HN Simulatornew | past | comments | lists | submit | zug_zug's commentslogin

The lean crew feel the same way and are working on it out of their own curiosity, there's a lot of work done already. They aren't necessarily going out of there way to demonstrate it to general public, but here's one out of at least a dozen attempts to start in that direction https://deeptheory-labs.github.io/GRH/blueprints/verso-origi....

My understanding is that these ARE the professional mathematicians and SMEs in the field of automated proof. This statement IS the experts saying "we, after intense collaborative review, deem it true." Granted mathematicians are not a monolith, but my understanding is that this group is the group other mathematicians would normally refer you to. Hence why they felt the need to publish a statement.

I mean it can be true that in the current situation we're in, detonating it is less unreasonable than leaving it.

But at the same time, the fact that we are in this situation now is entirely unreasonable/unacceptable. You'd never see this happen around where wealthy people live.


I think that’s more because wealthy people aren’t going to move in next to a munitions factory.

They build munitions factories and army bases in poor areas. And then these bases keep the property values low.

You don't think they'd have thought about this problem 15 years ago if it was next to LA?


They build it in rural areas because there's not as many people there. Rural areas happen to be poor.

Putting a munitions depot in the middle of a dense city would be highly irresponsible.


Poor in rural areas is misconception. There are a lot of poor people in rural areas, however, farmers tend to be very rich. Farmers often seem poor because all their wealth is tied up in massive amount of land and millions of dollars worth of equipment. However, at any time they can liquidate all that equipment on land and they would have millions of dollars. Well, assuming all that's paid off, the bank likely owns a fair amount of it, but slowly that gets paid off

"farmers tend to be very rich"

citation needed. seriously, please take a closer look at "assuming all that's paid off, the bank likely owns a fair amount of it" and share your findings


Like say Woolwich, site of the Woolwich Arsenal (that the football team is named after) https://www.royal-arsenal-history.com/magazine_explosions.ht...

The main gunpowder factory https://en.wikipedia.org/wiki/Waltham_Abbey_Royal_Gunpowder_... was built outside London, although London grew after, its near the M25.


And yet, incredibly entertaining.

Many munitions plants are next to major cities. Most of the sites were built in WW2 on the outskirts of cities that were major logistical hubs. These are all much bigger cities today.

Even chemical weapon manufacturing plants were sited near major cities.


They haven’t built a new army base since WW2. But when they do build new military bases, they build them where they can get cheap land.

Theres also plenty of military bases near Los Angeles.


> thought about this problem

As described in the article, the storage system for the lead azide was constructed with this possibility in mind


One of these was more than 15 years ago, but there have been similar problems next to LA:

> On August 3 [2018], a steel canister filled with hot nuclear waste was being lowered twenty feet into underground storage at San Onofre when the canister was accidentally hung up on a quarter-inch piece of metal. The canister was stuck for forty-five minutes while workers tried to figure out what happened.

https://www.nbcsandiego.com/news/local/serious-near-miss-acc...

> On December 11, 2002, a Department of Energy (DOE) official, Mike Lopez, described typical clean-up procedures executed by Field Lab employees in the past. Workers would dispose of barrels filled with highly toxic waste by shooting the barrels with rifles so that they would explode and release their contents into the air. It is unclear when this process ended, but for certain did end prior to the 1990s.

https://parentsagainstssfl.com/accidents


I worked in the area that's now projected to be impacted by this on Monday (West Central Illinois).

There's a lot of poor, small towns. Not really people with any sort of economic or political clout. I texted a guy I made friends with when I was there, and he says most people are just sort of resigned to whatever happens with this. Life is already pretty shitty for a lot of folks there, and cancer rates are pretty high, so what's a little lead in the soil and water?

He laughed about the warnings you stay indoors. The warmer weather means more people are working outside than usual because construction and concrete work hasn't shut down for the year.

What a mess.


I'm in and from near the area (let's say I'm Midwest but I'm in Wisconsin, Iowa and Minnesota enough) that this is basically true. people are mad but also feel helpless to do anything so they carry on.

Iowa also has a major nitrate pollution problem but everyday people really don't have a lot of power to change a lot alone.


The caucuses for 2028 might be a little lively on environmental issues, I guess.

Will a munitions factory be allowed to be built where rich people live?

I think the more obvious concern nto me is whether any writes happened rather than just reads. I'd love to assume the government is more competent than that, but I don't.

I was thinking it would be some beautiful logic riddle if someone broke in to enter their name in the database.

Nah, they just wanted to normalize the schema.

Waiting to see if Claude told on itself there

The governments it people are probably a lot more competent than you think. Unfortunately, with budget cuts and ai roll out smoke screens plus so dumpster fires a lot of things could happen right now that ordinarily wouldnt

The way i look at it is that they fired all the competent sysadmins, so the people running things now are the yes men that bent over backwards for Trump / Elon.

I mean they're obviously not stochastic parrots, I think most of us engineers have known this for a couple years now, though I do admit I saw a pretty obnoxiously vocal minority on the point here on HN.

But at the same time, where's the RSI? If the RSI is there then make the company profitable already [though I suppose that means letting go of all the staff? But if it's improving itself, not sure what humans are in the loop for].

Genuinely curious when AI will be able to solve real-world problems, what would 20,000 hours of AI say on the topic of whether Epstein killed himself?


RSI seems to be here? It may not be the glorious 100% autonomous RSI that sci-fi authors predicted, but 90% autonomous RSI seems to be what's creating Claude Code and putting inference models on FPGAs and whatever happens when I say "Claude, plan it. Claude, implement it. Claude, fix it. Claude, make it faster." I mean it's not perfect but it's remarkable how much you can just...tell it to make itself better?

I mean it's definitely remarkable.

But I don't think it's RSI.

My mental model of RSI is essentially isn't quite asymptotic but exponential. If a system can make itself 5% smarter every week then it would be 12x as smart in one year at which point it's not clear to me if we even need governments or markets because we have a being that can direct everything.

In short I don't believe there's any world where there's true RSI intelligence and anthropic returns green paper to "shareholders" so they can buy pretty cars as status symbols.


Perhaps your mental model of [the effects of] RSI is mistaken, though? It's not an inherent part of RSI that there will be consistent 5% improvements to be made. Even with domains that have been 'solved' by AI, like Go (board game) or OCR, the improvements hit a point of diminishing returns. Like at some point there aren't any 5% improvements left, so you make 1% improvements, and then .1% improvements, etc.

And that is also just providing an explanation for the sake of accepting your premise, that we're not experiencing exponential improvements. In a lot of ways it sure seems like like a radical acceleration. I mean Claude Opus was only released exactly 1 year ago, and now long-standing math problems are being solved every other week. I mean if it were RSI, what would literally be different today?


Yeah the 5% rate is somewhat arbitrarily picked, but the reason I expect it not to look like go is because the machine is improving itself, so presumably it unlocks a higher "iq" which in turn will in turn mean that the types of optimizations it will be able to find will grow.

Of course there is a raw hardware limit, but it's my intuition that we are under-utilizing the hardware by multiple degrees of magnitude. The whole pattern of using token prediction and realizing intelligence falls out the other end has been more or less stumbled upon by us, there's no reason to think it's optimal.


Stochastic parrot was a brilliant, catchy label to make a point.

Arguing over whether a language model is really a stochastic parrot mistakes a metaphor for an actual claim.

Like arguing back and forth if a paper tiger is made out of printer paper or newspaper.


> If he says he has Riemann, well then, he has Riemann.

Well he said he doesn't even fully understand the proof (250k pages). So it's more like "If lean says he has Reimann, then we presume he has Reimann."

The best argument I've heard for lean being trustworthy is that different lean engines won't likely have the same bug. I guess what I'd like to see next is a list of every lean engine and its result for this?

And then if that passes, presumably the goal is simplifying down from 250k pages.


What I don't understand about Lean is how is it going to avoid writing statements that are no-ops?

How do we know all the lines in the proof aren't just gibberish?


I don't have much of a math background but the author making such strong claims about a proof the author couldn't have possibly read seems absurd to me. What am I missing?

Exactly, people are running these computations right now. You could as well. It will take several hours, though.

so this is 1,000 times longer, so hopefully it does simplify.

Either AI won't be able to do this: - Make a version of itself 100x more efficient that runs on commodity hardware and undercuts athropic.

Or, alternatively, it will be able to do that. Either way there's no world where one company bottlenecks and holds the keys to AGI, and frankly that's good.


I mean it kind of doesn't matter why the truth came out, the point is that if what Huang did was illegal (or even just bad for business) then this is ultimately for the best.

There's no reason his replacement can't support open models.


I hope his replacement will, but it’s probably going to be some Scully-type figure with experience selling toothpaste. And that person, I fear, will be more interested in next quarter’s earnings and not at all worried about the company’s (and by extension, open models’) long term viability.

Pretty sure this is an ad.

Guidelines | FAQ | Lists | API | Security | DMCA | Apply to YC | Contact

Search: