flir

Ok, so they're uniquely careful, and they also get to that threshold first (whatever it is).

What happens when everyone else (who is not so careful) gets to the same threshold three months later? How does them getting there first stop that happening?

It's an utterly self-serving argument and it's not even internally consistent.

> proof of humanity

I'm just wondering how you'd do that. Only thing I can see working is a web of trust.

That's the problem every Elite after the first one had.

The first one was dogfighting in space, all the others were so. much. space.

Training on prompts I can understand - that's kinda baked into the premise, and they've been explicit about it.

Publication, though? Slimy is right.

But the interesting question to me is: once they had a solution, what should they have done with it? I see two choices: bury it, or contact the mathematicians whose prompts they were listening in on.

"More leaders means more leadership"

Heh.

Also: make it stop!!!

Ok half was an exaggeration, but I think $10 might make quite a lot of them jump to Claude. Or DeepSeek. Remember, we're talking about a global user base here.

I settled on GPT Plus/Codex for my own work, while my employer settled on Claude on Bedrock (where I am, according to their metrics, using about $20 of compute a day, so I can believe that GPT Plus is a loss-leader). If I hit too many 5-hour timeouts I'm not jumping to Pro I'm jumping to Claude.

(Funnily enough, I just hit one. It'll reset in... three hours. I wasn't even hammering it, just writing some tests. I think I'm gonna just go reactivate my Anthropic account for a month).

I've spent my entire career trying to automate my job.

(I get what you're trying to say, I'm just finding the irony amusing).

Or more efficient models/faster hardware close the gap between here and profitability.

I wouldn't be surprised if every $10 you add to the subscription cost halves your audience.

And both are illegal.

(Depending on your local jurisdiction. But GP said "Online Safety Act" so I'm making an assumption).

There's nothing a swarm of LLM instances can achieve that a single LLM instance can't. It's the same software, but running in parallel. That doesn't unlock Mysterious Cosmic Powers.

"just for replication rate to vary depending on some property of the thing replicating"

That's a heritable trait, not selection pressure. Remember, we're talking about an infinite population size here, but if the pop. size was capped then the mechanism that decided who got to replicate and who didn't would be the selection pressure.

Can you ELI5 why you think the varying replication rate is, by itself, a selection pressure? Because your point's obviously going over my head, and your last comment did nothing to communicate it more clearly.

Add Opus 5.0 to your list. (GPT 5.6 Terra tried to give me some kind of driving-at-night-with-a-starfield thing, but failed quite hard).

(Just realised I should have said "heritable variation" not "adaptive radiation" up there. Oh well.)

"Evolution doesn't require selection pressure" is a new one on me, tbh.

Not convinced you've got evolution there. You've got adaptive radiation, but where's the natural selection?

The low-black-hole universes aren't going to stop "procreating" because high-black-hole universes exist. At most, all you can say is that if you select a universe at random, it's likely to be a high-black-hole universe.

And I think you can use that fact to argue the theory's wrong: because surely our own universe is an LBH universe? A trillion is rookie numbers. There must be universes that produce vastly more black holes than ours, so the chances of us ending up in such an LBH universe at random are vanishingly small.

> And universes aren’t constrained by a shared environment with limited resources – newborn universes aren’t all competing in a valley with a limited amount of grass.

Yeah, that's what I'm trying to say. Without competition, it's not evolution.

Got curious, so I did a very unscientific experiment. Playing 3 1080 music videos in FF pulls about 60W here. In Chrome, about 50W. No videos playing is about 30W.

If it's uBlock or 3.33W, I know which I'm choosing.

Sounds peaceful.

But seriously, web is asymptotically approaching zero usefulness precisely because of all those "capabilities". In those circumstances, can't you see a million nerds muttering "we'll build our own interweb! with blackjack and hookers!" And one of them might even get lucky, and get a tipping point. Gemini's just an example.

I don't think web's ever going away for corporate stuff, but I can imagine the cultural party moving next door, to a less capable protocol.

You might still in a power usage graph.

On the bright side, there will finally be a compelling reason to move to an alternative protocol like Gemini.

Problem is, no matter how many sub-agents you break the LLM into it's still stochastic parrots all the way down. I've seen no evidence that compartmentalised harnesses can reach solutions that a monolithic harnesses cannot, so I'm thinking the compartmentalisation is just an excuse to throw more tokens into the token furnace - it doesn't unlock a step change in capability. If there is evidence to the contrary, I'd be interested in seeing it.

There have been some successful attempts to do science like this (https://github.com/AstroPilot-AI/Denario). Denario generated 1k+ "solutions" on its way to winning the FAIR Universe challenge, and they all had to be fed through a fitness function of some kind to assess them. It also got stuck on a local maximum and had to be kicked in the ass by a human to get it off that.

When writing software, that fitness function is relatively closed (when I run it, does it do what I want?). With something as open-ended as managing a business... might as well roll dice, IMO.

My belief is that 20's LLMs have a lot in common with 90's GAs: the more clearly we can define "success" for a given task (the fitness function) the more successful they can be. Open-ended problems are somewhat beyond them right now, and, I suspect, forever.

(The comment that's currently at the top of the discussion, about making an "AI Boss" that remembered everything about the company, and fed the user three tasks a day? That's not an AI Boss, that's an AI Assistant with different framing.)

The necessary computer science, you could probably implement. The necessary computer engineering, however...

logged-out users have been blocked if using a VPN for a while.

> And from what I have seen, a good number of exhaust fans are straight up fake and just ventilate back in to the room rather than ducted outside.

Culinary Cargo Cult.

If this doesn't turn into universal paperclips about 150 years after an industrial revolution, I'm going to be severely disappointed.

We're heading for two devices, IMO. A useful one and one for communicating with bureaucracies.

for-prophet

[Gnomic comments go over people's heads (I'm guilty of writing these too). Gotta think in UX terms - write to make your audience's lives easier, not harder.]

You're right, and I think a lot of these projects hit the same problem: "what can this offer me that centralized platforms can't?"

Take the Gemini protocol: if someone invents a "Gemini killer app" it can be replicated over the web because the web's features are a superset of Gemini's features. Medium is the centralized equivalent of traditional blogging. Spotify is eating podcasting. The whole indie web could be emulated by Meta (it's not currently big enough for them to care). Much like capitalism eating subcultures, the existing LCD internet will consume, commoditize and re-sell all alternatives.

The only exceptions I can think of are decentralization itself (centralization can only offer an emulation of decentralization) and freedom from corporate censorship, and not enough people care about them to make the switch.

I'd love to hear counterexamples if anyone has them, because I don't like this conclusion.

Swytch Bike is one. Hub motor.

History of creativity, I reckon. There's always been places and times that are just fertile. 470BCE Athens, the Renaissance, 1900 Vienna, 1990s Bristol...

*bangs head on desk* You're reading a description and hearing endorsement.