Marc Andreessen 🇺🇸

@elonmusk @KatieMiller

Crypto is at a pivotal moment. As Congress debates landmark market structure legislation, Marc Andreessen and Chris Dixon discuss the decisions that could shape where the next generation of financial infrastructure is built, and whether the United States remains the global leader in crypto innovation: 00:00 Intro 05:31 From crypto subculture to financial infrastructure 08:37 Why crypto needs rules now 12:20 The regulatory war on crypto 15:41 How CLARITY could prevent another FTX 22:42 Why criminals using crypto may be easier to catch 26:20 Privacy, blockchains, and the invention of HTTPS 30:41 Government ethics and crypto 34:52 The banking lobby’s stablecoin fight 37:05 Why every major bank is building on blockchains 41:04 Developer liability as a killshot 45:23 How CLARITY provides oversight 49:30 What happens if CLARITY fails? 50:50 Regulation vs. innovation 54:12 Why America should lead 55:36 What CLARITY could unlock @pmarca @cdixon @rhackett

BIG Update: I’m joining @a16z as an Investing Partner on the a16z @speedrun team! 🎉 I’ll be focusing on early-stage investing as part of speedrun, where we invest up to $1M in exceptional founders building great companies, & give them unfair advantages to succeed. If you are or know a founder who wants to win, I'd love to meet you! I'm especially excited about compelling novel ideas in AI dev tools, agent RL & evals, data, & infra in general. For those who have known me over the years, you know how much joy I get in supporting those around me. From working countless nights to support my brilliant teammates at @scale_AI, to spending thousands of hours in my leisure time making free videos for millions of students in Bangladesh, I have given my all to support those doing their life's most important work. That has always been my personal mission. So, when I met the speedrun team, a team that works relentlessly to make early bets & support founders from the very beginning of their journey, I knew this was the dream team for me. I am thrilled to join this shared mission with @andrewchen, @Tocelot, @tkexpress11, @emilybenn12, @kenanhsaleh, @far33d, @marcussegal, @ndrewlee, & the rest of the team! It's time to build.

AI has helped resolve an important question in statistics. In the area of multiple hypothesis testing, the goal of controlling the false discovery rate (FDR) has been introduced in a seminal paper by Benjamini and Hochberg (1995). They also introduced a method (the Benjamini-Hochberg or BH method) and proved it controls the FDR. This method has been widely adopted in modern high-throughput science, including in genomics, astronomy, economics, etc. The paper has has garnered more than 130,000 citations to date. However Benjamini and Hochberg showed FDR control only when the data for the individual tests are *independent*. In practice, these data are often dependent; a good example is data on genetic variants due to linkage disequilibrium. Later work has focused on extending the validity of the BH procedure, e.g., to a form of positive dependence by Benjamini and Yekutieli (2001). The question of when the BH procedure controls the FDR has remained open. Over the last twenty years, many authors, including Reiner-Benaim (2007), Kim and van de Wiel (2008), Benjamini (2010), Sarkar (2023), Sarkar and Zhang (2025), have conjectured that the BH procedure controls the FDR for two-sided tests using any correlated Gaussian data. These authors have presented both theoretical and empirical evidence supporting, but not directly showing, the conjecture. With the help of AI (specifically GPT-5.6 Sol Pro), I have settled the question in the negative: The Benjamini-Hochberg procedure does *not* generally control the false discovery rate at the desired level for correlated two-sided Gaussian tests. This was done by exhibiting a Gaussian factor model for which, at a nominal level alpha=0.01, the false discovery rate is proved to be FDR>0.0104. There is a lot of interesting commentary to be made: 1. This result should be of interest to everybody in the field of statistics. Emmanuel Candes of Stanford University once called the false discovery rate and the Benjamini-Hochberg procedure "one of the two most important developments in statistics after 1950" (the other being James-Stein shrinkage). The present conjecture is probably the most central question about FDR/BH that was unresolved to date. 2. GPT-5.6 one-shot the problem after 90 minutes of reasoning, whereas with 5.5 I was not able to solve it even after iterating with multiple parallel agents for perhaps 20 hours. So the capability improvement is quite real. Exciting times to live in! 3. The argument is not especially surprising, but it does combine an asymptotic approach (standard for FDR analysis, see e.g., Genovese and Wasserman, Efron, etc) with a numerical certificate in a way that would be pretty non-standard in the field. Once we have the specific example, then straightforward simulations also support that the false discovery rate is indeed higher than the nominal value (see attached fig). 4. The current degree of violation over the nominal level is relatively small (0.104 vs 0.1). So the importance of this result is mainly conceptual. The practical implications remain to be determined. Overall, an exciting development! Preprint is available here (faculty.wharton.upenn.edu/wp…) and will be on arxiv tonight; supporting code is here (github.com/dobriban/BH).

Sunset Boulevard is closed indefinitely in West Hollywood after a massive, 100-year-old water main ruptured early Thursday morning, creating a large sinkhole directly under the traffic light at Sunset Boulevard and Holloway Drive

Vibe research will be the biggest trend in 2026. I am starting to see people casually do it. AI is getting good enough to do this as well.

Excited to share our new working paper: how much did the US standard of living really rise over the twentieth century — and when? With @bhwittenbrink Using 5.1 million Sears catalog listings, we build a quality-adjusted price index for consumer goods, 1900–1990. Two headline findings: growth was much larger than official statistics imply, and its peak came before WWII, not after. (1/9)

Today, we're proud to announce the future home of Port Alpha: Brownsville, Texas. Our next-generation shipyard moves from vision to reality through a planned investment of more than $3 billion to establish one of the world's most advanced shipyards, built for software-defined shipbuilding and autonomous maritime systems. Port Alpha is expected to generate more than $160 billion in regional economic impact for Cameron County and $264.5 billion for the State of Texas, while creating up to 10,000 direct jobs — which makes it one of the largest economic development projects in modern Texas history. More than a shipyard, Port Alpha represents a new model for American shipbuilding — combining advanced manufacturing, software-defined production, and autonomy at unprecedented scale. We're proud to continue building in Texas. Learn more: medium.com/saronic-technolog…

This is actually crazy, many of us thought about this lower bound for many years ... reddit.com/r/math/s/lrKVCd6W…

The best time to read the @bismarckanlys Brief on Moonshot AI the creators of Kimi K3 was September 24, 2025. The second-best time is now.

1700s-1940s: free market capitalism 1940s-present: state capitalist social democracy cc @benlandautaylor

National security, finance, and every aspect of life will soon be disrupted by self-taught masters of the leading AI models. Every country and individual should be prepared, writes Tyler Cowen. thefp.com/p/tyler-cowen-ai-m…