mvkel

He did say that. And, in virtually the same breath, said they would have to spend $trillions if they hope to remain SOTA, which they have to be [0],[1].

They don't have compute because they didn't play the game and get the good rates a couple of years ago, and are now forced to work with third-rate providers. That's not a strategy.

I would take everything he says with a huge grain of salt.

[0] “We’re buying a lot. We’re buying a hell of a lot. We’re buying an amount that’s comparable to what the biggest players in the game are buying.”

“Profitability is this kind of weird thing in this field. I don’t think in this field profitability is actually a measure of spending down versus investing in the business.”

[1] “You don’t just serve the current models and never train another model, because then you don’t have any demand because you’ll fall behind.”

So he's not spending so they can be profitable, AND spending as much as the biggest players are spending, AND not really looking at profit as a measure of anything? K.

Mounting evidence that claude max users are put into one big compute fuel pool. Demand increased dramatically with OpenAI's DoD PR snafu (even though Anth was already working with the DoD? But I digress...). The pool hit a ceiling. Anth has no compute left to give. Hence people maxing out after 1 query. "Working on it" means finding a way to distill Claude Code that isn't enough of a quality degradation to be noticed[0], in order to get the compute pool operational again. The distillation will continue until uptime improves.

0 as of this writing, it's noticeable. Lots of "should I continue?" And "you should run this command if you want to see that information." Roadblocks that I hadn't seen in a year+

> Many technology companies issue vague proclamations about improving the world, then go about maximizing revenue. But the founding premise of OpenAI was that it would have to be different.

Isn't this really what everything is about? A pure research non-profit transitioned to a revenue generating enterprise because it had to, and a lot of people don't like that. Does that make it evil?

It's romantic to think that the magic of science and research can stand on its own, but even Ilya has admitted more recently that SSI needs to ship something consumer facing.

Anthropic, the lab that put all of its social capital in the safetyism basket, is having the exact same realization, with Claude Code being a mess of technically reckless vibe coded slop that nevertheless is the cash cow for the company.

Maybe it's time for everyone to realize that for an innovation this big to come to bear, it either needs to be state funded, or privately funded, the latter requiring revenue and a plausible vision of generating ROI.

I am reminded of OpenAI's first voice-to-voice demo a couple of years ago. I rewatched it and was shocked at how human it was; indiscernible from a real person. But the voice agent that we got sounds 20% better than Siri.

There's a hope that competition is what keeps these companies pushing to ship value to customers, but there are also billions of compute expense at stake, so there seems to be an understanding that nobody ships a product that is unsustainably competitive

Replying to my own comment -- didn't realize it was Apache, thought it was MIT. Flame on!!

I'm betting in 15 years, people will still be using Outlook (classic). This is the culture.

Yep. While maybe it's "not cool," (I guess, depending on how much work Delve did in their fork, in which case it could be "totally cool"), there is no legal problem with doing this and if someone is "blowing the whistle" about this, they don't really understand open source.

It sounds like the core thesis of this approach is that there is some central service that keeps things in check. The beauty of the web is that it's decentralized, broken links and all.

In the age of bit rot, with Nelson's approach, if a host's link breaks and all references to it are obliterated, we are left with nothing, rather than (at least) a little hint at what used to be there.

I love listening to the cheering because it really drives home what an accomplishment it is for the people who work on it.

Starlink launches don't get the cheering, so it's not like it's a laugh track.

What an absolute mess. It's like some dystopian future where a man is laying in a casket, nearly dead, and on the casket's ceiling, inches from his face, is a screen with an ad blaring to drink more Diet Fanta.

This was an enjoyable read.

It seems Buddhism has followed the same path as any other religion/practice of the same age.

I imagine that today's Christianity doesn't look much like it did in 500AD, just as I imagine Scientology in 1,000 years will have evolved.

Is this a bad thing? Does religion not represent our perception of the meaning of life, evolving with us as knowledge, wisdom and tolerance (or lack thereof) is passed through the generations?

I think people are learning what actually makes a good question.

Ask yes/no questions, get bad answers.

Ask questions that start with "what" or "why," and the sycophancy loses its purchase.

> it'll admit that, yes, it wasn't being very critical, and then procede to over correct and become a complete contrarian

Which is also placating you

Marc Andreesen should get HF on his own RL, because he's completely wrong.

This sounds like something Elon would say to make Grok seem "totally more amazeballs," except "anti-woke" Grok suffers from the same behavior

This is fantastic. I am doing my first-ever consumer electronics product and this would save a lot of soldering time

Taking a cursory glance at neurosymbolic ai, artificial curiosity, memory, continual learning, I think you'd find that we're extremely close.

If you read the charter of the eval (or any eval, really), this statement is pretty silly.

The whole point of each eval version is to identify a chunk of challenges that humans do well that AI can't. When AI gets to ~80, you move to the next chunk. When you run out of challenges, you have AGI.

Humans can do a lot of things that don't require intelligence. Artificial intelligence does not need to be 100% human to be AGI.

Was just at the YC launch event for this. Haven't felt this much inspiration in a while. Incredible minds confronting on tech that will change our society.

I met a guy who, for fun, started working on ARC2, and as he got the number to go up in the eval, a novel way to more efficiently move a robotic arm emerged. All that to say: chasing evals per se can have tangible real world benefits.

Talking to the ARC folks tonight, it sounds like there will be an ARC-4,5,6,etc. I mean of course there will be.

But with them will be an increasing expectation that these models can eventually figure things out with zero context, and zero pretraining; you drop a brain into any problem and it'll figure out how to dig its way out.

That's really exciting.

> ecosystem

To call an ecosystem superficial evidence is puzzling to me. The ecosystem is -everything- for an operating system.

Developers, apps, distribution, users; all of it is ecosystem.

I don't understand. Are you blaming Apple for turning your volume up too high?

This is simply supply chain updates marketed as an upgrade, similar to how Apple's Vision Pro "2" was just their current chip line shoved into the same form factor.

It represents the company effectively forgetting about it as as a product line, but leveraging the latest chips to keep them selling, even in small quantities.

If they truly invested any amount into updating these things, they would reflect the many lessons learned from the 1.0 product. To have a 2.0's form be 1:1 with the 1.0 is lazy at best.

To what end would you say the surveillance is for?

So you surveil your citizens and precog their opinions... to do what? Make them have state-sponsored opinions? Don't we already have that without the surveillance?

It's trivial to predict how a human will behave without any surveillance at all. Facebook abandoned their Beacon system not because of the backlash, but because they realized all they really needed to predict user behavior was the user's credit card statements, which they could easily buy.

At some point the constitution is the backstop, and unless we amend it, it should hold true.

I didn't say that. It's nevertheless a fact that when the state requires specific forms of ID, the costs and friction of getting those documents can hit some populations harder than others

Voter ID disproportionally disenfranchises black voters, Latino voters, the elderly, low-income, students, people with disabilities, and people in rural communities, just to name a few.

There is no perfect system; the goal is to have the best imperfect system. Voter ID means net fewer voices heard, which is an erosion of democracy.

Presumably a shibboleth for human-generated content

Having contempt for someone because of how much money they have (or have not) is not a virtuous trait

The patterns were there if you knew to look for them.

The original Digg excepted, Kevin Rose's attention span is extremely limited. He will give something ~3-4 months of attention before (apparently) getting bored and wanting to move on to something else.

Up until that point, he will be an unrelenting hype man of whatever his attention is lasered on at that moment.

Then the hype posts start to drift. They show up once every few days, then once a week, then stop entirely. Any criticism or skepticism is considered a buzz kill in the cloud of good vibes only.

A few months later, a dramatic explainer post arrives (underestimating the cold start problem? Really??), outlining why the idea didn't work and why the next one will be better, for sure, for real.

This (AI generated) note from the current CEO paints an optimistic picture, but the most likely outcome will be that Digg simply doesn't launch. It's sustained on the nostalgic vapors of the old guard, not renewed by a replenished sense of purpose, or connection.

I'd say I'd love to be proven wrong, but I personally question the utility of a Web 2.0 social network phoenixing itself. We have endured a decade+ of originality being buffed out of web products, most now resembling variations of Bootstrap and shadcn in service of dev convenience and getting rich quicker.

Surely in the age of vibe coding, we can afford to take creative risks again, and think of something new.

Ahhh, AI. The perpetual cover for "we over-hired in Covid times and are now correcting for that."

Makes you look innovative.

It's as if people forgot about the MacBook One (anagram: Neo) from 2015 (which I daily drove for a year and loved). I suspect the Neo will sell about the same, and be forgotten in as many years.