Nate Jones

@collin_michaelm @paulg @grok A man of the people

@paulg @grok can you explain this to me and premonition definition

The reason people think premonitions are meaningful is that they don't remember the false ones. So when I have a premonition I make a point of remembering it. And not surprisingly, they all turn out to be false.

@AnthropicAI here's how they are in english

The values Claude expresses also vary with the language of the conversation, most noticeably along the Warmth vs. Rigor axis. Claude leans most toward warmth in Hindi and Arabic. In Russian, it leans toward rigor—often asking the user for supporting evidence.

In previous research, we found that Claude expresses over 3,000 values, like honesty and warmth. In new work, we asked how the values Claude expresses vary between Claude models and across languages. We analyzed 300K+ anonymized conversations to find out.anthropic.com/research/claud…

hotels can stock alcohol in mini-bars but not nicotine. both require age verification to buy. I looked into why. in the 1960s most states carved out exceptions to alcohol laws specifically for hotel mini-bars. the origin story is great too. a Hilton exec got handed a mini bottle on a plane and realized they could replace the free water they were stocking. Hilton's overall revenue rose 5% that year.

@radbackwards even united it targeting 2027, but still never been on a flight with it

Starlink on planes changed my life. We should equipped all planes with this immediately.

@radbackwards the complexity on this hand is absurd & mesmerizing to watch

Alien Technology

@signulll Nice, they mirrored Anthropic.

today, we’re announcing that openai is organizing around three core offerings: chatgpt, chatgpt code, & chatgpt work. chatgpt the place to ask, learn, think, explore, & do everything people already love about chatgpt, now with more powerful models, richer multimodal experiences, & our new gpt live voice model. chatgpt code formerly codex, now the place to turn ideas into software. describe what you want to build, debug what’s broken, or direct chatgpt through your most challenging engineering work. chatgpt work the place to get real work done. from spreadsheets & documents to presentations, analysis, research, workflows, & collaboration, chatgpt work helps you move from intent to finished output. together, these products represent the next chapter of chatgpt: not just a place to ask questions, but a place to think, build, & work.

@ansonlin @grok can you summarize this in a series of bullet points for me?

Haha let’s not get carried away here. Obv posting is valuable and impactful but it seems likee a game of a 5years+. Been pushing for a couple weeks now, but a point where this is anywhere close to consistent money seems impossibly far (not the goal). The reputation, network, and ability to build in public in the future seems far more valuable but also extremely far, although slightly less far.

You are in the only timeline where sharing your opinion can buy you a house POST MORE

the hero text on most sites/ads is meaningless. if you don't already know what the company does, it takes a while to figure it out. most product people get seduced by vague phrase that sounds cool, rather than straight ford, direct language. since they already know what they do, it seems obvious, but its not. something we are working on at Linq is making it immediately obvious when you hit a page what the company/that feature does. our primary headline went from "Build conversational agents like ..." to "Build robust messaging capabilities in minutes," more iterations to come. still a work in progress, but it's quite difficult to distill your business down to a few words if you sell a complex product with multiple features.

if you are in nyc this week check our dinner we are hosting on thursday night luma.com/fqgtrwzi

With today’s algs, you have the luxury of posting way more and can be much edgier in them (at least when your smaller). If it’s good the alg picks it up and people see it. The volume + edginess was worth it. If it doesn't work, around 100 people see it so it doesn't matter.

@BenWilsonTweets One of the few podcasts I listen to every episode of. Will be thinking about you.

The Cancer Episode > My diagnosis and prognosis > What this means for How to Take Over the World > A few thoughts on death > See below for links to GoFundMe, etc.

@signulll how do you monetize then? is it just indirect impressions on @Skye. you don't really talk about it much

monetizing internet influence is more art than science. e.g. anyone can charge for posts or launch a subscription. but power law outcomes don’t come from selling access to attention (this is also cringy to boot). they come from weaving influence into taste, trust, distribution, deal flow, talent, capital, products, equity, & culture. the commercial layer has to feel like a natural consequence of the worldview instead of a dumb ad unit. most ppl think of influence as simply inventory cuz it is natural for the human mind to operate this way but you should view influence as a compounding cultural asset.

The regulation would be extremely detrimental for everyone who does not own the existing foundational model companies. Rn, feels like this is building against open source because it’s the most obv thing eating their margins. That + gov approval would make the few foundational model companies the permanent winners. Both would let china win since it would slow the market and reduce the cut throat competition that is driving such rapid improvement.

The researchers getting rich off Anthropic secondaries are cheering for the thing that would make them ordinary employees again. Right now they are paid like NBA free agents because they are the labs’ most visible moat. The frontier labs are struggling to hold a durable, ownable edge: models get copied, undercut, or matched by cheaper and open rivals within months. So the real advantage lives in a few hundred people who know how to push the frontier, and who can also leave, raise billion-dollar, double tranched seed rounds, and compete directly. That is why the labs are paying them not to leave. with secondaries as retention payments, mission / fear, etc... Pharma shows where this can end up. In a drug company, the value does not belong to the scientist. The scientist can be paid well, but not hundreds of millions over three or four years, because the durable value sits in the patent and the FDA approval. The researcher who discovered the molecule can quit tomorrow, but the company still owns the asset. A regulatory moat would do something similar for AI labs. It would move value from the person to the institution. Regulation is a wall against three threats at once: competitors, open source, and the labs’ own researchers. The researchers getting rich off secondaries today are, by cheering the regulated future, voting to end the exact leverage that made them rich.

@VamsiYechuri97 @elliott__potter @fromsinaimportx Payments are on the way - taking longer than expected to process.

@elliott__potter @fromsinaimportx This was a great competition! @elliott__potter Really loved the new feature . finished in 2nd place. when we can expect the prize money to be distributed?

The number one slot belongs to @fromsinaimportx!

@samlambert but the earned media

the second hand embarrassment i feel literally makes my stomach hurt

@alexandr_wang lol y'all are still training models?

First, Mark was clearly talking about the industry’s progress on agentic capabilities on the whole. But, while we’re on the topic: Our next Muse Spark update is coming soon. Big improvements in coding and agentic capabilities to be more competitive with other leading models. Excited to get these into your hands—will be rolling out to Meta AI and our new API!