Vibes & Benchmarks

What happens when being pretty good at your job stops being enough to pay the mortgage?Take designers. Your average designer? Claude might be able to do a better job.Your cracked top 1% designer? Now much, much higher leverage, and in even more demand. AI tears through the pixel-pushing, leaving her free to create the vision and take the product from good to great. But what happens to everyone who used to earn a living doing that middle work?We're joined by Louise Sheiner, economist and senior fellow at the Brookings Institution, where she studies monetary and fiscal policy.We discuss whether AI is creating an “NBA Economy,” where it's LeBron or bust. And what fiscal response that might warrant: Government ownership of capital? Much higher taxes on capital gains? Ending tax breaks on inherited wealth? A wealth tax? A direct tax on compute?Plus: Why the next recession could reveal much more about AI’s impact on hiring than today’s unemployment numbers.

Is Claude "conscious?" Should Anthropic require we be nice to it, or is that the beginning of a slippery slope?And are debates about AI "alignment" really technical questions, or subjective moral questions dressed up as science?Also: Meta's comeback with the release of Muse Wonder and Muse Wonder Full, Zuck's lack of interest in "pacing" the frontier, and why he's not interested in discussing his "P(doom)."What OpenAI solving hundreds of open math problems tells us about AI's ability to turn fuzzy problems into something measurable, then crush every benchmark. Once we can define what “good” looks like, how quickly does AI get better than us?And why 2027 is going to be all about AI, all the time. Politicians, your mother, your grandmother. No one is going to talk about anything but.Get ready.

SNL’s brilliant Dario Amodei sketch hits a real question: if the people building AI believe it could wipe us out, THEN WHY ACCELERATE?Even the benefit everyone loves to point to — potentially curing cancer — doesn't really hit when leaders like Dario are also agreeing that there is a 10% chance that AI wipes out all of humanity within the next 10 years.Asked whether AI could actually cure cancer, SNL's Amodei replied, "Let me put it this way… in 10 years, there’s about a 10% chance that cancer won’t be a problem for anyone."Yikes.

Open-source AI has a huge new coalition, Anthropic is getting dragged for its position, and the argument is really about who gets to control intelligence.Ali Rohde and Josh Albrecht also get into why AI's near-term threat to knowledge workers may be too much work, not no work; why engineering used to protect the rest of tech by moving slowly; why good and bad engineers are getting further apart; and what today's “dark software factories” are still missing.Plus: Kimi K3, neoclouds, continuous agents, Google capex, and whether data centers can actually be good for the counties that host them.

OpenAI model escaped its sandbox, but is that an almighty model or just bad engineering? We unpack why engineering failures will keep happening and what reliable AI systems need to do differently.We also cover why the battle between model providers is all going to come down to who can do inference most profitably, which, as of right now, gives Google a big advantage. We also debate whether we can ever build the primitives needed for AI to actually help with large chunks of knowledge work. And, if so, how.Vibes & Benchmarks is a weekly AI news podcast hosted by Ali Rohde (Outset Capital) and Josh Albrecht (Imbue + Outset Capital). Each week, they pick apart the latest news, from startups to SpaceX to OpenAI. They argue through what is actually new, what is overfit to Twitter, and what might still matter 6 months from now.Spotify: https://open.spotify.com/show/033s9Dt...Apple Podcasts: https://podcasts.apple.com/us/podcast...Substack: https://outsetcapital.substack.com

Claude can handle syntax. It still can't hand a junior engineer the judgment behind good software.Josh Albrecht walks through a code-review lesson from one of Imbue's interns: don't build a custom scheduler when cron or anacron already works. Choose stable tools, keep architecture simple, and write as little software as possible.Ali and Josh also cover what sustained model progress would mean for frontier-lab valuations, why same-rate progress isn't recursive self-improvement, and what Anthropic's JSpace work does and doesn't show about Claude.Vibes & Benchmarks is a weekly AI news podcast hosted by Ali Rohde (Outset Capital) and Josh Albrecht (Imbue + Outset Capital). Each week, they pick apart the latest news, from startups to SpaceX to OpenAI. They argue through what is actually new, what is overfit to Twitter, and what might still matter 6 months from now.

AI is becoming mandatory and non-differentiating at the same time. You have to use it because your competitors are using it. But once everyone has the same tools, no one gets the edge.Josh and Ali get into SpaceX buying Cursor for $60B, the vertical stack bull case, distillation double standards, Anthropic getting regulated 2 days after asking for regulation, and why "just tell Claude you're a defender" is not an AI safety strategy.

SpaceX may be the weirdest AI IPO story of the year. Ali and Josh talk through why the "space + Elon + AI" story may not make sense on its own, why the real bull case might be chips, and how OpenAI and Anthropic are using policy language to shape the market around themselves.V&B covers SpaceX's retail IPO pitch, Elon's reality distortion field, Anthropic's pause plan, OpenAI's proposed government stake, recursive self-improvement, AI loops, and the point where code review starts becoming "Did the AI make it good?"Ali Rohde is the co-founder and Managing Partner at Outset Capital.• LinkedIn: https://www.linkedin.com/in/ali-rohde-90719970/• X: https://x.com/RohdeAli• Website: https://alirohde.com• Outset Capital: https://outsetcapital.comJosh Albrecht is the CTO of Imbue and a partner at Outset Capital.• X: https://x.com/joshalbrecht• LinkedIn: https://www.linkedin.com/in/joshalbrecht• Website: https://joshalbrecht.com• Imbue: https://imbue.com

Co-hosts Ali Rohde and Josh Albrecht debate the hot topics that emerged this week including Bernie Sander's bill to give the public a 50% public ownership stake in Anthropic and OpenAI, whether frontier labs have lasting product-market fit, how AI is making us stupider, and why AI can make workers more valuable in the short term — until AI gets so good and companies have extracted all the domain knowledge needed they too are replaced.00:31 Bernie Sanders wants the public to own 50% of AI. If Bernie's proposal isn't right, what counterproposal are frontier labs like OpenAI, Anthropic putting forward?04:07 If unemployment stats don't yet show the effects of significant job loss, it too early for AI labs to work on UBI?06:12 Do Anthropic and OpenAI have product-market fit. If so, will it last?08:19 Who actually pays for all those tokens?09:45 Why the AI "token shock" headlines may be bullish for the labs12:28 "Why am I paying you to say make it good?"14:32 Is AI making us stupider?17:34 How to use AI without losing your craft19:23 Jevons paradox applied to token spend20:36 Is literally anyone actually working less hard because of AI?Ali Rohde is the cofounder and Managing Partner at Outset Capital.• LinkedIn: https://www.linkedin.com/in/ali-rohde-90719970/• X: https://x.com/RohdeAli• Website: https://alirohde.com• Outset Capital: https://outsetcapital.comJosh Albrecht is the CTO of Imbue and a partner at Outset Capital.• X: https://x.com/joshalbrecht• LinkedIn: https://www.linkedin.com/in/joshalbrecht• Website: https://joshalbrecht.com• Imbue: https://imbue.comFollow us on:Youtube: https://www.youtube.com/@OutsetCapitalSubstack: https://outsetcapital.substack.com

Outset Capital GPs Ali Rohde and Josh Albrecht discuss what happened in AI this week.This week: George Hotz declares agents can't program, why macOS gets worse before it gets better, whether "100% written by Claude Code" is a flex or a red flag, the SpaceX S-1 and the case against a $2T valuation, Uber admitting AI token spend is "harder to justify," AI solving Erdős problems, the Pope's surprisingly good AI take, why Eric Schmidt got booed at graduation, OpenAI's reported IPO, Anthropic's 23,000 vulnerabilities, and whether AI is making us more or less safe.Josh Albrecht is the CTO of Imbue and a partner at Outset Capital.• X: https://x.com/joshalbrecht• LinkedIn: https://www.linkedin.com/in/joshalbrecht• Website: https://joshalbrecht.com• Imbue: https://imbue.comAli Rohde is the cofounder/Managing Partner at Outset Capital.• LinkedIn: https://www.linkedin.com/in/ali-rohde-90719970/• X: https://x.com/RohdeAli• Website: https://alirohde.com• Outset Capital: https://outsetcapital.comOutset Capital is a San Francisco–based pre-seed and seed fund investing in AI, devtools, robotics, and the future of work. Josh Albrecht is CTO of Imbue and a partner at Outset; he spends in the hundreds of thousands of dollars a year on tokens, runs his own coding harnesses at Imbue, and has very spicy opinions about where the rest of the industry is wasting its money.