipated.

A lot of senior leaders in the LP space seem to be quite skeptical of what AI can do for investment decisions in its current state, yet junior team members seem to widely adopt it to write memos, to inform investment decisions, to create portfolio construction models, and so on. And so far, there’s been a general consensus that AI, at least with respect to investment decision-making, has yet to reach its desired state. In one comment at the same dinner, a senior allocator remarked that one of her direct reports submitted a fund construction model that was built via AI and suggested that in order to return the fund, they needed almost a quarter of the companies to become unicorns. And when questioned, the junior allocator saw nothing wrong with the model. Only to further defend their choice. Or as Brandon Sanderson says in the talk, the problem with AI “is because they steal the opportunity for growth from us.”

“The process of creating art makes art of you. My friends, let me repeat that. The book, the painting, the film script is not the only art. It’s important, but in a way, it’s a receipt. It’s a diploma. The book you write, the painting you create, the music you compose is important and artistic, but it’s also a mark of proof you’ve done the work to learn because in the end of it all, you are the art. The most important change made by an artistic endeavor is the change it makes in you. The most important emotions are the ones you feel when writing that story and holding the completed work.

“I don’t care if the AI can create something that is better than what we can create because it cannot be changed by that creation. Writing a prompt for an LLM, even refining what it spits out, will not make an artist of you because if you haven’t done the hard part—if you haven’t watched a book spiral completely out of control, if you haven’t written something you thought was wonderful and then had readers get completely lost because your narrative chops aren’t strong enough, if you haven’t beat your head against the wall of dead ends on a story day after day until you break it down and find the unexpected path—you’re not going to have the skill to refine that prompt. The machine will have done the hard part for you and it doesn’t care.”

Growth comes from making mistakes. It comes from the struggle. The “distance travelled,” to borrow a term Aram Verdiyan used before. This is why investors often prefer partnerships and co-founderships. It’s why many firms have “red teams.”

There is probably a day when AI can do our job. But for now, the art of investing is in the friction it takes to make a decision. The character-building moments. The moments where you question your own priors. So if AI enables you to have more nuanced dialogues with yourself, if it challenges the way you think in ways you hadn’t considered before so that you look for evidence that either proves or disproves the null hypothesis, then there will still be room for the use of AI in investing. Otherwise, if you’re regurgitating scripts based on singular uninspired prompts, then you likely won’t have a job for long.

Stay up to date with the weekly cup of cognitive adventures inside venture capital and startups, as well as cataloging the history of tomorrow through the bookmarks of yesterday!

Share this:

Share this:

Share this:

Share this: