📖 Venture Chronicles - February 2026
Overlooked #214
Hi, it’s Alex from 20VC. I’m investing in seed & series A European vertical solutions (vSol) which are industry specific solutions aiming to become industry OS and combining dynamics from SaaS, marketplaces and fintechs. Overlooked is a weekly newsletter about venture capital and vSol. Today, I’m sharing the most insightful tech news of February.
I curated updates and insights around three themes:
Vertical Software
General Venture Capital
Entropy - other news and personal topics of interest
Vertical Software
Toast reported its 2025 financial results. - Toast
$2bn ARR growing 26% YoY. 164k locations growing 22% YoY. $51.4bn GPV growing 22% YoY.
“Toast signed an agreement with MTY Food Group to roll out Toast across more than 1,000 Papa Murphy’s US locations. MTY selected Toast for its modern, flexible platform that can support multiple operating models on one system, along with the flexible product features needed for a growing pizza chain. This expands Toast’s existing relationship with MTY. Toast currently serves their Wetzel’s Pretzels and BBQ Holdings locations.”
“Our product team released over 500 new features, including ToastIQ, our conversational AI assistant. Customer feedback and adoption has been tremendous. ToastIQ not only generates reports and insights about restaurant performance, it executes tasks directly in Toast ranging from menu management to inventory updates.” “Less than 4 months post launch, over half of all Toast locations have used ToastIQ, collectively sending over 8 million queries and tens and thousands of locations are already using it each week.”
“We continue to grow market share year after year and now power 20% of SMB and mid-market restaurants in the U.S. This has nearly doubled over the past 3 years.”
“2/3 of our demand is inbound and existing customers are the largest source of referrals.”
“In 2026, we will continue to invest in the platform to support the needs of our largest customers, including the launch of our drive-thru product, which is planned for later this year.”
“Our SaaS net retention rate remained in a healthy range at 109% in 2025, led by solid contributions from upsell and location expansion from existing customers.”
“Payback periods across our new TAMs are above the core today, given we are earlier in building our go-to-market and product offering. As we scale and mature in these areas, we are confident each one is on a path to under 20 months. We have a proven track record and clear road map to improve payback periods. In the core, payback dropped from 22 months in 2019 to 14 months in 2023.”
If “you you think about everything Toast offers: it’s software, it’s also hardware. It’s fintech, so things like lending and payments, payroll, with heavy regulatory and compliance needs.”
“Restaurants outsource work around generating demand with marketing or bookkeeping, payroll and tax. And we think there’s an opportunity there because a lot of that is our data that’s powering those experiences.”
9fin is raising $150m at a $1bn pre-money valuation. It generates $50m in ARR implying a 20x EV/ARR multiple. It previously raised a $50m series B at a $500m valuation in 2024. 9fin is a vertical SaaS in financial services for credit traders. - Bloomberg
“Founded in 2016, 9fin offers what it calls an “AI-native platform” oriented toward credit traders, investors and advisers, helping market practitioners to benchmark transactions and originate new business opportunities more efficiently.”
Lawhive raised a $60m series B led by Mitch Rales who is the cofounder of Danaher Corporation. It’s a an AI-native law firm that combines a network of human lawyers with a proprietary AI platform to automate routine legal work. It offers consumer legal services (e.g. family law, landlord-tenant disputes, property transactions, and consumer rights). It grew its annual revenue 7x to reach $35m with 400-500 lawyers supported by the platform. - Fortune
Gyde raised a $60m round led by Lightspeed with the participation from Optum, Crystal Venture Partners, Virtue and MVP Ventures. It’s an AI-native brokerage platform acquiring and augmenting brokers with AI tooling. - Gyde, Business Wire
“Through acquisition, Gyde is initially partnering with best-in-class health insurance agencies that serve the Medicare Advantage, Employee Benefits, and Individual markets, keeping existing teams in place post-acquisition and investing alongside them to grow the business.”
“The platform includes GydeOS, Gyde’s broker-facing platform, and Gia, its Intelligent Assistant. Together, they deliver a consistent, high-touch, and customer-oriented experience at unprecedented scale. By combining intelligent automation with a service-centric approach, Gyde creates a scalable way to deliver more proactive, personalized, and continuous support as consumers and employers navigate their healthcare and financial journeys.“
Harper raised $47m in a combined seed & series A. It’s a YC-backed company and Harper’s series A was led by Emergence. It’s an AI-native commercial insurance broker. - Techcrunch, Harper
“They thought about building AI tools for existing brokerages. Then they decided to just use that technology to build an AI-native insurance brokerage.”
“The YC blog wrote that the future of agencies “will look more like software companies, with software margins.” That’s exactly what Harper is — an almost fully autonomous licensed commercial insurance agency. It matches small- to mid-sized businesses with more than 160 insurance carriers to assist with workers’ compensation, as well as general and professional liability.”
“A typical sales team at a human-led brokerage handles 20 to 30 deals a month, but AI enables Harper to handle more than 1,000 customers a month. Harper has more than 5,000 customers to date.”
“What makes Harper different is that it hopes to target middle America. The real-world businesses like daycares, manufacturers, car dealerships, local bars and restaurants.”
“We raised this money because of what we’ve built. In 13 months, we’ve served more than 5,000 businesses across America, including manufacturers, healthcare providers, hospitality groups, transportation fleets, construction firms, and the bars and restaurants that anchor every Main Street in this country.”
“Harper uses AI to do what used to require an army of analysts: reading applications, routing submissions, following up with underwriters, managing quotes, answering questions at 2 am. Our proprietary models handle the judgment-heavy, unstructured work that slows traditional insurance down.”
“Harper isn’t selling software to legacy brokerages and hoping they figure it out. We are the brokerage. We own the customer relationship. We do the work. And because AI can execute judgment at scale, we can deliver the quality of service that used to be reserved for Fortune 500 companies to every manufacturer, every healthcare facility, every fleet operator in middle America.”
Jake Saper shared Emergence’s framework for AI-era software moats. - Jake Saper
“Pre-Claude, getting humans to do their jobs inside your software was a powerful moat. But if agents are doing the work, who cares about the human workflow?”
“MCPs are rapidly commoditizing the glue between systems. Being the connector used to be a moat. Soon, it’ll be a utility.”
“Network effects is the most underappreciated moat in AI, and potentially the most enduring. But not all network effects are equal. Workflow-based ones (more humans in the tool) are vulnerable. Data and trust-based ones (every user makes the product smarter) actually accelerate with agents. LinkedIn, Doximity, and Slack all still have this. Relatively few AI native companies have prioritized it...yet.”
“Seat-based pricing and workflow stickiness were the backbone of the last era. Both are eroding. The next era rewards companies tied to outcomes, context, and trust.”
General Venture Capital
Invest Like the Best recorded a podcast episode with **Dan Sundheim who is founder and CIO of D1 Capital**. - Colossus
“The only telltale sign was reading Jeff Bezos’ 1997 shareholder letter, which was like – the clarity of thought, his understanding of what he wanted to achieve and how to create value for shareholders was greater than almost any public CEO I dealt with. If I had read that and almost ignored everything else, it would’ve been a really important sign, and very profitable. Dario struck me like that. […] I place a lot of weight, rightly or wrongly, on clarity of thought and the ability to communicate as a CEO. What you want to achieve and how you’re going to achieve it.”
“I think the real debate [on LLM providers] is, these are extremely capital-intensive businesses, capital-intensive to a degree that we’ve never seen before in the history of business. And the question is, you’re spending a ton of capital and the ultimate return on that capital is unknown.”
LLMs are improving exponentially, making their real-economy impact hard to predict. The key is to identify which companies will retain durable moats protected from LLMs.
Software will have to evolve. It will probably be a worse business model going forward but some software companies will be successful at riding the AI wave like Walmart has been successful in riding the e-commerce wave - even if it was a painful process to get there with massive investments & an impact on margins. Software systems of records are more resilient but they need to integrate AI to keep thriving in the long term.
Gokul Rajaram recorded a podcast on Invest Like the Best. - Invest Like the Best
“Gokul is one of the most prolific product builders of the last 20 years. He’s built the core ads and product businesses at Google, Facebook, Square, and DoorDash, working at each company during its most formative scaling periods. Alongside his operating career, Gokul has invested in more than 700 companies, giving him an unusually broad view into how products are built and scaled.”
“What is interesting about product development is that 10 years ago or even five years ago, there were very clearly defined roles. Product managers articulated what to build, designers designed it, and engineers built it. Over the last few months, I’ve been talking to many companies. But over the last two months, in particular, December ‘25, and January ‘26, it’s become very clear that something has fundamentally changed. And what that thing is, is the notion of a long horizon, a long-running agent. […] Because these agents now are resilient to failure, and you don’t have to be very technical to use them, this changes the expectation of product teams.”
Clear role boundaries between PM, design, engineering are collapsing due to AI coding agents.
Designer and product manager roles are merging.
Product managers now define customer needs and “why,” then prototype and code alongside engineers.
Product managers increasingly commit AI-generated code to production repositories.
“The actual product is built bottoms-up by engineers, researchers, and product managers, and designers all working together on the code itself.”
“Until recently, software was used more as a tool by humans. We finally have software that is agentic in nature, which means it can do the job of people. What industry are there roles of people that are highly paid, that are doing somewhat of a repetitive job, and that can be done by software? Every three months, the answer gets deeper and deeper.”
“You want to target a high-value workflow. You want to target a workflow that is deep, that is complex, and that requires custom data. I think one of the challenges with this whole space is that the models are becoming so good that, if you try to build a company that is light, that is not a hard problem, the foundation model companies are going to eat you.”
“There are a few things around durability. One, you need to have ownership of a scarce asset. A scarce asset could be a license of some kind, it could be a regulation of some kind where you have unique insight into it. Second, you might basically own a control point. A control point is a thing that controls how people interact with money or with data. Third, you want to maybe have hardware, which is hard to replace. Fourth, maybe you want to be part of an essential workflow. Fifth, you want to have network effects.”
“For many years, incumbent systems of record all had APIs that if you enter that industry, you could build an agent company on top of these APIs. In 2024, things changed. These companies started seeing that these agent companies, AI companies that are being built, they are starting to take on the functionality out of these companies and are treating them like a dumb database. So you started seeing last year that these companies are cutting off access to APIs. […] They’re blocking access to APIs. They’re offering their own agents for free, bundled. I think that is a great and effective strategy. Or they’re charging these AI agent companies to access the data. Just to access data, the API was free. They’re saying now it’s $2 an API call. They’re trying to make the model of these agent companies unviable.” “So agent companies have no option but to also start building and offering a system of record.”
“The software companies that should be the most worried right now, is where they are pricing the product based on utility. Zendesk is a good example. Literally, Zendesk prices seats, and each seat comes with utility. In other words, each seat corresponds to a customer service agent that takes a certain number of customer tickets. So, that company should be worried because I can have an AI agent sit right next to Zendesk and you can slowly siphon off. Instead of paying for 50 Zendesk seats, you can pay for 20, and I can have 30 AI agents sitting next to Zendesk. For these companies, you need to change your pricing model to be based on outcome, and you need to actually build the product to be based on outcome. It’s easier said than done because literally you’re going from a $20 or $30 per seat to maybe charging a buck, or $0.50 or $0.20 per ticket resolved and you don’t know how that’s going to turn out.”
“The companies that are less exposed are ones where the utility is not based on seats, but it’s based on data that has been collected and captured over a period of time. The more timeless the data is, the more protected they are (e.g. ERP/CRM vs. Slack).”
“In the age of AI, stickiness, I think, comes from a few sources. One, you need to have network effects (e.g. DoorDash). […] The second example of stickiness is when you have financial or money moving through you. I think that’s another way to be sticky. Many of the systems of records, for example, Toast, have payments going through them. […] The third stickiness is from hardware. You can actually have hardware. Toast is a good example where Toast gives you hardware for free. The fourth one is access to a unique asset (e.g. Bret Taylor’s unique access to F500 companies).”
Jack Altman interviewed the entire Benchmark partnership. - Uncapped
Benchmark is built on several non-consensus pillars: (1) a small and equal partnership, (2) a focus on ultra-early-stage investing (pre-seed to series A) and (3) a refusal to raise massive funds that might compromise their ability to serve as “sparring partners” for founders.
Scaling of capital degradates the quality of the relationship with the entrepreneur and the cash on cash multiple for investors.
By not participating in future rounds, Benchmark eliminates conflicts of interest, aligning themselves fully with founders on dilution and outcome size.
Partners wants to be seen as proxy co-founders. (1) Being the first call: building enough trust so that when an entrepreneur hits a patch of bad news, the Benchmark partner is the first person they notify. (2) Continuous engagement: boards often average 10+ years of service which provides a stable relational balance through multiple executive teams. (3) Truth seeking: comfort and “slaps on the back” are less valuable than direct & constructive feedback that makes the entrepreneur better.
Odin shared data on emerging managers and arguments on why they are outperforming the venture market. - Odin
“The single best opportunity in venture capital today is investing in operators’ funds that are small in size, have differentiated access, and exercise independent judgement. I’d bet this cohort of managers will outperform 90%+ of traditional VC’s with billions in AUM.” Ankur Nagpal, Founder of Carry
Reasons for outperformance: (1) easier to deliver larger multiples on smaller amount of capital, (2) better compensation incentive (you need your carried to generate generational changing money), (3) unique edge over the market, (4) unique focus, (5) more contrarian investments with more attractive prices.
“For the same reason that VCs like to invest early, [LPs investing in emerging managers] are effectively buying a call option on future potential. If you skip the first fund, and the performance is impressive, you’ll have a much harder time getting access to the second fund.”
Alex Rampell shared his framework to assess entrepreneurs. - Alex Rampell
“Materialize Labor: “Who are your first 5 hires?” I always try to ask this question. The best answers are where the person literally has five people ready to jump off their existing ship(s) and join and those people are great...it shows that they are a magnetic attractor of talent. But even without that, do they know whom to hire? Fastest ramp is where the 5+ are already ready to jump and those 5 are themselves great.”
“Materialize Capital: another description for “heat” but with more of a future lens -- how good of a presenter are they? How do I handicap their future fundraising?”
“Is this a “fast boil” company or a “slow boil” company? In other words, is there a good chance (based on the entrepreneur, TBH team, and idea/product) that they can get to discernible traction early?”
“How “deep” are they in the domain? If their exposure is nascent, did they really take the time to learn the history, present, and have a theory for the future of the space? I know we call this the idea maze, but I consider it more a sign of intellectual rigor plus ability and willingness to challenge their beliefs and take input from all sorts of people.”
“Do they want to learn/win, or do they want to think they’re right? Similar to the above, I really look for people who want to spar/duel around the best way to accomplish a solution -- they are headstrong, which is great, but also want to learn of alternatives -- somewhat the manifestation of “strong ideas, weakly held”.”
“Marathon, not a sprint -- they will not give up. What in their past convinces me that, when they’re going through hell, that they’ll keep going? Have they faced adversity? What kind and how? Even if silver-spoon-fed, what motivates them and what have they done that requires incredible tenacity?”
“Hire fast, fire fast, decision fast — can they make FAST decisions? Are they decisive or indecisive?”
“Do they have some path to get their first five customers? Who wants to run their business on enterprise software written by a company with 9 months of cash?”
AI threatens private-equity backed software companies. - FT
“Niche software companies have been at the epicentre of a decade-long dealmaking boom — but the rise of AI technologies threatens to make them obsolete.”
“The release of the model, Claude Opus 4.6, caused many to question the business models of companies selling specialised software services to large and midsized corporations. Could Claude be used to recreate valuable software businesses serving call centres, salespeople, human resources departments and a whole host of data analytics companies that had only recently been sold for billions, or even tens of billions of dollars?”
“Takeovers of software companies by private equity funds accounted for about 40% cent of trillions of dollars in deal activity over the past decade by some estimates.”
“To make matters worse, the apex of the software deal boom came at the end of a decade of ultra-low interest rates during which takeover valuations roughly doubled from 2012 to 2022. But a combination of US interest rate rises precipitating a collapse in some tech valuations and the more recent threat of AI now raises serious questions about whether big investments in private software held by pension funds, endowments and retirees are at risk.”
“Initially, lenders had been reluctant to finance software buyouts given that such companies have few physical assets and generally do not earn high profits by traditional accounting measures. But several novel private credit firms such as Ares, Golub and Blue Owl took notice and invented specialised loans for software buyouts.”
“Apollo went a step further. Last year, it started cutting its exposure to software deals and even began shorting the debt of some companies it deemed exposed to AI risks.”
“Orlando Bravo has remained optimistic about the sector’s future, arguing that AI will be a beneficiary of many niche companies that can leverage the technology to cut internal costs and build new products. He also said that most of the value in niche software companies comes from their understanding of specific industries and how to win sales, or implement the product inside IT departments.”
“Bravo, who is sitting on tens of billions of dollars in unspent investor cash, told the FT he was planning a deal spree to take advantage of plunging valuations.”
Anthropic raised a $30bn series G led by D.E. Shaw, Dragoneer, Founders Fund, Iconiq and MGX at a $380bn post money valuation. - Anthropic
“It has been less than three years since Anthropic earned its first dollar in revenue. Today, our run-rate revenue is $14 billion, with this figure growing over 10x annually in each of those past three years.”
“The number of customers spending over $100,000 annually on Claude (as represented by run-rate revenue) has grown 7x in the past year.”
“Two years ago, a dozen customers spent over $1 million with us on an annualized basis. Today that number exceeds 500. Eight of the Fortune 10 are now Claude customers.”
“Claude Code was made available to the general public in May 2025. Today, Claude Code’s run-rate revenue has grown to over $2.5 billion; this figure has more than doubled since the beginning of 2026. The number of weekly active Claude Code users has also doubled since January 1.”
“Claude remains the only frontier AI model available to customers on all three of the world’s largest cloud platforms: Amazon Web Services (Bedrock), Google Cloud (Vertex AI), and Microsoft Azure (Foundry). We train and run Claude on a diversified range of AI hardware—AWS Trainium, Google TPUs, and NVIDIA GPUs—which means we can match workloads to the chips best suited for them.”
Databricks raised a $5bn round at a $134bn valuation. In 2025, it reached $5.4bn in ARR growing 65% YoY while being cash-flow positive. $1.4bn in ARR (26%) is coming from AI products. Databricks has a 140% Net Dollar Retention, 800 customers with $1m+ ACV and 70 customers with $10m+ ACV. It will use the funding for product development on Lakebase (Postgres database to build data & AI applications) and Genie (conversational AI assistant). - Databricks
ElevenLabs raised a $500m series D at a $11bn valuation led by Sequoia with the participation of both a16z and Iconiq. - WSJ, ElevenLabs
It ended 2025 with $330m in ARR. It reached $100m in 20 months, $200m in 10 months and $300m in 5 months.
“The startup aims to continue growing into a broader platform for conversational AI, including selling tools to business customers who are using AI-generated voices for customer service, sales and marketing.”
“Its customers include Deutsche Telekom, which uses ElevenLabs voice agents for customer service, and delivery service Deliveroo, which uses the technology to streamline courier onboarding. The government of Ukraine is also using the software for several initiatives, including in education.”
“With the new funding, ElevenLabs is doubling down on ElevenAgents, its enterprise platform for voice and conversational AI, to support customer experience, sales and marketing, and internal workflows with interactive voice agents.”
Mistral reached $400m in ARR growing 12x YoY and is planning to surpass $1bn in ARR by the end of the year. It has 100+ enterprise customers. Mistral is also investing €1.2bn to build an AI data-center in Sweden. - FT
“We are diversifying and spreading our capacity across Europe,” Mensch said. “Europe has realised that its dependency on US digital services was excessive and at breaking point today. We bring them leverage because we bring them models, software and compute that is fully independent from US players.”
“Concern has intensified in European boardrooms and capitals that US President Donald Trump’s foreign policy could force a “tech decoupling”. The EU today relies on overseas providers, most of them American, for more than 80 per cent of its digital services and infrastructure.”
“It has expanded efforts to build and run its own AI data centres, instead of relying on US “hyperscalers” such as Amazon, Microsoft and Google to take its products to market. This “vertical integration” helps pay for chips that can train its next generation of models — running customer workloads by day and training new AI systems by night — as well as offering European customers comfort that their data is stored on local servers. Providing sovereign AI infrastructure was “top of mind for everyone”, Mensch said.”
“Mistral’s customers include ASML, TotalEnergies, HSBC and several European governments including France, Germany, Luxembourg, Greece and Estonia. About 60% of revenues come from Europe, with the rest from the US and Asia.”
Entropy
Stripe published its 2025 annual letter. - Stripe
“Last year, businesses running on Stripe generated $1.9 trillion in total volume, up 34% from 2024, and equivalent to roughly 1.6% of global GDP.”
“Metronome joins our Revenue suite, which is on track to hit an annual run rate of $1 billion this year.”
“The sorting machine is now whirring faster: winners and losers are being anointed more quickly and more intensely. Today, the most profitable third of publicly listed companies in the US account for two-thirds of total market capitalization, the highest share since data began in 1963. And much of this is a story of profit concentration, not just valuations: the top 10% of the S&P 500 by market cap now account for roughly 59% of the index’s total profits, which is elevated relative to recent history.” “Each sector has its own particular dynamics, but the pattern is clear: a cohort of companies is pulling away.”
“[The cohort of companies created in 2025 on Stripe] is by far the highest performing and fastest moving we’ve ever seen, growing around 50% faster than the 2024 cohort. The number of companies reaching $10 million ARR within 3 months of launch was double the 2024 count.”
“After years of relative calm, the number of iOS apps released in December 2025 jumped by 60% year over year. Even code production is accelerating: pushes to GitHub, which grew roughly 10%–12% in prior years, surged 41% between Q3 2024 and Q3 2025.”
“In 2025, the price of Bitcoin dropped precipitously (and is now down 50% from October), but stablecoin payments volume doubled to around $400 billion, 60% of which is estimated to represent B2B payments. Bridge, the stablecoin orchestration platform we acquired, saw volume more than quadruple.”
Stripe’s 5 levels of agentic commerce.
Supercell published its 2025 annual letter. - Supercell
Supercell generated $3bn in revenues and $1.1bn in EBITDA.
“It was a near-record year that virtually matched our 2024 results. The year was driven mostly by an incredible, historic year of Clash Royale. On the other side, launching new hit games remains as hard as ever – not just for us, but across the entire industry, especially in the West.”
“Are we satisfied with the status quo – relying on optimization and incremental improvements to the great games we’ve already made? Or are we brave enough to dream bigger and be more ambitious about what this industry could become?”
“Royale’s re-engaged players doubled. New players grew almost 500%. Every major measure of playtime and engagement grew significantly. And maybe most importantly, Clash Royale once again became the game to play with friends around the world. It felt like 2016 again.”
“Billy Ren, one of our market insights specialists, recently analyzed mobile games launched since 2020. Out of roughly 53,000 games launched, only 22 (about .04%!) grossed $1 billion or more – which is roughly what meeting our ambitions looks like. Here’s what’s striking: 20 of those 22 came from developers in China, Japan, and South Korea. Only 2 were developed in the West: Royal Match by Dream Games and Monopoly Go by Scopely.”
“Why is the West falling behind? Yes, local Chinese developers have a natural advantage in the world’s biggest market. But we in the West should look in the mirror. The fact is, we have not brought radical new gameplay innovation to the market – not in the way Clash Royale, Pokémon GO, or Brawl Stars each did in their time.”
“Here’s the reality: live game excellence alone doesn’t grow an industry. It maintains one. For the market to truly expand, we need to bring new players in. People who don’t currently play mobile games.”
“In 2025, we literally doubled our investment in new games and innovation compared to the prior year and this year we’re expecting to roughly double it again. While many companies have scaled back, we’ve moved in the opposite direction.”
Uber bought Getir from Mudabala for up to $1bn to double down on its expansion in Turkey. Getir was founded in Turkey in 2015 to deliver groceries in less than 10 minutes. It became one of the largest quick-commerce players alongside GoPuff and Gorillas. Its valuation peaked at $11.8bn before Mudabala took control of the entire business to avoid a bankruptcy. - Bloomberg
The so-called “global intelligence crisis” fears are overblown. AI adoption is slower, economically constrained, and more likely to boost productivity than trigger mass job destruction. - Citadel
“Displacing white collar work would require orders of magnitude more compute intensity than the current level utilization. If automation expands rapidly, demand for compute definitionally rises, pushing up its marginal cost. If the marginal cost of compute rises above the marginal cost of human labor for certain tasks, substitution will not occur, creating a natural economic boundary. This dynamic contrasts sharply with narratives assuming frictionless replication of intelligence. Even if algorithms improve recursively, economic deployment remains bounded by physical capital, energy availability, regulatory approvals, and organizational change. Recursive capability does not imply recursive adoption.”
“At its core, AI-driven automation is a productivity shock. Productivity shocks are positive supply shocks: they lower marginal costs, expand potential output, and increase real income. They are in isolation disinflationary and growth-enhancing in the medium term.”
“There is little evidence of AI disruption in labor market data as of today. In fact, the forward-looking components of our labor market tracking have improved and AI data center construction appears to be driving a pick-up in construction hiring.”
“The economy contains a vast array of tasks: physical, relational, regulatory, supervisory – that are costly to automate. Even cognitive automation faces coordination frictions, liability constraints, and trust barriers. It seems more likely that AI will be a complement rather than a substitute for labor is many areas.”
Emily Kramer (ex. head of marketing at Asana and Carta) wrote a post on running executive dinners as a GTM strategy to close customers. - MKT1
“Many of [companies doing events] noticed was that the most valuable moments were the side conversations: the dinner the night before, the coffee meeting in the lobby, the drinks after a long day on the floor. At some point, people started asking: Why not just do those things directly? You control the budget, the venue, and the guest list. You can get the right mix of existing champions, active prospects, and warm accounts. You’re not trying to collect 500 badge scans from people who will never respond to follow-up emails. You’re connecting directly with ~12 people who are actually in your pipeline, actually evaluating you, and giving them a reason to move forward. Seems like a better use of money.”
“The other reason these dinners are a go-to strategy is not just the reaction to the trade show reboot post-Covid, it’s of course, AI-related too. In the age of AI-generated everything, this kind of authentic human connection is a strong differentiator. A dinner forces real conversation. You can’t automate your way through it. That’s why it works.”
“People who go to these dinners tell me they want to either meet their peers, somebody interesting, or they want to learn something new that they can take back to their business. That’s ultimately what they care about. And ideally, it’s both. You’re helping facilitate those connections, and you’re helping facilitate the exchange of ideas.”
“Once you have your target accounts identified, treat each dinner as a campaign, not a one-off event. This is the same approach you need for any marketing initiative: you need fuel (the content and experience) and engine (how you get people there and follow up). An event campaign needs touchpoints before, during, and after.”
“Rather than going out with an empty invite list, seed a few people to build interest in the dinner. Are there friendlies you can bring in—like an existing customer, investor, or influencer—to get a couple of logos? Now you can use that in the promotion of the event itself, and that helps to drive RSVPs faster than if you had no one.”
Colossus published a deep-dive on 3G Capital. - Colossus
“He asked them how big Burger King was relative to McDonald’s. They each said half the size. The actual number was one-sixtieth. The brand was bigger than the business. If they could fix some obvious problems—like not selling cheeseburgers at a loss—they’d make money. Their model projected they would triple their money in five years. […] Fifteen years later, the investment is up nearly 30 times. Burger King is now one piece of a $45 billion public business Behring and Schwartz have built: Restaurant Brands International (RBI).”
“The firm became synonymous with zero-based budgeting, where every expense must “be justified from scratch. To 3G, it was a principle of ownership: Spend every dollar as if it were your own.”
“The cost discipline is real. They practice it themselves. 3G’s Midtown East office, which I visited recently for rare on-the-record interviews with Behring and Schwartz, is notable mostly for what it lacks.”
“Frugality is one piece of a model centered around the simple idea that employees should feel and act like owners. The model has produced the biggest investment bank in Brazil, the world’s largest brewer, the third-largest restaurant company, and turned hundreds of employees into multimillionaires. In 3G Capital, it has also produced a rare kind of investing partnership, one where each fund holds exactly one company, the partners are the largest investors in every fund, they work the businesses themselves, and they have never lost money on a deal.”
“It’s about a group of investors who circle the globe looking for old people worth learning from and young people worth betting on, who teach the latter to want impossible things, then give them years to make good on the wanting.”
“He became CEO in July 1998, and spent his first 10 days conducting 15-minute interviews with the top 150 managers. From these conversations, he identified 30 mid-level managers to help lead the transformation. He tore down walls, auctioned off company cars, let go of chauffeurs, and started recruiting students. Then he designed a compensation system in which top management would decide on five key objectives each year, and every manager would have the same number of personal targets linked to those goals. Individual performance was tracked weekly. The results were posted publicly, Behring’s included.”
Dwarkesh Patel wrote about the challenges with running data-centers in space. - Dwarkesh Patel
“The whole reason to go to space is energy. Yes, panels in space get about 40% more irradiance—but the real advantage is that you can put your satellites in sun-synchronous orbit, where they face the sun continuously. No nights, no clouds, no need for batteries (which is the majority of cost in a solar-storage system). Solar on Earth has a roughly 25% capacity factor, meaning panels only generate a quarter of their peak output on average. In space, you get close to 100%. The logic is that if the launch costs continue to drop, it will become cheaper to put GPUs in orbit than to build power plants and batteries on Earth.”
“But here’s the problem with this argument. Energy is only about 15% of a datacenter’s total cost of ownership. The chips themselves are around 70%.” You also need to replace chips when they die which will be extremely expensive to do in space.
Elon Musk believes that it will simply not be physically possible to scale power production to the scale needed for AI on Earth.
“I’m willing to accept Elon’s argument that if launch costs become sufficiently cheap and we can repair GPUs in space, then there’s a viable path toward orbital data centers. But it seems especially difficult to imagine a situation in which orbital data centers end up significantly cheaper, because, again, most of the cost of a data center is the GPUs. For most compute to shift to space, all of the following things would need to be true: (1) Power generation on Earth hits a ceiling, or AI demand outstrips every terrestrial option. (2) Chip production scales faster than anyone expects, so we have the silicon but not the electricity. (3) Starship reaches thousands of launches per year.”
The Economist wrote a portrait of Daniel Kretinsky. He is a Czech businessman who made a fortune in energy and who is building a diversified portfolio of European assets in retail, media, logistics and sports. - The Economist
“Last February the Czech billionaire took Metro, a big German wholesaler, private. In April he bought International Distributions Services (IDS), the parent company of Britain’s Royal Mail. In November he became a major shareholder in TotalEnergies, a French oil giant. And on January 26th he launched a takeover offer for Fnac Darty, a French electronics retailer.”
“Over the past few years he has quietly become one of the continent’s most influential businessmen, amassing a fortune of around $10bn in the process.”
“Although he still believes in Europe’s potential, he fears a lack of pragmatism will drag it down.”
“The heart of Mr Kretinsky’s business empire is EPH, a privately held energy company that is one of Europe’s biggest—and dirtiest. The company focuses on generating electricity mainly through coal- and gas-fired power plants and the storage and transmission of natural gas in various countries across the continent.”
“Mr Kretinsy’s speciality became scooping up dirty assets at bargain prices that listed utilities were forced to sell under pressure from investors and governments.”
“Mr Kretinsky has also acquired various media assets: he controls a swathe of the Czech media market as well as Editis, a French publishing house, and magazines including Elle and Marianne.”
“He believes Europe still has many advantages, including the scale of its market, the quality of its institutions and the standard of its education. But he worries that taxation and red tape are deterring capital and talent.”
The Financial Times interviewed Kalshi’s cofounder and CEO Tarek Mansour. - FT
“Tarek Mansour believes deeply in prediction markets — or so he tells me again and again. He says they will, variously, play the role of modern oracle, innovative asset, public educator, new journalism and political saviour. At the moment, however, most of the money flowing through them is gambling on American football.”
“The idea behind prediction markets such as Kalshi is simple and powerful. Participants trade shares of some future event (for example, will it rain in New York today?). The shares are redeemable for $1 if the event occurs and $0 if it does not. Beforehand, the price (74 cents, say) can therefore be read as a probability of the event occurring (74 per cent). Thus in theory, if large and liquid, these markets quantify and broadcast the “wisdom of the crowd”.”
“You have a financial market to price companies, you have one for physical commodities, for intangible commodities like rates and credit and currencies,” he says. “What if you built one for pricing simple questions about the future?”
“In addition to politics and sports, Kalshi hosts a selection of what might be called cultural markets — in other interviews, Mansour has been quick to mention Taylor Swift — including at the moment on the Oscars, existence of aliens, Epstein files, and whims of the unknowable mind of Donald Trump.”
Thanks to Julia for the feedback! 🦒 Thanks for reading! See you next week for another issue! 👋
This newsletter is so insightful 🙏