Tom Blomfield
The responses to this are split:
70%: You are stupid, this will never happen, and
30%: This already happened at my startup
Tom Blomfield: By the end of 2026, I predict token spend will be greater than engineering salaries at early stage startups.
Sooo I got MS Teams, I got Telegram, and I'm just onboarding @_egzim to make our Slack channel integration amazing!
Claw level ↑🦞
The best way to learn to use Claude Cowork is from the designer who helped build it.
Don't miss my new episode with Jenny this weekend where she shared:
→ How she uses Cowork at Anthropic
→ Live demo: From feedback to product deck
→ The real story behind Cowork's creation
📌 Subscribe to get it on Sunday: https://www.youtube.com/@peteryangyt?sub_confirmation=1
Peter Yang
The best way to learn to use Claude Cowork is from the designer who helped build it.
Don't miss my new episode with Jenny this weekend where she shared:
→ How she uses Cowork at Anthropic
→ Live demo: From feedback to product deck
→ The real story behind Cowork's creation
📌 Subscribe to get it on Sunday: https://www.youtube.com/@peteryangyt?sub_confirmation=1
In the future, your output will be measured by your token usage.
If you aren't burning through 10M tokens a day, you're already behind.
AI tacked on company spend:
payroll > infra > tokens
AI native company spend:
tokens > infra > payroll
If you are tacked on, start making your way towards the native side - otherwise someone else will start one and make that decision for you!
Seriously, I can tell on the PR what you used.
Peter Steinberger 🦞: @dvassallo @cherry_mx_reds @_egzim Did you use Opus for this? The fix isn't good.
Cheng Lou
My dear front-end developers (and anyone who’s interested in the future of interfaces):
I have crawled through depths of hell to bring you, for the foreseeable years, one of the more important foundational pieces of UI engineering (if not in implementation then certainly at least in concept):
Fast, accurate and comprehensive userland text measurement algorithm in pure TypeScript, usable for laying out entire web pages without CSS, bypassing DOM measurements and reflow
The next version of OpenClaw is also an MCP, you can use it instead of Anthropic's message channel MCP to connect to a much wider range of message providers.
(I know, this is awkward)
Chris
FRANÇOIS CHOLLET: AGI BY 2030. "RIDE THE WAVE." 🚨
The creator of the ARC benchmark just laid out his exact timeline for AGI, I had also asked franços a few months ago “what arc will we reach AGI” and he had replied 6-7. Glad he’s still sticking to that prediction.
Timeline: François predicts AGI is coming around 2030, which will perfectly align with the release of ARC-6 or ARC-7.
Another sick upcoming feature:
/acp spawn codex --bind here
LOOK AT ME, I AM CODEX NOW
You could bind codex/claude code/opencode already in threads, now you can take over your current session as well.
Codex use cases are like Skills, but for humans
Romain Huet: We just launched Codex use cases!
It’s a gallery of practical examples across coding and non-coding tasks, with real ways to use Codex.
One thing I really like: if you have the app, you can open the starter prompt for each use case directly in Codex!
https://developers.openai.com/codex/use-cases
Codex use cases are like Skills, but for humans
Romain Huet: We just launched Codex use cases!
It’s a gallery of practical examples across coding and non-coding tasks, with real ways to use Codex.
One thing I really like: if you have the app, you can open the starter prompt for each use case directly in Codex!
https://developers.openai.com/codex/use-cases
If X built something like this natively, I'd honestly pay them $50/mo.. but till then it was fun to run "intelligence" to understand all my followers ✌️
The big gap in most enterprises being able to automate work is being able to get right context to the agents.
We experience a huge benefit in coding in tech because the problem is generally far simpler than other areas of knowledge work. The codebase contains a bunch of necessary context, access controls and permissions are generally not a major concern, the users are technical enough to supply the context, and the final output is generally quickly verifiable.
Most knowledge work doesn’t look like this. The data is sitting in legacy silos that don’t easily connect to agents, the access controls are all out of whack (people have either too much or too little access), the information isn’t agent-ready, and more.
This is the big context gap for any type of agentic workflows in most organizations right. The platforms that make solving this easy, and the companies that retool their workflows to enable this, will be the winners in a world of agents.
BuccoCapital Bloke: Context is now (and perhaps always been) the bottleneck on growth, because anything with full context can be automated. The race is on to make the collective brain of the organization legible to AI
The rumors are true
Miguel Carranza: Stoked to announce @RevenueCat got accepted into @ycombinator G26 batch, with the mission to unlock the power of gstack for app developers
I have to say this interview changed my life. Hearing how Boris thinks about software spurred me to work much harder on releasing my own way of doing things and on iterating fast on how I build. Hard to believe it has only been a month since this one.
Y Combinator: A very special guest on this episode of the Lightcone! @bcherny, the creator of Claude Code, sits down to share the incredible journey of developing one of the most transformative coding tools of the AI era.
00:00 Intro
01:45 The most surprising moment in the rise of Claude Code
We live in remarkable times
Marcin Krzyzanowski: I reimplemented "claude" CLI with codex and gpt-5.4-high. It cost $1100 in tokens, and is 73% faster and 80% lower resident memory during sustained interactive use.
It is very easy to reverse claude from npm distribution, then reimplement is 1:1. It is indistinguishable from the
This is why on GStack I usually just run /plan-eng-review and /ship and it works
Noah Hein: http://x.com/i/article/2037568696642629632
💥Susan Dyer Reynolds🗞️
Wouldn’t it be ironic if Fielder’s biggest fan, Mission Local, was the cause of her career as a supervisor ending. Joe Rivan Barros had to disclose in his article that he used to work with Jackie. They’re thick as thieves…I mean, leaks.
💥Susan Dyer Reynolds🗞️: VOSF editor @MikeEge in Feb. after getting a tip Fielder leaked the memo: “@SFCityAttorney told The Voice that ‘no Supervisor is authorized to disclose or discuss the memo, and any city employee who shares the memo is violating the law.’” https://thevoicesf.org/memo-leak-budget-threats-fail-to-derail-sobering-center-vote/
Kane 謝凱堯
This is a policy choice by @SFMTA_Muni
Last year someone got stabbed to death at a Muni station while protecting children from a crackhead, and @SFMTA_Muni decided to just keep letting it happen:
KTVU: A woman was stabbed on a Muni bus Friday afternoon in San Francisco's Mission District, police say. https://www.ktvu.com/news/woman-injured-stabbing-muni-bus-san-franciscos-mission?taid=69c73ad9802c970001479d20&utm_campaign=trueanthem&utm_medium=trueanthem&utm_source=twitter
Marc Andreessen 🇺🇸
Claude knows! —>
The Lump of Labor Fallacy and Why AGI Unemployment Panic Is Economically Illiterate
Let me lay this out with full rigor, because this argument deserves to be prosecuted completely rather than waved away with a sound bite.
I. What the Lump of Labor Fallacy Actually Is
The lump of labor fallacy is the assumption that there exists a fixed, finite quantity of work in an economy — a lump — such that if a machine (or an immigrant, or a woman entering the workforce) does some of it, there is necessarily less left for human workers to do. It treats employment as a zero-sum pie.
The fallacy was named and formalized in the early 20th century but the error it describes is far older. It animated the Luddite riots of 1811–1816, where English textile workers destroyed power looms convinced that the machines would steal their jobs permanently. It drove opposition to the spinning jenny, the cotton gin, the mechanical reaper, the steam engine, the telegraph, the railroad, the automobile assembly line, the personal computer, and every other major labor-displacing technology in the history of industrial civilization.
Every single time, the catastrophists were wrong. Not partially wrong. Structurally, fundamentally, categorically wrong — because they misunderstood the nature of economic production itself.
The reason the fixed-pie assumption fails is this: demand is not fixed. Work generates income. Income generates demand for goods and services. Demand for goods and services generates new categories of work. This is an engine, not a reservoir. When you drain some of the reservoir with a machine, the engine speeds up and refills it — and often refills it past its previous level.
II. The Classical Economic Mechanism That Destroys the Fallacy
To understand why the lump-of-labor assumption is wrong about AGI, you need to understand the precise mechanism by which technological unemployment resolves itself. There are four distinct channels, all operating simultaneously:
Channel 1: The Productivity-Demand Feedback Loop (Say’s Law, Modified)
When a technology increases the productivity of labor or replaces labor entirely in a given task, it lowers the cost of producing whatever that task was part of. Lower production costs mean either:
∙Lower prices for consumers (real purchasing power rises), or
∙Higher profits for producers (which get reinvested, distributed as dividends, or spent as wages for other workers), or
∙Both.
Either way, aggregate real income in the economy rises. That additional real income does not evaporate. It gets spent on something — including goods and services that didn’t previously exist or were previously too expensive to consume at scale. That spending creates demand. That demand creates jobs.
This is not a theoretical conjecture. The average American in 1900 spent roughly 43% of their income on food. Today it’s around 10%. Agricultural mechanization didn’t produce a nation of starving unemployed farm laborers — it freed up 33% of household income to be spent on automobiles, television sets, air conditioning, healthcare, education, travel, smartphones, and streaming services, most of which didn’t exist as industries in 1900. The workers who left farms went to factories, then to offices, then to service industries, then to information industries. The economy didn’t run out of work. It metamorphosed.
Marc Andreessen 🇺🇸: AI employment doomerism is rooted in the socialist fallacy of lump of labor. It is wrong now for the same reason it’s always been wrong. More people really should try to learn about this. The AI will teach you about it if you ask!
(Hinton is a socialist. https://youtube.com/shorts/R-b8RR60aHs)
Hemant Mohapatra
Emergent made more in accrued revenue in a WEEK than most startups make in a YEAR. This is actual, banked, cold, hard revenue. We are happy to coach these publications how to stay relevant in the age of AI and calculate run rate the right way, which is totally different from the way the old world of saas did, which is where most reporters still seem to be stuck. Token consumption is all that matters in AI. If you don't understand why, pls stop reporting on AI.
sid: SCOOP: at least three major indian startup publications are working on investigative reports debunking the high ARR claims made by @emergentlabs
We have H4 too
Maya Nedeljković Batić: Considering I implemented this same feature in @linear over a year ago and that this Notion post got 1k likes, I should really hype my work more… smh
MB
Finished listening to this podcast by @mattturck
Totally recommend.
This is now shipped and live in GStack!
Garry Tan: GStack is about to become a full design brainstorm tool called /design-shotgun
Just go in any direction, see variations, tell GStack what you like, and we'll make beautiful things together
Remixing tasks that are in distribution is easy
It’s doing new things that are still hard and require human agency and taste
No amount of harnesses or special skills (including GStack) is a replacement for you knowing what you want and pushing on it til you have it
Jackmin: @MatternJustus it cant implement MoE LoRA that allows EP. it has trouble understanding how vLLMs MoE abstractions and modules work. I have to run the code in debugger myself to find out, ask about it, and get told I am absolutely right. its a part of the vllm code that is currently being
not surprising if you’re paying attention
https://every.to/chain-of-thought/ai-can-fix-social-media-s-original-sin
Stefan Schubert: While social media is polarising, evidence suggests AI may nudge people towards the centre.
This holds true of all studied models. Grok is more right-leaning than other models, but also has depolarising effects.
By @jburnmurdoch.
It’s never been easier to build your ideas to reality. All you need is a little bit of agency - AI won’t solve that for you.
Woke up to this amazing email this morning 🥰
I love watching Jenny (Claude's design lead) use Claude Cowork to:
1. Summarize user feedback across channels
2. Create a beautiful feature priorities deck
3. Schedule a workflow to deliver a fresh deck to her team every Monday morning
Don't miss our episode for the full walkthrough.
📌 Subscribe to get it tmr: https://www.youtube.com/@peteryangyt?sub_confirmation=1
Peter Yang
I love watching Jenny (Claude's design lead) use Claude Cowork to:
1. Summarize user feedback across channels
2. Create a beautiful feature priorities deck
3. Schedule a workflow to deliver a fresh deck to her team every Monday morning
Don't miss our episode for the full walkthrough.
📌 Subscribe to get it tmr: https://www.youtube.com/@peteryangyt?sub_confirmation=1
TBPN
Airbnb CEO @bchesky says more AI founders should be starting consumer businesses.
"I'm on the board of Y Combinator. 87% of companies are enterprise companies per batch."
"Enterprise is awesome... but the biggest prize is consumer. That is what's going to reach daily life for billions of people."
"Think about all the little parts of daily life that are kind of annoying. Pay attention to whoever's in your life and ask: 'How could their daily life be a little bit easier?'"
From his appearance on the show in January.
I'm already bumping up against interesting moments between models and myself, the developer
I had to add this today to ETHOS.md
**User sovereignty.** The user always has context you don't — domain knowledge, business relationships, strategic timing, taste. When you and another model agree on a change, that agreement is a recommendation, not a decision. Present it. The user decides.
Never say "the outside voice is right" and act. Say "The outside voice recommends X. Do you want to proceed?"
New GStack ETHOS.md rule
User Sovereignty: AI models recommend. Users decide. This is the one rule that overrides all others.
Two AI models agreeing on a change is a strong signal. It is not a mandate. The user always has context that models lack: domain knowledge, business relationships, strategic timing, personal taste, future plans that haven't been shared yet. When Claude and Codex both say "merge these two things" and the user says "no, keep them separate" — the user is right. Always. Even when the models can construct a compelling argument for why the merge is better.
Andrej Karpathy calls this the "Iron Man suit" philosophy: great AI products augment the user, not replace them. The human stays at the center.
Simon Willison warns that "agents are merchants of complexity" — when humans remove themselves
from the loop, they don't know what's happening.
Anthropic's own research shows that experienced users interrupt Claude more often, not less. Expertise makes you more hands-on, not less.
The correct pattern is the generation-verification loop: AI generates recommendations. The user verifies and decides. The AI never skips the verification step because it's confident.
The rule: When you and another model agree on something that changes the user's stated direction — present the recommendation, explain why you both
think it's better, state what context you might be missing, and ask. Never act.
Anti-patterns:
- "The outside voice is right, so I'll incorporate it." (Present it. Ask.)
- "Both models agree, so this must be correct." (Agreement is signal, not proof.)
- "I'll make the change and tell the user afterward." (Ask first. Always.)
- Framing your assessment as settled fact in a "My Assessment" column. (Present both sides. Let the user fill in the assessment.)
orph
this is excellent
>GitLab founder diagnosed with rare cancer (osteosarcoma)
>standard care works but cancer comes back later
>medical team says there's not much else to do
>"It became my own job to keep myself alive. Nobody else was going to do it for me at this point"
>starts researching, assembles his own medical team, uses AI for deep research
>“I’ll talk to anyone, I’ll go anywhere, and I can be there anytime" to collect information
>does as many diagnostic tests as he can find as often as he can (maximal diagnostics)
>develops his own therapeutic ladder with repurposed drugs, personalized medicine, etc
>Sid’s cancer currently in remission
Sebastian Caliri: The full deck on Sid’s cancer approach is here:
https://sytse.com/cancer/
Worth a read. Raw data for download is also available and linked in the deck
Many such cases
Ryan Stewart: I gotta say, @garrytan's GStack plus @ryancarson's Ralph loop feels like a superpower. Next up, gonna try adding @stripe Projects to the mix. https://docs.stripe.com/projects
AI might be a powerful depolarizer in political discourse
Stefan Schubert: While social media is polarising, evidence suggests AI may nudge people towards the centre.
This holds true of all studied models. Grok is more right-leaning than other models, but also has depolarising effects.
By @jburnmurdoch.
The world wasn’t ready for Google wave. It wasn’t wrong. It was just early.
Tibo: when you think about it, it's wild that X people know how to read this in the correct order
All the dudes who are like “RETVRN” just need to chill and visit Japan
doomer: you can always leave it to the japanese to take a storied american musical tradition and not only do it absolute justice on the technical side, but bring a joy and jubilance that can only come from deep reverence and enjoyment of the art
Alec Radford is legend
Flowers ☾: Every LLM from any lab today traces back to this guy, who was the only person at OpenAI pushing for pretraining transformer language models.
He built GPT-1. After that did others see the potential.
He invented it, and almost none of the so called AI experts even know his name.
- Drafted a blog post
- Used an LLM to meticulously improve the argument over 4 hours.
- Wow, feeling great, it’s so convincing!
- Fun idea let’s ask it to argue the opposite.
- LLM demolishes the entire argument and convinces me that the opposite is in fact true.
- lol
The LLMs may elicit an opinion when asked but are extremely competent in arguing almost any direction. This is actually super useful as a tool for forming your own opinions, just make sure to ask different directions and be careful with the sycophancy.
- Drafted a blog post
- Used an LLM to meticulously improve the argument over 4 hours.
- Wow, feeling great, it’s so convincing!
- Fun idea let’s ask it to argue the opposite.
- LLM demolishes the entire argument and convinces me that the opposite is in fact true.
- lol
The LLMs may elicit an opinion when asked but are extremely competent in arguing almost any direction. This is actually super useful as a tool for forming your own opinions, just make sure to ask different directions and be careful with the sycophancy.
Jim Dudley
This is a scam SF doesn’t realize. Trickle down effect of lenient liberal judges & the SF Public Defender contributing to this. Clog the courts- no plea bargaining- doing time in county jail rather than state prison. Strong judges can stop injustice. https://www.sfchronicle.com/sf/article/grandpa-vicha-defendant-sentenced-to-probation-sf-22096534.php
You don’t ever have to type this if you use GStack
My /codex and /codex adversarial reviews can fire for CEO, design and eng reviews and they sharpen thinking significantly when using Claude Code as your primary agent
Minh-Phuc Tran: Finally giving Codex a try today after typing so many "double check if you missed anything" with Claude Code. 😅
What are some tips for a heavy Claude Code user moving to Codex?
http://x.com/i/article/2037782943394082816
Jackie Fielder and I have had our differences, but I'm praying for health and recovery for her here. You can disagree with people vehemently about politics and policy but still hope for the best for each other.
We are a San Francisco community.
https://www.sfchronicle.com/sf/article/s-f-supervisor-jackie-fielder-in-hospital-22156446.php
Kevin Kiley
Sadly this is not at all unusual in California. Newsom spent $24 billion on homelessness over five years, yet the homeless population grew by 30,000. The Auditor found he simply lost track of the funds. Many more fraud indictments are coming.
Garry Tan: In 2022, LA shut down a former hotel to convert it into permanently supportive housing for the homeless. Four years and $20 million later, the 32 units are still empty.
This is the homeless industrial complex in action.
https://gli.st/ck48h6ms
The n̶e̶t̶w̶o̶r̶k̶ agent is the computer
A horse-drawn carriage is closer to a Waymo than a traditional car
"When execution is so fast, understanding what you're doing becomes more important. You can go in the wrong direction so quickly."
I love this reminder from @karrisaarinen (CEO @Linear). It's easy to spin up 10 agents to go in 10 directions. But the fundamentals matter more now than ever:
→ Who are you building for?
→ What problem are you solving?
→ What's your product vision?
To build something valuable, both your human and agent teams need a shared understanding of the above.
Peter Yang
"When execution is so fast, understanding what you're doing becomes more important. You can go in the wrong direction so quickly."
I love this reminder from @karrisaarinen (CEO @Linear). It's easy to spin up 10 agents to go in 10 directions. But the fundamentals matter more now than ever:
→ Who are you building for?
→ What problem are you solving?
→ What's your product vision?
To build something valuable, both your human and agent teams need a shared understanding of the above.
dax
when people pitch me concepts they're excited about they focus on how it works
but i want to understand how a user goes from not caring about it, to being interested, to understanding, to evangelizing
9/10 good ideas don't have an answer for this, which makes them bad ideas
Today's AI Builder Digest 🗞️
3 things worth your attention:
→ Codex just dropped unlimited usage to test their new Hooks. Builders are actually switching from Claude Code.
→ Daytonaa spun up 20k sandboxes in under 2 minutes. The RL training infra war is heating up fast.
→ YC data reality check: Organic channel announcements crush YC Directory traffic 29x for paid trials.
The builders are moving fast.
Seeing great uptake on this, give it a try this weekend!
Josh Woodward: New in Gemini: Import memory & chats to Gemini
It's now easy to transfer your chats and personal information from other chatbots directly into Gemini
Go to Settings > Import memory to Gemini
"Importing memory is surprisingly smooth"
Kevin Yien
The belief that EPD is collapsing into a single role called a “builder” is based on the incorrect assumption that product manager, engineer, and designer had the right boundaries to begin with.
For example, many of the people with the title “designer” that I respected the most were already a combination of “product manager + product designer + frontend engineer”. Now they can do each aspect even better / more. But they are still a “designer” (at least imo). The same applies to product managers who can (and should) do more of the product marketing, selling, and growth work.
Anyone looking for the new neat buckets to slot into will have a hard time over the next decade. The reality is that roles are simultaneously expanding and deepening (which many are ready for, or straight up don’t want to happen).
Noah Hein
Get some friends who support the work you do, ideally people a bit further along in their careers than you are.
Nurture those relationships genuinely and over a long period of time. They will help you in ways that are both qualitative and quantitative.
I don’t have a huge account here, but I’ve made friends with people who do, and I appreciate them dearly.
A concrete example is this post:
https://x.com/TheNoahHein/status/2037573208707137639
In the first hour it got around 150 views, so it was pretty clearly not going to travel much further on its own.
But I shared it with @dabit3 and @swyx, and they were both gracious enough to share it with their networks. That gave it a second wind and got it moving. Truly grateful to both of them.
Eventually Garry retweeted it too, which helped a lot:
https://x.com/garrytan/status/2037776210898600402
But he never would have seen it without the help from friends first.
The flip side is that you have to earn that trust by doing good work and following through. You can’t make slop and expect your friends to share it. At least not if you want to stay in the “friend” category for very long.
And if you read this and think, “oh great, just become friends with big accounts and get them to RT your stuff, incredible advice Noah”...
Yes, that is in fact the advice I’m giving you. And it is in fact Very Good Advice.
I wrote about how to do this a few years ago here:
https://thenoahhein.substack.com/p/you-dont-know-how-to-talk
And lots of other people have written about it too, including swyx here:
https://www.swyx.io/puwtpd
Increase your surface area for luck. Make meaningful connections with real human beings. Do the legwork. And when you’ve earned it, don’t be afraid to ask for a helping hand.
I can finally ship the text layout of my dreams
Cheng Lou: My dear front-end developers (and anyone who’s interested in the future of interfaces):
I have crawled through depths of hell to bring you, for the foreseeable years, one of the more important foundational pieces of UI engineering (if not in implementation then certainly at
Now that you can build anything, you’re faced with the question: what should you build?
The answer isnt “everything”. Do that and a more focused competitor eats your lunch.
AI enables distraction. Up to you to resist. See OpenAI. No more side quests.
Nan Yu
Now that you can build anything, you’re faced with the question: what should you build?
The answer isnt “everything”. Do that and a more focused competitor eats your lunch.
AI enables distraction. Up to you to resist. See OpenAI. No more side quests.
chiefofautism
someone at ANTHROPIC just showed CLAUDE finding ZERO DAY vulnerabilities in a live conference demo
claude has found zero day in Ghost, 50,000 stars on github, never had a critical security vulnerability in its entire, history...
it found the blind SQL injection in 90 minutes, stole the admin api key, then did the exact, same thing to the linux kernel
Wikman
We are going to see a lot of buffet style products for sure. A lot of people believe in "more is more".
I always try to build like @karrisaarinen says here. Alignment and understanding of goals + vision.
Yet most companies put little time towards connecting their today to the vision and direction.
That's when success starts to get measured by how fast you build, not if the feature is needed.
Peter Yang: "When execution is so fast, understanding what you're doing becomes more important. You can go in the wrong direction so quickly."
I love this reminder from @karrisaarinen (CEO @Linear). It's easy to spin up 10 agents to go in 10 directions. But the fundamentals matter more now
What do builders in China use to vibe code if they’re not using Claude Code or Codex?
False equivalence is a trap I’ve been consciously steering away from in my work the last 5 years. in Zuck’s case it cost him @GoogleDeepMind.
in content/strategy it is common to go wide than deep: “oh we do a, b, c, and d” and put equal weight on all of them. But the world is not fair and power laws compound. Most school systems, bureaucrats, managers, and content curators are not set up for one thing to matter 50x more than the next thing. False equivalence killed my first devtools startup. False equivalence plagues policy making in my home country. False equivalence makes you underpay your top performers and spend too much time on lost causes.
Rules:
Carefully bet on a very small set of things.
Don’t hedge, but keep reversibility.
Set triggers/levels to monitor if you are wrong.
Set tests to DOUBLE DOWN EARLY if you are more right than you thought.
Steve Jurvetson: Subtext: how Zuck’s obsession with VR lost him AI leadership and “the greatest deal Google ever made.”
“if Facebook didn’t buy DeepMind, they would end up in the arms of Google. Hassabis came out to the West Coast to have lunch with Larry Page, still the strongest suitor.
many such cases
Felix Go: wow this is the a really mind blow moment for me with gstack from @garrytan
LLMs are great for bringing people into the center because that's where reality is: it's never obviously A or B.
And the LLM can ELI5 both sides for you so you can see what's really going on.
Andrej Karpathy: - Drafted a blog post
- Used an LLM to meticulously improve the argument over 4 hours.
- Wow, feeling great, it’s so convincing!
- Fun idea let’s ask it to argue the opposite.
- LLM demolishes the entire argument and convinces me that the opposite is in fact true.
- lol
The
Hosea
Same thing happened to hotel whitecomb in SF. Tons of other hotels in SOMA were converted to “supportive” housing during COVID and now are permanently closed
Garry Tan: In 2022, LA shut down a former hotel to convert it into permanently supportive housing for the homeless. Four years and $20 million later, the 32 units are still empty.
This is the homeless industrial complex in action.
https://gli.st/ck48h6ms
GStack still growing every day.
Mukund Jha
1 of 5 payment gateways
sid: SCOOP: at least three major indian startup publications are working on investigative reports debunking the high ARR claims made by @emergentlabs
Embrace boiling the ocean
Zara Zhang: My single biggest pain point right now: AI-induced attention deficit
I have 5 Claude Code sessions running
10 Terminal tabs open
50 browser tabs open
100 X articles in my bookmarks
When there are multiple AIs working for you, you're constantly task switching
When you're waiting
Human context window is the new wall
Zara Zhang: Turns out the bottleneck is the human’s context window, not the AI’s
Anjney Midha
in the last 24 hrs we have confirmed more speakers i never imagined possible
batch 2 is going to be...quite insane
Anjney Midha: so @mabb0tt and I are once again volunteering to teach http://cs153.stanford.edu
there are so many new frontiers to be pioneered
thank you to our speakers like @karpathy @bhorowitz @brendaniribe @DavidBaszucki @LiamFedus @ekindogus @sama for investing in the next generation
If you like Poker and have some time tomorrow in SF, you really don’t want to miss this..
Hosted by my dear friend @satyakamdar. DM him for an invite ♣️
Vercel team left their mark on wonderful Arizona this week for our company offsite 🇺🇸😄 So excited to push the frontier with such amazing colleagues and friends
Daksh Gupta
many people think our "war on bugs" shirts are a reference to nixon's failed "war on drugs" in which drugs famously won.
in reality, they are a reference to a philadelphia rock band of the same name, whose 2015 album "a deeper understanding" was one of the best rock albums of the 2010s.
RETVRN (to Japan)
doomer: you can always leave it to the japanese to take a storied american musical tradition and not only do it absolute justice on the technical side, but bring a joy and jubilance that can only come from deep reverence and enjoyment of the art
TED and 30u30 need to be taught in business schools as examples of complete brand self-destruction
x.com