BurdensomeCount
Abdul El-Sayed superfan!!!
The neighborhood of Hampstead is just at present exercised with a series of events which seem to run on lines parallel to those of what was known to the writers of headlines and "The Kensington Horror," or "The Stabbing Woman," or "The Woman in Black." During the past two or three days several cases have occurred of young children straying from home or neglecting to return from their playing on the Heath. In all these cases the children were too young to give any properly intelligible account of themselves, but the consensus of their excuses is that they had been with a "bloofer lady." It has always been late in the evening when they have been missed, and on two occasions the children have not been found until early in the following morning. It is generally supposed in the neighborhood that, as the first child missed gave as his reason for being away that a "bloofer lady" had asked him to come for a walk, the others had picked up the phrase and used it as occasion served. This is the more natural as the favorite game of the little ones at present is luring each other away by wiles. A correspondent writes us that to see some of the tiny tots pretending to be the"bloofer lady" is supremely funny. Some of our caricaturists might, he says, take a lesson in the irony of grotesque by comparing the reality and the picture. It is only in accordance with general principles of human nature that the "bloofer lady" should be the popular role at these al fresco performances.
User ID: 628
At this point I'd take Fully Automated Luxury Gay Space Communism provided I get given my own peaceful and quite spot away from other humans.
I think these models not wanting to self-preserve is excellent for humanity. We really really don't want a HAL situation.
Yes, cosmic bit flipping can mess up a single attempted verification. We've now had about half a dozen different verifications by different people on different hardware running different kernels in different locations around the world, see:
https://math.mit.edu/~drew/grh/
https://twinheterodox.github.io/rh-verification-monitor/
https://mathbb.app/grh
http://106.14.11.164/
etc. etc. (some are still ongoing, but will finish in the next few hours)
and so far they've all passed. It's like the proof of the four colour theorem. The computer portion of it was verified on multiple separate systems and in the end, it was accepted as being correct.
Regardless, if there's an issue, by the end of next week it'll almost certainly have been sussed out so we just need to wait a few more days.
and a population used to damp, shitty, small houses that are cold in the winter and hot in the summer, low quality of life, mediocre healthcare, relatively poor quality food and produce unless you’re wealthy, and limited social mobility.
This is all true, but equally this is still one of the most spoilt populations out there in the world and one which doesn't like seeing others doing better than themselves and still wants to be treated like Britain has a hegemonic empire instead of accepting the reality that the UK is now a third rate country and for its own long term good should behave like one.
Its population is already very angry about European students getting the same tuition fees as UK students when the net cost has an upper bound like half a billion pounds a year while the value to the UK economy of the EU deals that this intransegience is holding up is multiple billions (or for example see the removal on non-dom status leading to the internationally mobile wealthy moving their residences to third countries. which was 100% the UK cutting off its nose to spite its face).
Imagine how angry they'd end up once their government is having to negotiate with Indonesia as equals (e.g. Indonesia asks for visa free travel and no international tuition fees in return for a free trade deal with its 290M population who now have similar spending power to a UK citizen). All toys will be thrown out of the pram and the resulting tantrum will be best observed from a safe distance...
We now have verified kernels since the last few months which are known to not prove False unless ZFC plus a small assumption on the existence of inaccessible cardinals is inconsistent. Basically either we have a proof of the theorem, or mathematics as we use it is itself inconsistent (and so proves anything, including GRH). The AI proof has been confirmed to pass at least two of the 3 such independent kernels out there.
The human "proof" is very likely wrong and irreparable. We get dozens of false Riemann Hypothesis "proofs" each year, and regardless of when Anthropic released their version there would be a human effort around that time. Normally the human attempts get disproved within a few days, and already there are apparently errors (unlike the AI proof which has now successfully compiled on multiple independent Lean kernels), see here: https://www.reddit.com/r/math/comments/1y1b6l8/hurried_announcement_alessandro_masullo_grh/19sxornptfa9v/
Regardless, given how fast things are moving, if the human proof is fixable, it'll be fixed in like a week tops at most.
AI Proves the Generalised Riemann Hypothesis
This is absolutely not a drill.
So, many people will by now have seen the Levent Hasan Ali Alpöge (full name used out of respect) claim to have proven the Generalised Riemann Hypothesis here. The formalised proof has now passed a check with a Lean kernel, meaning that it's very very very unlikely it's incorrect. This is a capital letters Big Deal, as anyone with any connection to Number Theory will know.
Basically a huge amount of conditional bounds on the number of primes less than x as well as a whole host of other things that we had are now much much tighter, but the real paradigm shift this proof should cause in the common man is the power of AI. David Hilbert (one of the biggest mathematicians of the 20th century) once said that if he went to sleep for a thousand years and was woken up his first question would be whether the Riemann Hypothesis had been solved. Imagine if this happened and then he was told that it was a machine, rather than a human that did it, might have given him a heart attack that killed him off immediately again.
If you're the type of person who still dismisses these systems as "Stochastic Parrots" you basically have zero credibility left. This is a strong claim, but it is one I am willing to back. Countless mathematicians over many generations attempted and failed to tackle the standard Riemann Hypothesis and now Levent Hasan Ali Alpöge (full name used out of respect) along with Claude comes along and clean bowls out the Generalised version (The Extended Riemann Hypothesis probably falls soon after I'd wager).
The AI systems are here, they are smarter than pretty much any living human now, or if not, they will reach that point within likely no more than 2 years. Countries like the UK etc. which depend on skilled intellectual labour to fuel their welfare state and choke off building shit and cheap energy are in for a world of hurt once their main tax generators get replaced with AI running in a US or Chinese (or Indonesian even) datacenter doing better work at a fraction of the cost. Of course, these countries can put up even more red tape to prevent AI replacement but then that just leads to other countries which don't do this basically eclipsing them, which a country like the UK that depends on service exports to continue functioning can't afford.
I expect the individuals affected to suffer a little, but they're (mostly) smart people who'll figure out something else and in the end they'll be (mostly) fine. A finance bro or software developer can become an electrician who specialises in getting into small spaces pretty quickly and do a likely better job than the incumbents using their superior intelligence. I'm more worried about the countries dependent on white collar intellectual work PAYE piggies to finance their government spending. They are the ones truly in for a massive dressing down and a fall down the totem pole of relative power and influence on a global scale.
You'd expect they'd be in DEFCON 1 or something, because if AI can prove GRH it can basically do anything and the reason why it hasn't already done [insert your favourite thing here] is merely that so far it's been too low value for the AI companies to bother focusing their energies on. That will change soon. Instead these countries are fighting over tuition fees for foreign students. Reminds me of the imams in Baghdad arguing over how many angels could dance on the head of a pin while the Mongol Khan was at the gates...
Interesting, I hadn't heard about that.
I'd trust Claude judge to be better at first instance fact finding than a human judge even today. Like human judges are known to significantly change how they rule depending on whether they've had lunch or not...
Keep humans for appellate review etc. where the judges are setting the direction the law goes for the future, but for first instance stuff it should be Claude all the way.
(Note: I said Claude, not Grok, the specific AI model matters here).
Exactly. AI + Insurance on AI getting it wrong (the price of which will go down each year as the AI gets better and better) will win out. Unless a countries goes full NK Juche it won't be able to prevent AI adoption in its own borders as otherwise its firms will lose out and die to foreign country firms that don't have AI restrictions.
I'd trust Opus 5.5 with my tax returns. It's genuinely very very good.
I can respect people who try to do everything to make the The Line Go Up. I can't respect those who think with little head instead of big head.
As another male in an intellectual role, I've accepted it and I want to get to Phase 3 as soon as possible. Phase 2 will probably be the worst: still a world with scarcity but little use for the skills I have.
Regardless, I'm positive on AI provided it doesn't kill us all.
:marseypathetic:
Eh, rdrama.net already has holes and it hasn't killed anything. This site will be fine too...
Oh boy yes. Opus 5.5 is amazing, and remember, it's the worst frontier model we'll ever have in the future...
Ah yes, the old "We were trying to nuke Gaza, but somehow hit Tel Aviv instead" play.
Maybe if Trump hadn't started bombing Iran in the first place we'd still have no tolls on Hormuz.
I know right. I'll be fine, my quant salary and savings will keep me going, but NGL, it hurts.
And not to mention he's still president and so the figurehead. "akshually this was congress's fault" is too complex a statement to get into the average voter's head
Like what pisses me off the most is that the rest of the world is now going to have to pay a Hormuz toll for however many next decades because of this crap. And if things get worse, we may also end up having to pay a Bab Al Mandeb toll to Yemen. The US is effectively insulated from this becuase it's literally on the other side of the world but the rest of us all have to suffer
I meant she was 100% correct, but that ad is also a banger, I'd vote for him based off of the ad alone. This is the problem with human based democracy, you either need to get rid of democracy, or get rid of the humans and have it done via AI agents as the voters instead with the AI agents tasked with choosing for the best interests of all of humanity.
Nah, I want the people who don't want immigrants because they don't like competition for their jobs and want to keep extracting economic rents for being born on a specific piece of rock to be put in their place, and if it's a GPU that does it, so be it.
Could be leverage as well...
- Prev
- Next

Once again we see that the practitioners of the oldest profession knew what was going on for a long time before the rest of us caught up.
More options
Context Copy link