Another great recent interview with $NU's CEO. Nubank started because the traditional banking experience was terrible. David Vélez faced huge delays, revolving doors, and extreme frustration just to open a simple bank account in Brazil. He realized that major banks were failing the consumer. To solve this, Nu focused on fighting complexity and empowering people. Today, this mission has turned Nu into a massive financial giant. They now have 135M customers across Brazil, Mexico, and Colombia. In the first quarter of 2026 alone, Nu reached a record $5B in revenue and $871M in net income. From the beginning, tech gave Nu a massive edge. In 2012, the rise of smartphones and cloud computing allowed Nu to build a bank without spending billions on physical branches and old computer systems. This digital-first model makes Nu 20 to 30 times more efficient than traditional banks. This incredible efficiency resulted in a record-low operating efficiency ratio. Now, Nu is pushing this advantage even further through AI. They are completely rebuilding the bank around it. AI now handles real-time credit pricing in under one second and has sped up engineering testing cycles by 90%. Because Nu operates with such low costs and deep data insights, they can safely lend money in ways that others cannot. They are aggressively growing their credit card and unsecured lending businesses. These products made up 98% of their new exposure in the quarter. Nu does not worry about minimizing late payments. Instead, they use AI to price risk extremely well. This ensures their loans remain highly profitable over the long term. They keep loan durations short so they can react to market changes much faster than incumbent banks. They are also using these advantages to capture the massive small business market at nearly zero acquisition cost. They are quickly growing credit limits to win over high-income customers. Nu does not view itself purely as a bank. Their true goal is to make customers love them. Because 100% of their customers are digital, Nu can easily sell non-banking products. They are now expanding into marketplaces, travel, and telecom services like NuCel to make daily life simpler. Nu is also taking its winning formula to new countries. Their operations in Mexico and Colombia are growing very fast. They believe their Latin American culture helps them build strong emotional connections with customers. This gives them a major edge over traditional banks. They are also testing an expansion into the US. This move is a low-cost bet that risks less than 100 basis points of their efficiency ratio but offers massive upside. Despite their massive success, Nu is still at the very beginning of its journey. The global financial market produces around $7T in profits, and traditional banks still control 97% of it. By combining a pro-consumer culture with top-tier AI and an unbeatable cost structure, Nu plans to keep taking market share from legacy banks. They aim to eventually serve hundreds of millions of people across many countries. It is pretty wild to see that despite their growth over the last decade that a large part of their future growth will involve winning over even more customers from legacy banks.

Just got back from a Life Science investor conference. The mood of sub-$300m VC funds is not good. They are having a very hard time raising money. Many people got laid off just in the past 6 months. Apparently it is the same in EU and US. But Mega funds have no problem raising.

Today we are launching two revolutionary products: Dual and Phase. These devices will enhance how humans dream. Prophetic Dual retails for $449 and starts shipping at the end of this year. Prophetic Phase retails for $1299 and starting shipping middle of next year.

sorry but this is seriously fucking impressive china just shipped a claude code-level ai model small enough to run on your laptop. it codes better than opus 4.5 and its tiny. 27B beating models 15X LARGER than it. best part? shit is fully open source. no cloud, no rate limits, no api keys this model plus kimi k2.6 tells me open source has caught up to the frontier models how the fuck did china pull this off?

What's especially cool is that the flagellar motor exploits the indivisibility of 5 by 2 to run on a fuel of rising entropy as protons diffuse into cells. For me as a physics and math person, this is biology at its best. quantamagazine.org/what-phys…

Running a company: 2020: can you survive a pandemic? 2021: still here? we’re going to give all of your competitors $100m series A rounds. 2022: wow, you made it? okay, all engineers cost $600,000/year now. 2023: nice job! okay, SVB failed and we’re going to take away your bank account. 2024: a survivor I see. but can you pivot from ai to crypto to defense tech back to ai-enabled defense tech in a 12 month period to stay relevant? 2025: unfortunately all of your competitors have raised $2b series B rounds. oh and only 500 engineers are relevant and they cost $100m/yr each. 2026: well, well, well. you’re still in business? let’s deploy the thunderclap of godlike LLMs from the heavens so all of your customers can rebuild your app in 2 hours. can you survive?

Introducing Project Glasswing: an urgent initiative to help secure the world’s most critical software. It’s powered by our newest frontier model, Claude Mythos Preview, which can find software vulnerabilities better than all but the most skilled humans. anthropic.com/glasswing

Sequence is all you need. Protein language models set a new state of the art for predicting the clinical effect of genetic variations. x.com/vntranos/status/203901…

We spent $15,000 on billboards targeting one person: the guy controlling all the chemical spend at a saltwater disposal company in Texas. We mapped his commute and bought every billboard between his house and the oil field. When we finally called, he said "I see your billboards everywhere." That landed us our first oil field contract. At the time our entire operation was a $10,000 reactor built from PVC pipes from Home Depot, turning corn sugar into industrial chemicals. People keep trying to throw it away. It still works. That leaking reactor started a multibillion-dollar company. @ycombinator visited our plant in Houston. The original PVC reactor is still on the floor next to the Bioforge.

Introducing Humanoid Atlas, the Bloomberg Terminal for humanoids. Every OEM, every supplier, every dependency humanoids.fyi

My take on the whole "AI cures cancer in dog in Australia". It's a very interesting story, but perhaps not for the reasons that are being noted. In 2007, Freeman Dyson published an essay in The New York Review of Books called “Our Biotech Future.” It contains one of the most memorable predictions about the future of biology I’ve ever read. “I predict that the domestication of biotechnology will dominate our lives during the next fifty years at least as much as the domestication of computers has dominated our lives during the previous fifty years.” Dyson believed biology would eventually follow the trajectory of computing. At first, powerful tools live inside large institutions - universities, government labs, major companies. Over time those tools get cheaper, easier to use, and more widely distributed. Eventually individuals start doing things that once required entire organizations. “Biotechnology will become small and domesticated rather than big and centralized.” He even imagined genome design becoming something almost artistic: “Designing genomes will be a personal thing, a new art form as creative as painting or sculpture.” Dyson's words rang in my mind as I read the "AI cures dog cancer" story. Much of the coverage framed this as an example of AI discovering new science. But that’s not really the interesting part of the story. The scientific pipeline involved here is actually well known. It closely mirrors the workflow used in personalized neoantigen vaccine research that has been under active development for years. The steps are fairly standard: sequence the tumor, identify somatic mutations, predict which mutated peptides might be recognized by the immune system, encode those sequences in an mRNA construct, and deliver them to stimulate an immune response. The biological targets themselves were almost certainly not new discoveries (I have been unable to find out what they are, but mutations in targets like KIT which are common might be involved). Partly therein lies the rub, since the hardest part of drug discovery, whether in humans or dogs, is target validation, the lack of which leads to lack of efficacy - the #1 reason for drug failure. In neoantigen vaccines, the proteins involved are usually ordinary cellular proteins that happen to contain tumor-specific mutations. AlphaFold which was used to map the mutations on to specific protein structures is now a standard part of drug discovery pipelines. The challenge is identifying which mutated peptides might plausibly trigger immunity. What is interesting though is how the pipeline was assembled. Normally, this type of workflow spans multiple domains - genomics, bioinformatics, immunology, and translational medicine - and in institutional settings those pieces are distributed across specialized teams, document sources and legal and technical barriers. Navigating the literature, selecting computational tools, interpreting sequencing results, and designing a candidate mRNA construct is typically a collaborative process. In this case, AI appears to have helped compress that process, pulling together data and tools from different sources. Instead of requiring multiple experts, a motivated individual was able to assemble the workflow with AI acting as a kind of guide through the technical landscape. I’ve seen something similar in my own work while building lead-optimization pipelines in drug discovery. The underlying science hasn’t changed, but the friction involved in assembling the workflow can drop dramatically. Tasks that once required stitching together multiple tools, papers, and areas of expertise can now often be executed much faster with AI helping navigate the terrain; and by faster I mean roughly 100x. That kind of workflow compression is powerful, to say the least. When the cost of navigating technical knowledge drops, more people can realistically assemble sophisticated research pipelines. This story is a great example of what naively seems like a boring quantitative acceleration of the research process. In that sense, therefore, the real novelty here is not the biology but the combination of three things: a non-specialist orchestrating a complex biomedical pipeline, AI acting as a navigational layer across multiple technical domains, and the resulting decentralization of capabilities that were once confined to institutional research environments. But I think the story also points to something deeper, which is a challenge to modern regulatory environments. Modern biomedical innovation does not operate solely according to what is scientifically possible. It is structured by regulatory frameworks - clinical trials, safety oversight, institutional review boards, and regulatory agencies. Those systems exist for important reasons, but they also assume that the development of therapies occurs primarily within large, regulated organizations. When individuals begin assembling pieces of these pipelines outside those institutions, the relationship between technological capability and regulatory oversight starts to shift. The dog in this story sits outside the human regulatory framework. That fact alone made the experiment possible. In other words, the story is not just about technological capability; it is also about how certain forms of experimentation can occur when they bypass the regulatory pathways that normally govern biomedical innovation. One is reminded of another Australian, Barry Marshall, who received a Nobel for demonstrating through self-experimentation that ulcers are caused by bacteria. This raises an interesting question: what happens when the tools for assembling sophisticated biological workflows become widely accessible while the regulatory structures governing them remain institution-centric? That tension may ultimately be the most important implication of this moment. Regulatory frameworks will need to adapt to this kind of citizen science. Seen in this light, the story about the AI-assisted vaccine is less about a breakthrough in cancer therapy and more about a glimpse of the early stages of something Dyson anticipated nearly two decades ago: the domestication of biotechnology. If AI continues to reduce the cognitive overhead required to navigate biological knowledge and assemble complex pipelines, the boundary between professional research and motivated individuals may begin to blur. That shift will require careful thinking about safety, governance, and responsibility. But it also carries an exciting possibility. Dyson imagined a world in which biological design might eventually become something like a creative craft practiced not only by institutions but also by curious individuals experimenting at smaller scales. For a long time that vision felt distant. Now, it feels like we may be seeing the first hints of it.

Over the course of history--scribbling, scratching, typing and clattering--we, humanity, have managed to jot down, memorialize and commit to paper (or disk) 270 quadrillion written words. In 2026, AI will surpass that output. The singularity is here.

Thrive X raised $10 billion today and the announcement reads like a monastery brochure. “We do not view this as a milestone.” “We are not the main character.” “We do not hedge.” Meanwhile, the math behind Thrive tells the story of the most aggressive concentration strategy in venture capital history. Josh Kushner started with $5 million from Joel Cutler at General Catalyst in 2010. He was 24. His AUM was $2 billion in 2020. Today it’s north of $25 billion. The fund progression tells the story of compound conviction: Fund I ($5M, 2010), Fund II ($40M, 2011), Fund III ($150M, 2012), Fund IV ($400M, 2014), Fund V ($700M, 2016), Fund VI ($1B, 2018), Fund VII ($2B, 2021), Fund VIII ($3.3B, 2022), Fund IX ($5B, 2024), Fund X ($10B, 2026). Every single fund roughly doubled. For 16 years straight. Princeton’s endowment calls Thrive one of their top 10 best-performing relationships. UTIMCO data shows a 34.5% net IRR and 2.8x TVPI on the growth fund. Now look at the bets. In April 2012, Kushner invested in Instagram’s $50 million Series B at a $500 million valuation alongside Benchmark, Sequoia, and Greylock. Facebook announced the $1 billion acquisition 72 hours later. Thrive doubled its money before most VCs finished their diligence calls. The offered allocation was roughly 30% of his $40 million fund. At 26, he was willing to put a third of his entire vehicle into a photo app with zero revenue. That willingness to concentrate became the playbook. In 2022, when OpenAI needed money, Sam Altman didn’t call Andreessen. Didn’t call Thiel. He called Kushner. Thrive put in $130 million at a $29 billion valuation. It was the only term sheet OpenAI received. OpenAI’s current valuation target is $830 billion. That $130 million position has appreciated roughly 28x in three years. Then Thrive did it again. Led the $86 billion tender offer. Then led the $6.6 billion round at $157 billion, committing $1 billion with an option for $4 billion more at the same price. Total OpenAI exposure: approximately $1.3 billion across three rounds. If OpenAI hits its IPO target, that position alone could be worth north of $15 billion. The Stripe bet is equally aggressive. Thrive committed $1.8 billion at a $50 billion valuation in early 2023, more than half of Fund VIII’s $3.3 billion pool. Traditional portfolio theory says cap single positions at 10 to 15% of fund size. Kushner put over 50% into one company. Stripe’s valuation hit $107 billion by late 2024. In 18 months, that single investment roughly doubled. GitHub: nearly 10% stake, $150 million invested, sold to Microsoft at $7.5 billion. Robinhood, Affirm, Nubank, Databricks ($43 billion), Anduril, Cursor, SpaceX. A public market Carvana trade that returned $522 million over three years. Thirty-nine unicorns. Twelve IPOs. Fifty-two acquisitions. From a firm that makes a handful of deals per year. Most VCs spray capital across 30 to 50 companies per fund, hoping for 2 or 3 outliers. Thrive runs a small team managing $25 billion. When they commit, they go so deep that any individual position can make or break the fund. A 24-year-old with $5 million turned concentrated bets on Instagram, OpenAI, Stripe, and GitHub into one of the highest-performing track records in the history of the asset class. And the gap between Thrive’s approach and conventional venture is only getting wider.