@HarryStebbings
🎤 @twentyminutevc, 🏦 @20vcfund, @projecteurope_😇 @fuseenergy @linear @wearelegora @factoryai @lovable @airwallex @mercor_ai @workos
Why Wix customers won't churn to vibe-coded solutions: "Most small businesses aren't going to vibe-code their entire software stack. Running a business requires complex workflows that take years to build, not just generating code with AI. The real challenge isn't creating an app, it's recreating the deep business logic that platforms like Wix have already developed." @Avishai_ab How do you think about this @antonosika @matanSF @tobi
How public markets are getting both Wix and SaaS wrong: "The market is underestimating what makes the best SaaS companies valuable. The real moat isn't just the software, it's the trust, data, and relationships built over years with enterprise customers. You can recreate features with AI, but you can't instantly recreate the trust that companies like Salesforce have earned." @Avishai_ab How do you think about this @benioff @nikesharora @dharmesh
First term sheet signed for an Israeli company where 20VC is lead. 15% ownership. Two of Israel’s greatest operators. Mega market. True pre-seed. LFG.
"An unconventional sleep schedule can be a competitive advantage for a CEO. Having a few uninterrupted hours when everyone else is offline creates space for deep thinking, planning, and high-leverage work. For leaders, protected time to think is often more valuable than another day full of meetings." @Avishai_ab Cmon, @bryan_johnson hit us, what is the optimal sleep schedule for founders????
The SaaS bloodbath has been beyond brutal. Wix for more than any other. This company is doing $2.1BN in ARR and they are valued at $2.1BN. Like WTF. And, they have Base44, one of the leaders in vibe coding, now doing over $170M in ARR. What is going on? How is this being priced at $2.1BN? I sat down with @Avishai_ab and have shared the biggest lessons below. 1. How Public Markets Are Getting Both Wix and SaaS Wrong Public markets routinely misprice tech cycles, hyping vibe-coding startups while discounting core SaaS engines. They ignore the moat of enterprise trust. Giants like Salesforce win because institutions trust them with sensitive data, creating security and compliance barriers automated code projects cannot easily replicate. 2. Was the Buyback a Terrible Decision? Timing the stock market is a fool’s errand. If your balance sheet has cash and your team is focused on core products instead of distracting M&A, a buyback during a dip can be rational. Capital allocation should be judged over three years, not a three-month panic. 3. How Do You Keep Talent When Your Market Cap Is Cratering? Stop treating talent attrition as a failure. It can be a healthy refresh. Bull markets make companies forget that elite teams are forged by solving hard problems, not coasting. When market caps fall, protecting your top tier can uncover hidden stars and make room for a hungry new generation. 4. How Do the Business Models Compare Between Wix and Base44? Scaling newer AI platforms requires ruthless focus on unit economics over expensive, generic LLM usage. The margin breakthrough comes from fine-tuning and combining smaller, custom models to match top-tier performance. Within two years, infrastructure cost declines could make AI COGS a non-issue for platform margins. 5. Why Wix Customers Will Not Churn to Vibe-Coded Solutions Mainstream SMBs have neither the desire nor capacity to build their own tech stacks with automated coding tools. Seemingly simple workflows, like vertical business logic for a hair salon, are surprisingly complex. Even elite engineering teams struggle to rebuild them quickly, proving vertical SaaS remains highly defensible. 6. Why Teams of the Future Will Not Be as Small as People Think The idea that future tech giants will run on skeleton crews managing thousands of AI agents is overhyped. The ecosystem gives too much credit to standalone AI capabilities. On granular business workflows, models still stumble, which means dense, capable human teams remain essential. 7. I Would Be Terrified if I Were in University Today, and My Advice to Students Entering the workforce today is uncertain, but current LLMs will not instantly replace white-collar jobs. Today’s models are strong at reframing data, not deep reasoning, and still make frequent mistakes. True disruption to human capital likely requires two or three more major breakthroughs. (links below)
Spotify 👉 open.spotify.com/episode/4Tc1M3… Youtube 👉 youtu.be/TdQyVXN0GF8 Apple Podcasts 👉 podcasts.apple.com/us/podcast/20v…
Meta Deepseek Anthropic OpenAI All developing their own chips. Forgive me, I am a moron, does everyone have to own the full stack? @AnjneyMidha @jonathanross321
"When companies don't have the best product, they often compete through services instead. Strong enterprise relationships let them sell consulting and implementation, helping customers adopt technology built by others. It's a strategy IBM pioneered, and one that's becoming increasingly common in the AI era." @rodriscoll Does a services-led strategy simply imply inferior product and movement from innovator to incumbent @matansf @ceo_clickhouse @apoorv03 @sundeep
Every week I speak to 3-5 LPs about various different things. They all ask; which manager of the last 2-3 years has broken out above all others? So easy. - @saranormous beyond crushed with Conviction: - @MattEvantic is in literally every 🚀 in Europe and US in last 12 months. Mega respect to both. Incredible.
"If AI really replaced half of all white-collar jobs, a small redistribution tax wouldn't come close to addressing the consequences. The scale of economic and political disruption would require a far bigger response than a token solution. The broader point is that if you truly believe the most extreme AI displacement forecasts, you also have to believe today's policy proposals are nowhere near sufficient." @rodriscoll Do you think this is right and fair @altcap @mmurph @deedydas @kevinweil
It is easy to invest money. It is harder to raise it. It is hardest to return it. A lot forget this when pontificating BS on X.
If you are in tech, this is the one show you have to listen to every week. Value per minute, this will give you a better ROI than anything else you can listen to. AGENDA: - Sam Altman Offers Trump 5% of OpenAI - Alex Karp Sounds the Alarm: Enterprises Fear Frontier Models - Deepseek Building Own Chips: The Rise of Chinese Open Source? - Meta Compute Launches Cloud Business My notes below on this weeks show with @jasonlk & @rodriscoll: 1. Are All the Hyperscalers Adopting Consulting Strategies Because Their Product Isn’t Good Enough? Tech giants fall back on deep enterprise relationships when pure product adoption stalls. Rory calls this the classic IBM Global Services playbook: monetizing trusted distribution to deploy other people’s technology. With Microsoft embedding 6,000 engineers for AI pilots, the industry is shifting from pure software toward intensive change management. 2. If You Think 5% Is Going to Make Up for the Job Losses From AI, You’re Having a Laugh Tech platforms trying to rewrite the U.S. tax system are detached from political reality. Rory notes that if AI replaces 50% of white-collar jobs, a 5% corporate equity stake amounts to only about $140 per person. Believing a minor slice of equity will pacify mass economic disruption is delusional. 3. Why Would Any Great Employee Join a Company That Does Not Offer Tenders? The talent war has moved beyond illiquid paper wealth. Jason emphasizes that top-tier operators now demand a clear path to secondary liquidity within 24 months. As companies like OpenAI offer regular tender options, elite builders refuse to get locked into stagnant equity stacks. 4. Why AI Frontrunners Are Actively Welcoming Washington Oversight AI frontrunners are flipping the old tech playbook of “leave us alone” by actively volunteering for regulatory alignment. Jason views OpenAI’s 5% equity offer as a masterful communication play. It sets government ownership expectations low before political pressure forces a harsher outcome. 5. The Extinction of Downstream Valuation Fear for Modern Founders Founders no longer fear aggressive valuations the way they once did. In past cycles, builders avoided high prices because late-stage investors held predatory downside protections. Today, growth investors often accept 1x returns on failed bets without drama, reducing structural fear and increasing founder velocity. 6. Meta Compute Launches a Cloud Business to Sell Access to Its AI Infrastructure Overbuying compute to build proprietary models and then renting it out as hosted infrastructure is the ultimate hedge. Meta’s 10% stock surge mirrors SpaceX’s infrastructure monetization playbook. Hyperscalers are learning that becoming a neo-cloud utility can be the perfect Plan B while raw demand remains tight. (links below)
Spotify 👉open.spotify.com/episode/1YnExW… Youtube 👉 youtu.be/LUkbZHSQIog Apple Podcasts 👉 podcasts.apple.com/us/podcast/20v…
I remember investing in Mercor at a $2BN valuation. Some very smart investors told me what a high price that was. Yesterday, they announced $2BN in revenue. This is just the start. This space is so much bigger than anyone sees. Excited for them to hit $20BN in revenue soon. 😜
Every day I walk past this in my gym and it really pisses me off. Nice idea but you don’t get a medal for participating. You need to win and you can leave nothing on the field. That is the message that should be here.
If you are not building a personal brand as a CEO today you are leaving alpha on the table. It’s that simple. Customers, investors, hires. Will all come because of your brand. Start today.
"The biggest lesson in venture is simple: don't project your own ideas onto the founder. Even if you think you know the right strategy, it's the founder's company, not yours. The best investors back teams because they believe in their judgment and ability to execute, not because they expect them to build the company the way they would." @mignano What is your single biggest lesson from 10 years investing @kirbyman01 @rabois @rebeccakaden @infoarbitrage
"Most enterprise workflows don't require frontier AI models. Outside of coding, tasks like summarization, document generation, and briefing can often be handled effectively by lower-cost open-source models. The real value is knowing when you need frontier performance and when cheaper models are more than good enough." @mignano What percent of enterprise workflows requires frontier models today @nikesharora @matansf @ceo_clickhouse @lqiao
"To be great at seed investing, you have to put your ideas into the world. Sharing your thinking signals to founders what you believe in and what kinds of companies you're looking to back. Content isn't just marketing, it's a way for the best founders to find you before everyone else does." @mignano Love to hear your thoughts on this @packym @jasonlk @honam @TurnerNovak
I missed out on $100M+ by passing on the seed rounds of Suno and Granola. Mike @mignano did both. He is the single best example in the last 5 years of an operator successfully turning into a GP at a large fund. It is so hard to do. He crushed it. He has been a friend for 10 years and we shoot the shit today on the biggest questions of the day. I summarized my notes from the chat below. 1. Do You Have to Produce Content to Do Seed Well? To excel at seed, you need to put your ideas into the market so founders know exactly what you are looking for. Publishing your theses acts as a bat signal for early-stage founders before they even launch a product. Success comes down to building a great network, sharing your point of view, and betting on people. 2. Single Biggest Lesson About Venture The biggest lesson in venture is to avoid projecting your own operator ideas onto a founder. It is their company, not yours. Making investment decisions based on your personal roadmap for them leads to bad bets. Venture success requires trusting the team’s independent judgment and execution plan. 3. What Percent of Enterprise Workflows Actually Require Frontier Models Versus Open Source? Roughly 80% of non-coding enterprise tasks can run well on non-frontier models. Standard workflows like summarization and document generation are well suited to open source, which is improving quickly. This lets enterprises optimize token budgets without sacrificing meaningful performance. 4. Why We Believe Suno Has Unlimited Upside Suno has unlimited upside because it unlocks a new consumer behavior: creative entertainment. Users generate music simply for the joy of creation, with no need for distribution or monetization. Like YouTube or TikTok, platforms that democratize a creative medium can create immense, generational value. 5. Obliterate, Don’t Automate Avoid backing software that merely automates existing workflows to make old businesses slightly faster. The true generational opportunities come from completely reinventing industries, like putting an AI doctor in everyone’s pocket, rather than selling incremental efficiency to a corporate middleman. 6. The Shift From Infrastructure to Applications As the capital-intensive AI infrastructure buildout matures, massive value creation is shifting to the application layer. Similar to the internet after the broadband boom, we are entering an explosion of specialized software. Winning requires an opinionated thesis, early conviction, and exceptional speed. 7. Why Startups Must Maximize Token Spend While incumbents restrict budgets, startup CEOs should maximize token spend. Frontier models give lean teams a critical advantage over slower giants. Handing low-level tasks to agents preserves small, hyper-efficient engineering teams and keeps them moving at winning speed. (links below)
Spotify 👉open.spotify.com/episode/0ixO7r… Youtube 👉 youtube.com/watch?v=Gjjnb8… Apple Podcasts 👉 podcasts.apple.com/us/podcast/20v…
Young Companies Have To Be In Person: "In particular for a young company, I think it is very challenging to build a culture, as a remote company. It feels great to be part of an amazing team, and it's different when your connection to that amazing team is via a Brady Bunch of Zoom squares. We have rituals that we've developed that we only would have developed if we were all working in person." @claybavor Love to hear your thoughts on this @typesfast @howard @dhh @BrendanFoody
"The unfair advantage young people have coming out of university; you've just had four years to spend unlimited time ... to master of these AI tools. A thousand companies would love to have you infuse what you know into how they're doing things. I can't remember a time when a young person with no work experience, but with the right mindset and experience using some of these tools, has ever been so valued. Some of our most effective employees are 22 or 23 years old and have been completely AI-pilled and have a comfort and facility with these tools, that many of our more experienced folks don't." @claybavor Do you agree and single biggest advice to young people coming out of university today @matanSF @nikesharora @jasonlk @cory
The greatest joy of my job is building friendships with the smartest people on the planet. I have interviewed 1,000+ of the best CEOs over the last decade. @nikesharora is one of the best (alongside Nik Storonsky, @alanchanguk). Nikesh's story is insane. He came to the US with two suitcases, $200, worked as a security guard, took notes for students and flipped burgers to pay tuition. Today, he is CEO of Palo Alto Networks, a $225BN cybersecurity giant. He is a dear friend and we sat down to chat the other day and he said he was happy for me to release my biggest lessons from the discussion. 🚀 6 Lessons from Nikesh Arora on Building a $225BN Company 1. Product Builds Brand. Not the Other Way Around. If you build a great product and execute well, brand follows. If you have a great brand and the product is poor, the brand dies. Sun Microsystems had a brand. Yahoo had a brand. It did not save them. In tech, brand is not a substitute for product. Brand is the residue of product excellence. 2. Consumer AI Is About Breadth. Enterprise AI Is About Depth. Consumers tolerate false positives. If ChatGPT gets something slightly wrong, you correct it, ignore it, or move on. Enterprises cannot do that. If an AI agent is making independent decisions in a company, false positives are catastrophic. The future of enterprise AI is deep context-rich intelligence. 3. Most Enterprises Are Using AI Wrong. The real opportunity is to rethink the workflow from first principles and let AI do 80% of the thinking. The winners will not be the companies that sprinkle AI on old processes. The winners will be the companies that redesign the company around AI. 4. SaaS Had No Opinion. AI Applications Will. SaaS applications are containers. You define the input. You define the output. AI applications will have opinions. That changes everything. Nikesh thinks many G&A functions — marketing, finance, HR — could have half the people over the next three years because AI will make the average employee dramatically more capable. 5. AI Transformation Has to Be Top-Down and Competitive. Nikesh runs a meeting twice a week called AIIO. The point is simple: get the top technical leaders in the company in one room and force the conversation. What are you doing with AI? Why are you building this? How are you using tokens? How are you changing the product? What have you done in the last three days? The brilliance is not the meeting. It is the social pressure. That is how you transform a 21,000-person company. 6. In Technology, Miss One Trick and You Survive. Miss Three and You Are Obsolete. Miss one trick, you can survive. Miss two tricks, you are partially impaired. Miss three tricks, you could be obsolete. That is the CEO mindset in 2026. Paranoia is not optional. The speed of learning is now the speed of survival. Thank you my friend. Forever learning from you. (Link in comments.)
YouTube: youtube.com/watch?v=v4GN1q… Spotify: open.spotify.com/episode/6LYUjX…
I would say how I feel about this but I am legitimately nervous I will get in trouble for doing so… This is a sad state of affairs. x.com/themerciannews…
Of all new media, Tara is the single greatest talent. The energy, the charisma, the obsession. Unparalleled. Watch this space. 🚀 x.com/tarakeeney/sta…
The single best follow in tech on X right now is @nikesharora. One of the best operators of our time spewing wisdom on a daily basis. x.com/nikesharora/st…
Dario just declared war on open-source. Anthropic's message is clear: open source could destroy the entire AI business model, and Chinese open-source models are the cause. I sat down with @jasonlk & @rodriscoll to discuss it, along with the biggest news in tech this week: - Anthropic & Dario Declare War on Open-Source - Coinbase Slashes AI Spend 50%: Is the Token Bubble Bursting? - Kalshi's $40BN Valuation & Impending IPO - Bending Spoons: The Smartest IPO of 2026 & the $100BN SaaS Roll-Up Play My notes below: 1. Anthropic Is Laying the Foundation for a Deal With the U.S. Government on Chinese Models Anthropic accused Chinese open-source AI competitors of “brazen theft” through model distillation that violates its terms of service. Rory noted the hypocrisy, given that U.S. frontier models originally scraped external IP. Still, Anthropic appears to be laying the groundwork for a regulatory trade: complying with domestic access restrictions in exchange for a federal ban on distilled Chinese models. 2. Jason Lemkin’s Response to Brian Armstrong’s AI Tweet Jason dismissed Coinbase CEO Brian Armstrong’s 50% LLM cost-cutting post as “performative social media,” arguing that savings matter little if core revenue is flat or shrinking. He believes AI must actively drive top-line growth. Rory defended the move as “cost management 101,” saying cash-conscious enterprise executives will quickly emulate it to curb runaway frontier model fees. 3. CEOs Are Struggling to See the ROI From AI Massive enterprise spending on AI tokens is failing to deliver the expected revenue or productivity gains, leaving CFOs searching for measurable operational lift. Jason noted that adding millions in AI spend can still produce the same growth rates as prior quarters. Rory argued that AI spending must clearly accelerate software delivery or create definitive bottom-line savings as boards push back on reckless “token maxing.” 4. We Are All So Aligned in Wanting AI to Win Jason warned that the U.S. economy is structurally addicted to AI, with 40% of the S&P 500 tied to the boom, making society eager to prop it up to protect 401(k) portfolios. Rory countered that protectionism artificially inflates intelligence costs for the broader economy. He compared it to banning IBM PC clones in the 1980s just to protect IBM’s stock price, calling the blocking of low-cost open-source alternatives “fricking dumb.” 5. Bending Spoons and the New Playbook for B2B Revenue Arbitrage Bending Spoons’ $20 billion public valuation marks a shift toward tech roll-ups that drive profitability through price hikes and cost-cutting rather than organic user growth. Jason predicts this playbook will expand into mature B2B SaaS. By acquiring sticky but underperforming platforms like Marketo, Asana, or PagerDuty and injecting hungry talent, operators can rapidly improve retention and capture massive revenue arbitrage. (links below)
Spotify 👉 open.spotify.com/episode/5B05Zi… Youtube 👉 youtu.be/uINvYTxGCQE?si… Apple Podcasts 👉 podcasts.apple.com/us/podcast/20v…
"After food, energy sovereignty is one of the most important strategic priorities for any nation. History shows that wars are often fought over access to essential resources, and today, energy is at the center of that competition. The more countries can produce reliable energy domestically, the more resilient, secure, and economically independent they'll become." What is more important energy sovereignty or model sovereignty @jameswise @vkhosla @alanchanguk @wolfejosh @traestephens
"America should be doing far more with its abundant natural gas resources. Natural gas remains the cleanest widely available fossil fuel, and expanding access to it could help countries move away from coal while meeting growing energy demand. The opportunity is to pair that energy abundance with distributed power infrastructure to deliver reliable, affordable electricity at global scale." Love to hear your thoughts on this @alanchanguk @shaunmmaguire @rabois @zebulgar
The true complexity of building a data center but why power should not be your constraint. "Building a hyperscale data center is far more complex than just securing power. The real bottlenecks are construction, cooling, copper, equipment, and the skilled labor needed to build it, making 12–18 month timelines the norm. In many cases, power isn't the limiting factor. The physical infrastructure is." What does no one know about building data centres that everyone should know @altcap @DavidCahn6 @JonathanRoss321 @andrewdfeldman
So @harmonic_ai just released their Hot 25 Report on the most in-demand early-stage companies. The top 3: 🥇 @resolveai reclaims the top spot — joining Lovable as the only two companies to top the list twice. 🥈 Salient skyrockets to #2, up 11 spots. 🥉 @aaruHQ rounds out the podium, as rival Simile debuts at #6. Biggest mover on the list: @BrainCo_AI - climbing 12 spots. Newcomers to watch: @periodiclabs at #7, @sandstonehq at #19, and @Jackandjillai at #25. Find the full report here: t.co/BWskVYv5cY
"Governments should be careful not to regulate AI so heavily that they fall behind. If one country slows innovation while others move at full speed, it risks losing both economic competitiveness and technological leadership. The challenge is finding the balance between sensible oversight and enabling innovation." How involved should governments be in the development of AI @rabois @torstenreil @AlistairPullen @Jameswise @vkhosla
In the last 24 hours, I have had 5 founders message me of varying-sized companies; some 10-person startups and one $200BN public company. All of them stated they have been able to cut inference spend by 75% or more with little effort, no performance change and better latency. The times they are a changing.
"When intelligence becomes commoditized, wisdom becomes the scarce asset. AI can provide answers, but it can't replace human judgment, empathy, or the ability to build meaningful relationships. In a world where intelligence is abundant, the people who create the most value will be those with wisdom." If intelligence is abundant, what do you think will be the most valuable asset @AnjneyMidha @omooretweets @kirbyman01 @anneleeskates
"Andy Grove's biggest management lesson was simple: talk to the people doing the work. If you walk the floor, they'll tell you exactly what's broken and what they don't understand. And once you've heard them, stop looking for someone else to give you the answers. Go figure it out yourself."
"There will always be moments when your company faces existential threats. The key is not to avoid those moments, but to refuse to let fear dominate your thinking. If you believe failure isn't an option, your focus shifts from worrying about survival to making tomorrow better than today." Single biggest advice to someone making big decisions that impact the life of their company @AlexHormozi @businessbarista @austin_rief @thesamparr
How to manage regret and learning "Don't replay bad experiences in your mind like a movie. Learn the lesson, leave the regret behind, and focus on what's ahead. Life gets a lot easier when you believe tomorrow can be better than today." Single biggest advice to someone struggling to get over a mistake from the past @pmarca @Codie_Sanchez @SahilBloom @ElishaDLong
When the smartest investor you follow puts 16% of his entire fund into a single stock, you pay attention. That is what Leo Aschenbrenner has done with Bloom Energy. The largest single holding in his portfolio, by a distance. So I had to understand why. KR Sridhar, founder and CEO of @Bloom_Energy. Former NASA rocket scientist. Worked on Mars missions. Then spent 25 years building an energy company almost nobody believed in. The numbers today: $2BN in revenue in 2025. Up 37% YoY. Market cap ~$93BN Market cap. I now completely understand why Leo is so all in. I sat down with KR, took my notes, and have shared the biggest lessons below. 1. The Biggest Management Lesson From Andy Grove The best management advice is simple: walk the floor and talk to your team. They will tell you exactly why things are not working. Do not drown in training binders when you can learn the product by engaging directly. Success requires deep empathy for every part of the business, from shop floor technicians to customers’ real pain points. 2. When Intelligence Becomes Commoditized, What Is the Valuable Asset? When AI makes intelligence ubiquitous, the rarest asset will be wisdom. AI can manufacture information, but it cannot provide happiness, judgment, empathy, or genuine human connection. Because humans are social animals, automated bots will never fully replace real interaction. 3. The True Complexity of Building a Data Center but Why Power Should Not Be Your Constraint A large greenfield data center can take 12 to 18 months to build because supply chains for copper, cooling, and specialized labor are strained. But with modular solid-state technology, power can be deployed faster than the facility itself. Energy should not remain the bottleneck when generation can move to the edge. 4. To What Extent Should Governments Act to Prevent Monopolies? Heavy regulation and government equity stakes can destroy innovative culture and block better startups from competing. Dominant positions often shift naturally over time, as Microsoft and Google show. The best path is a level playing field where free competition lets breakthrough entrepreneurs win. 5. AI Will Do More to Hurt Than Help Income Inequality AI may concentrate enormous wealth among fewer players, but technology has historically been humanity’s greatest equalizer for raising living standards. Intelligence abundance creates a non-zero-sum game that can lift all boats. Leaders still need empathy for the transition generation so they are not left behind. 6. Why Energy Sovereignty Is So Much More Important Than Anyone Thinks After food security, energy sovereignty is the most critical supply chain challenge on Earth. History’s major conflicts have often centered on water, food, and energy. Decentralized power generation at the edge can make communities more self-reliant and reduce exposure to geopolitical conflict. (links below)
Spotify 👉 open.spotify.com/episode/156Hsm… Youtube 👉 youtu.be/p6rjlXLDpcc Apple Podcasts 👉 podcasts.apple.com/us/podcast/20v…
"OpenAI and Anthropic succeeded by avoiding vertical integration, not by embracing it. Their advantage comes from letting cloud providers and infrastructure companies absorb hundreds of billions of dollars in capital expenditure. If they had tried to build the entire stack themselves, they would have tied up enormous capital and moved far more slowly." @rodriscoll Love to hear your thoughts @PatrickToulme @MainzOnX @jietang @vassallo
"The playbook for building the first 100 employees of a startup has fundamentally changed. In AI, speed and intensity have become a competitive advantage, and founders increasingly believe you can't build category-defining companies with low-intensity teams. The harder challenge isn't starting this way, it's transforming older companies whose people and culture were built for a different era." @jasonlk Do you agree that the way we build companies to 100 employees has fundamentally changed @destraynor @typesfast @t_blom @dessaigne
"OpenAI and Anthropic shouldn't be distracted by building their own chips. They've found one of the biggest technology markets in decades, and their focus should be on winning customers, not vertically integrating into hardware. With cloud providers and chip companies competing to supply them with compute, their highest-value use of capital is building better AI products, not designing silicon." @rodriscoll Love to hear your thoughts on this @andrewdfeldman @JonathanRoss321 @AnjneyMidha @mmurph
I could do some bland fundraising congrats tweet for Airwallex and Jack Zhang today with their raising $320M at an $11BN valuation. But I am not going to. The reason I love doing 20VC is because I learn from the smartest minds in the world and they often become friends. @awxjack is a big brother to me, a dear friend, a permanent fixture at every birthday party of mine and a firm favourite of my mother. Jack, I am so proud of you and the team today. This is just the start. LFG. ❤️
“The #3 closed-source LLMs is most in trouble. As enterprises adopt model routing, they're increasingly choosing between the top proprietary models and rapidly improving open-source alternatives. That leaves the number three closed-source provider squeezed from both sides, struggling to justify its place against leaders on one end and cheaper open-source models on the other." @jasonlk Love to hear your thoughts on this @matanSF @ScottWu46 @alexatallah @ceo_clickhouse @eladgil
The AI talent wars have never been crazier. Within 48 hours, Google lost two of their generational scientists to OpenAI and Anthropic. I sat down with @jasonlk and @rodriscoll to discuss it, along with the biggest news in tech this week: - Deepseek Raises $50BN - Wall St's $725BN AI Question - The Rise of Open Source & How it Threatens OpenAI & Anthropic - OpenAI Builds its Own Chip: Jalapeno My notes below: 1. The Number Three Closed-Source LLM Is Most at Risk In the closed-source foundation model race, the number three vendor faces severe pressure as multi-model routing spreads across enterprises. Historically, a third-place software or cloud vendor could survive by being cheaper or simpler. But today, that tier is being squeezed by highly capable open-source models, many heavily subsidized by China, making developers less likely to care about a closed-source number three. 2. The Playbook for Building a Startup to Its First 100 Employees Has Changed Scaling a startup to its first 100 or 200 employees has changed radically from the remote-work era. In today’s hyper-competitive AI landscape, companies built around relaxed schedules and 20-hour remote workweeks will not win. The modern model is lean, elite, highly compensated teams working intensely in person, often six or seven days a week, to survive nonstop product sprints. 3. Why It Makes No Sense for OpenAI and Anthropic to Build Their Own Chips Frontier AI labs have captured the greatest demand engine in modern technology and should focus entirely on winning enterprise customers. Designing custom hardware is a massive distraction when cloud giants like Oracle, Google, Microsoft, and Amazon are already competing aggressively to provide cheap compute. Those infrastructure vendors absorb the capital risk, letting labs prioritize product growth and customer adoption. 4. Why Vertical Integration Makes No Sense for OpenAI and Anthropic Closed-source labs scaled rapidly by staying asset-light and outsourcing heavy infrastructure needs. Their model worked because cloud hyperscalers absorbed hundreds of billions of dollars in cumulative CapEx on their behalf. Forcing backward vertical integration down to the chip level would introduce massive capital liabilities and undermine the operational efficiency that made the model work. 5. Show Me the ROI Next Year Enterprise software is moving from loose AI experimentation to strict financial accountability. Early corporate budgets funded unconstrained “token maxing” to build basic AI fluency, but the 2027 narrative will demand measurable ROI. CIOs will no longer allocate tokens based on compelling pitches alone; they will require verified departmental efficiency gains or clear revenue growth. (links below)
Spotify 👉 open.spotify.com/episode/2ydKtL… Youtube 👉 youtu.be/LDBff24uaDQ Apple Podcasts 👉 podcasts.apple.com/us/podcast/20v…
I have spoken at 100 of the best educational institutions in the world. So many ask: “how do I get into VC?.” I say the same thing; send me three deals per day for 90 days. Simple. Not one person did this. Not one. Until I went to speak at LSE and one incredible student (@NiallKiely20VC) promised me he would. I brushed it off believing he would be like all the rest who gave up after 7 days. Everyday he persisted. His founder recommendations got better and better. He became a part of the team without even joining. Immensely proud to welcome Niall to 20VC for his first week. He earned this with everyday and every founder recommendation. Hustle, persist and never give up.
I have spoken at 100 of the best educational institutions in the world. So many ask: “how do I get into VC?.” I say the same thing; send me three deals per day for 90 days. Simple. Not one person did this. Not one. Until I went to speak at LSE and one incredible student (@niallventure) promised me he would. I brushed it off believing he would be like all the rest who gave up after 7 days. Everyday he persisted. His founder recommendations got better and better. He became a part of the team without even joining. Immensely proud to welcome Niall to 20VC for his first week. He earned this with everyday and every founder recommendation. Hustle, persist and never give up.
Turned down a founder this week. Why? They were finishing the year at $1.5M ARR. Finishing next year at $5M ARR. Today, brutal as it is, that isn’t good enough to raise a good Series A. Opportunity cost of cash is real.
I first met Josh in Greylock HQ. I was 18 and was desperate to get into venture. The only thing; no one other than @mamoonha and @jasonlk took a meeting. He did. He gave me time. Advice. 12 years later we are still friends. 12 years later he is still one of the best operators, minds, thinkers. Can’t wait to find things to work together on now! ❤️
Today is my 30th birthday 🎈 I thought I’d reflect on the decade that has been: 💵 Raised $1BN 🚀 170 investments 🦄 18 unicorn investments 🎙️ 2,500 podcast episodes 🌎 110M downloads I have never been more hungry. I have never been more ambitious. This is just the start. To my 30s. LFG.
Why Remote Work is White Collar Fraud. "I have a three-year-old and a five-year-old. The idea that I could do any work at my house is like a total fantasy. The kids come home at 3pm, your work day needs to keep going. I'm highly against it." @typesfast
Why do we have barely any of our team in San Francisco today? "Our team in San Francisco is maybe 4% of the company, even though that's where we're founded and where I'm based. Cost and loyalty. Retention of good people has always been really hard." @typesfast
Two questions every startup needs to ask on growth: @paulg 1. Are they doing some kind of hack that's unsustainable and gonna stop working? 2. Is the market big enough for it to keep going? "If the answer is no there's no hack, this is genuine growth, and the market is really big, the thing will just keep going." @typesfast
Should founders speak to associates? The collusion in venture is wild. "I never had this attitude that you shouldn't talk to associates, mostly 'cause I'm not a dick. I just assumed those people someday will become partners and they'll remember that I was cool. Within an hour of telling one fund I was going a different direction, I got phone calls from three others. The amount of collusion that happens in VC, founders have no idea." @typesfast Should founders go straight to partners @t_blom @collinmathilde @harjtaggar @FDavidsonT
The SoftBank pitch that got a billion-dollar check. "First time I met him (Masa) would've been at his house in Woodside. Considering the size of the check, I thought it was a surprisingly short meeting. An hour. He pushed us to 'be cheaper than everybody.' Whatever the price of freight is, just be 10% cheaper, and if someone matches you, beat them by 10%. Which is like a terrible strategy. We would've burned so much money. But he just wanted to push you to go bigger." @typesfast
“We have a crisis of open source models in the Western world. Outside of China, there are no good open source models. We don't even have any in the US now. The talent, the capital and the focus to be best-in-class at pre-training, mid-training, post-training is an extremely scarce skillset. The answer for a lot of countries may just be to take an open weights model from China, post-train or fine-tune their own version, and have that be what they start from." @EverettRandle Why does the West have such poor open source models and is it a national security threat to rely on open-source Chinese models @demishassabis @ylecun @aidangomez @lilianweng @janleike
I interviewed @pmarca a couple of months ago and he said something that changed so much of my relationships and life. Ask yourself: “how is this situation all my fault?” Simple as it seems and sometimes wrong, it totally changes mentality, empathy and communication. Thanks Marc, you the man. ✊
How does Benchmark rationalize not being in any of the core model providers? "It f**king sucks. It's a complete and utter failure on our part. You can't be in a situation where you have a chance to make a 30x on scaled capital in four or five years and you don't do that, that's always a failure. It stings especially when all your friends are sending you their implied look-through ownership of Anthropic, OpenAI and SpaceX, eye-watering numbers." @EverettRandle
Another day, another Project Europe team signs a term sheet with a Tier One. 15 of 22 Project Europe companies have raised follow on funding in the first 12 months. The single most successful community in Europe right now.
Every $10BN tech company that interfaces with a consumer in some way will spend $25M+ on sports team partnerships per year. There is an avalanche of money that will flow from tech to sports teams.
Founders don’t speak badly of competitors/incumbents. They have done something right to get to where they are. State to investors 1. Where they have done well? 2. That you have analysed their successes as well as their failures. 3. State a plan for how you build upon their successes and learn from failures. The best founders are students of history.
“Everyone misunderstands the SpaceX valuation. He sells the market on these incredible long-dated call options. If you were just looking at the numbers in the P&L alone, it'd be very, very hard to even get a $2 trillion valuation as fair market value. He says 'I'm gonna figure this massive thing out and that's where the next trillions of dollars of value are going to come from.' Historically he's had such a track record that the market's willing to believe him." @EverettRandle Do you agree that this is where the pricing premium is or is there something we miss @GavinSBaker @shaunmmaguire @altcap @BillAckman
“Everyone misunderstands the SpaceX valuation. He sells the market on these incredible long-dated call options. If you were just looking at the numbers in the P&L alone, it'd be very, very hard to even get a $2 trillion valuation as fair market value. He says 'I'm gonna figure this massive thing out and that's where the next trillions of dollars of value are going to come from.' Historically he's had such a track record that the market's willing to believe him." @EverettRandle Do you agree that this is where the pricing premium is or is there something we miss @GavinSBaker @shaunmmaguire @altcap @BillAckman
“Elon's AI strategy has been the best of anyone. He solved his 'what am I gonna do with this compute' problem not once, but twice. Roughly $2 billion a month of compute from Anthropic and Google, filling the gap in the Colossus revenue stream. Everyone else has spent two years thinking they should probably do something in AI. He's built two data centres, bought a company to fill them, and signed two huge contracts. It's the execution speed that's just so impressive." @rodriscoll
”This is a Rubicon moment in the history of AI. It's the first time that the US has regulated an AI model based on its capabilities.” @EverettRandle joins Harry and @rodriscoll (in person for the first time ever) to shoot the s*** on: - SpaceX Soars to $2.7T - Anthropic's Fable Banned by US Government - Wix and Adobe Hit All-Time Lows - Mistral Raising at $20B and The Case for Sovereign Models - Fin Acquired by Salesforce for $3.6B My notes below: 1. Everyone Misunderstands the SpaceX Valuation SpaceX’s $2 trillion-plus valuation does not make sense on the P&L alone, but that is the wrong lens. Elon’s playbook is selling long-dated call options to the market: full self-driving, Starship, orbital data centers, Mars. He has delivered enough of them that the market keeps buying in. 2. Elon’s AI Strategy Has Been the Best of Anyone While everyone else spent two years debating whether to “do something in AI,” Elon built two data centers, bought Cursor to fill the Colossus revenue gap, and signed $2 billion per month in compute contracts with Google and Anthropic. By September, xAI’s AI revenue run rate will exceed SpaceX’s entire launch business. 3. Why Elon’s Biggest Advantage Is Cost of Capital No bank would lend $24 billion with no contracts to build Colossus in under a year. Elon could do it because his cost of capital is uniquely low, earned through decades of delivering on impossible bets. That lets him make asymmetric wagers no one else can, reinforcing the track record that keeps his cost of capital low. 4. Why Anthropic Creates a Sovereignty Problem The US restricting Claude’s export based on capabilities is a Rubicon moment. If that logic extends to OpenAI and Google, governments may begin gating access to intelligence itself. Who gets access to ASI when it arrives: NATO allies, select citizens, approved institutions? 5. Any Liquidity for Pre-AI SaaS Companies Is Top-Decile Performance The Intercom/Fin $3.6B sale to Salesforce for is a masterclass in turning a zombie asset into real capital. Growing at 7% with no AI story, the equity was essentially worthless. If your board is not making this the top agenda item and asking, “Is this relevant for us?” that is a failure of governance. 6. We Have a Crisis of Open-Source Models in the Western World Outside China, there are essentially no world-class open-source models. The talent and focus required to be best-in-class across pre-training, mid-training, and post-training are scarcer than people think. For many countries, the realistic path may be fine-tuning a Chinese open-weights model. 7. How Does Benchmark Rationalize Not Being in Any of the Core Model Providers? They do not. Ev was blunt: “It f**king sucks.” Missing a 30x on scaled capital in four to five years is a failure, full stop, regardless of how well the rest of the fund performed. It is one of the most honest VC admissions you will ever hear. (links below)
@EverettRandle @rodriscoll Spotify 👉 open.spotify.com/episode/5K2te8… Youtube 👉youtu.be/kQn3GQBZ0Cs Apple Podcasts 👉 podcasts.apple.com/us/podcast/20v…
In-person podcasts are 10x better than remote. 10x might be underselling.
"The biggest thing I'd take from @elonmusk is his ability to focus relentlessly on the bottleneck. No matter how much is happening around him, he seems able to tune everything else out and concentrate entirely on the one problem that matters most. For founders, that ability to ignore the noise and obsess over the constraint holding the business back is an incredibly powerful skill." @AravSrinivas
"Agents won't kill advertising, they'll split the internet into objective and subjective decisions. For objective choices, like finding the cheapest flight, best insurance policy, or most suitable software tool, agents will increasingly make decisions on our behalf, reducing the role of ads. But for subjective choices, like fashion, furniture, travel experiences, or design, people will still want inspiration, discovery, and personal taste. Those categories will remain heavily driven by advertising and brand." @AravSrinivas What happens to advertising in an agent first world @scottbelsky @fdavidsont @alexschultz
"The future of software development isn't engineers building software, it's engineers building the factories that build software. Just as Tesla engineers design and optimize the factory rather than assembling every car themselves, software engineers will design the systems, guardrails, and workflows that generate code. Their job will shift from writing code to ensuring the software factory produces high-quality products without creating technical debt or bloat." @matanSF Do you agree and what does no one know that everyone should know about the future of software development @DanHollick @jsngr @steipete @sytses @Dimillian @thsottiaux
"The best organizations won't manage token budgets manually. Instead, they'll rely on orchestration layers that automatically route workloads to the right models based on performance, cost, and use case. Keeping track of which model is best for coding, finance, research, or support will become too complex for humans to manage directly, so intelligent routing will become a core part of enterprise AI infrastructure." @AravSrinivas How will the best orgs of the future manage token budgeting @mignano @nikesharora @typesfast @mmurph
"To stay competitive in AI, America needs to take physical infrastructure seriously. That means building more data centers, expanding power capacity, and making decisions based on facts rather than fear-driven narratives about AI and infrastructure. The bigger opportunity is not focusing on job destruction, but on enabling a new generation of companies that can create massive economic value with small teams and drive trillions of dollars of new GDP." @AravSrinivas What one single change would make the biggest difference to American competitiveness in the AI race @shaunmmaguire @mattocko @sethbannon @rabois @Sethwinterroth
"Export controls have probably helped in the short term by preserving a gap between frontier AI labs and open-source competitors. But the long-term effect is less clear. By restricting access to chips, the US may be forcing China to become far better at building its own infrastructure, data centers, and supply chains. The risk is that today's constraint creates a much stronger competitor tomorrow." @AravSrinivas Have export controls hurt or helped China @pmarca @vkhosla @KTmBoyle @DavidSacks @GavinSBaker
At this stage, I think it is undeniable that Revolut will be a $500BN company. (I do not say that lightly). They are a compound startup. They try 20 new products at a time. They create small 8-10-person teams for each product. They "invest" $2M into each team. They give them 24 months to execute. They monitor them on a weekly basis according to a set of agreed KPIs. What works, they double down on. A machine. 👇
"The best engineers in an AI world won't just ship features, they'll own outcomes. They'll understand customers, influence product adoption, and work across engineering, sales, marketing, and enablement to drive results. The engineers who thrive will be the ones who think like founders, not just builders." @matanSF Do you agree and does no one see that everyone should about what it will take to be a top engineer in 5 years @lennysan @sjwhitmore @jasonyuan @zan2434 @rsms
This show with @AravSrinivas is one of the best we have done. - The biggest problem today is power. - We will see large resistance to data centre buildout continue. - Micron will be worth more than Meta. - Export controls have helped China. - Monitoring token budgets is BS and only for losing companies. How is that for spice? I have added my notes from the discussion below but such a special show to do with Aravind in London. 1. Are Agents the End of the Internet Advertising Model? AI agents will disrupt industries built around objective transactions, but subjective spaces like travel, fashion, and shopping are more resilient. When intent is driven by vibes, exploration, and aesthetics rather than one correct answer, chat interfaces struggle to replace browsing, keeping much of the advertising model intact. 2. Biggest Takeaway From Elon Musk? An elite entrepreneur can identify the single limiting bottleneck in a business and focus on it with unusual intensity. True concentration requires discipline: you must ignore even important issues when they distract from the immediate objective that matters most. 3. Did Export Controls Hurt or Help US Competition With China? Export controls gave America a short-term edge by preserving a capability gap between open-source and frontier models. But they may have also forced China to become a stronger physical competitor by pushing its tech ecosystem toward memory-efficient architectures and deeper vertical integration. 4. How Will the Best Organizations Approach Token Budgeting? Enterprises will stop manually tracking model capabilities or micromanaging token budgets across teams. As model optimization accelerates, the best companies will outsource this complexity to AI orchestrators that automatically route tasks to the right models at the right cost. 5. How Can America Retain Competitiveness in AI? The West must build physical infrastructure aggressively, streamline power procurement, and counter alarmist narratives around data centers and job displacement. Leaders should explain the upside clearly: AI can let tiny teams build billion-dollar companies and drive major new GDP growth. 6. What Is the Single Most Important Metric in AI? The key metric is token value per watt per user. Raw model building and fine-tuning are becoming commoditized. The real venture-scale value will accrue to orchestration layers and agent harnesses that deliver high-quality outputs while using minimal data center power. 7. Why Is Moving Fast the Ultimate Expression of Humility? Founders often waste early cycles searching for a permanent moat, when speed is their only real defense. Moving fast is an act of humility because it forces constant contact with the market, repeated assumption testing, and a willingness to let reality overrule ego. (links below)
Spotify 👉 open.spotify.com/episode/2lcWvX… Youtube 👉 youtu.be/OxFyVcO1Yow Apple Podcasts 👉 podcasts.apple.com/us/podcast/20v…
Should we be worried about labour displacement at the hands of AI? "Short term, yes. Long term, no. In the near term, AI is creating real disruption, with large layoffs and thousands of people being displaced from existing jobs. Over the long run, there are far more problems to solve than people currently working on them. If AI makes engineers more abundant and productive, we can redirect that talent toward solving problems that were previously uneconomic or impossible to tackle." @matanSF Do you agree and what does no one consider that everyone should with regards to future of labour @pmarca @MattEvantic @shaunmmaguire @vkhosla
When I was a kid I hated Sunday night; another week of school loomed. Ever since I started 20VC, Sunday night has been my favourite. The start of a new week, more guests, more amazing founders. Do what you love and never work a day in your life.
"Consumers are better served when the model provider and the application provider are separate. If the same company owns both, it has an incentive to maximize token consumption rather than optimize for efficiency and customer value. An independent application layer creates competition between models, forcing them to be faster, cheaper, and better, while giving customers more choice and less vendor lock-in." @matanSF Why is this wrong and what would you say in response @bcherny @amorriscode @blelbach @trq212 @embirico @jxnlco
"Everyone gets FDEs wrong. The job of an FDE isn't to make the product work, it's to accelerate customer adoption and time-to-value. If you need FDEs just to deliver the product, you're not running a software company, you're running a services business with a bad product." @matanSF Do you agree and what do people misunderstand most about what it takes to do FDE motion well? @ssankar @chadwahl @nikogrupen @barrald @lkothari @LeoMehr @zkevinbai
"We spent $3,000 on an Eight Sleep for every employee because we optimize for output, not cost. When you're building a team of exceptional people, every improvement in sleep, focus, and decision-making is worth the investment. The hardest work isn't about grinding more hours, it's about making better decisions, and that starts with getting a great night's sleep." @matanSF What single decisions have had the biggest impact on performance for you @rabois @zebulgar @m_franceschetti @bryan_johnson
"The problem with grind culture is that it focuses on intermediate metrics instead of outcomes. Measuring success by hours worked is like deciding a basketball game based on who sweated the most instead of checking the scoreboard. Great companies hire great players and judge them on results. If you're forcing crazy hours and beds in the office, you're probably solving for the wrong thing." @matanSF Love to hear your thoughts @lmcorrigan1 @WillManidis @beffjezos @garrytan
"Dario's messaging around AI-driven job loss has done real damage. By framing AI as a force that could eliminate vast swathes of human work, it fuels fear, resistance, and calls to slow down progress. The criticism is that these narratives often serve fundraising goals in the short term, while overlooking the long-term reality that new industries, new jobs, and new opportunities tend to emerge from every major technological shift." @matanSF Do you agree? How should that message change moving forward? @JoshuaKushner @vkhosla @mmurph @Benioff @demishassabis
"So @ivankatrump is one of those rare investors who adds far more than capital. She combines an exceptional network with genuine generosity, helping founders with the kind of unglamorous, hands-on work that many investors won't do. The reason founders value her isn't just who she knows, it's how much time and effort she's willing to invest in helping companies succeed." @matanSF What does no one know that everyone should know about working with Ivanka @jaredkushner @davidsenra @francesca_lab @eladgil
Call me grumpy or jaded but I really only want to work with founders where what they are working on is their life mission.
Buy versus build: what are your lessons? "Just because you can build something doesn't mean you should. The best companies are ruthless about focusing on the few things they uniquely own and do better than anyone else. If it's not core to your business or competitive
To all of my insecure male friends in the valley: Yes, if you are not in SpaceX, Anthropic or OpenAI, yes, you are less of a man. KIDDING! Go touch grass, hug a loved one, change a nappy. Onto the next one.
If someone solve this problem for 20VC, I will wire them $500. 1. We create compositions in Descript. Clips of good moments. 2. This is done in an audio file by me. 3. My team then have to go through them and make the same clips in the video file of the show. This wastes
There is only one podcast you need to listen to every week to know everything you need to know in tech: AGENDA: - Largest IPO Ever: SpaceX - OpenAI Files to Go Public - Uber Cuts 23% of HR - Lovable Hits $500M ARR - Apple rebuilds Siri on Google Gemini at WWDC My notes below
Spotify 👉 open.spotify.com/episode/1aGM11… Youtube 👉 youtu.be/jxw7iRVpRoI Apple Podcasts 👉 podcasts.apple.com/us/podcast/20v…
Today is the day: SpaceX launches largest ever IPO in history. I sat down with @jasonlk and @rodriscoll to discuss everything: pricing, predictions, and much more on the biggest news this week: - OpenAI Files to Go Public - Uber Cuts 23% of HR - Lovable Hits $500M ARR - Apple
Spotify 👉 open.spotify.com/episode/1aGM11… Youtube 👉 youtu.be/jxw7iRVpRoI Apple Podcasts 👉 podcasts.apple.com/us/podcast/20v…
The Leo Aschenbrenner Effect "Everyone noticed it when @leopoldasch made such a large investment. We saw it as validation of what we’re building, but ultimately it’s just another opportunity to deliver. Every customer and investor gives you credit in advance, and your job is x.com/HarryStebbings…
Why Model Concentration is the Biggest Danger: "The biggest threat to Nebius is a world that becomes too consolidated. If 3 to 5 companies end up controlling everything, infrastructure providers become far less important and are reduced to serving a handful of dominant x.com/HarryStebbings…
How education changes in a world of AI "In a world where everyone has access to intelligence, memorising facts becomes far less important. The challenge is teaching people how to think, adapt, and continuously learn as industries and careers keep changing. Education will shift x.com/HarryStebbings…
