@sarker's banner p

sarker

hantavirus landfill tour guide

0 followers   follows 0 users  
joined 2022 September 05 16:50:08 UTC

Suddenly I cannot remember the color of your eyes

Or the things we said as we stood together for the last time


				

User ID: 636

sarker

hantavirus landfill tour guide

0 followers   follows 0 users   joined 2022 September 05 16:50:08 UTC

					

Suddenly I cannot remember the color of your eyes

Or the things we said as we stood together for the last time


					

User ID: 636

I mean, that was certainly my impression.

I think you made up a thing that people are doing and then wrote a post about how it's dumb that they are doing that.

If the model hallucinates a vulnerability that doesn't work, just throw it away and try again!

I mean, yes?

It's frequently the case that it's easier to check a result for correctness than it is to generate the result in the first place. This is especially the case if the problem can be formalized in Lean.

Yep. It's him alright.

No, he's talking about theorem proving, not formal software verification.

Who is saying anything about AI assisted software verification via Lean? I've literally never heard of anyone suggesting that we should do this.

I don't really understand your overall point. To whatever extent Phelps' success was said to be thanks to wheaties, it's obvious the same can't be said for Tao for the simple reason that wheaties are from 1924 (predating Phelps) and Tao won the fields medal in 2006 (predating ChatGPT). It's obviously a logical contradiction.

The other thing is that we are at the point where LLMs are solving open problems with minimal or no involvement from humans:

  • Automated theorem proving across open Erdos problems. Each solved problem cost a few hundred dollars, which is hardly an unbelievably large number.

  • A problem relating to sumset combinatorics. Gowers is also a Fields medalist and seems pretty impressed with ChatGPT here. Reading the post, it does not seem that it takes a Fields medalist to do the prompting here: "the era where you could enjoy the thrill of having your name forever associated with a particular theorem or definition may well be close to its end." This also doesn't seem to have cost an inordinate amount of money, although there isn't a figure given.

  • A disproof of a conjecture related to the planar unit distance problem. I'll grant that it's not explicitly clear what the human prompting looked like, but there's no mathematician whose name is attached to this so I can't imagine it was something that only a select few could have done.

This is not an exhaustive list.

In other words, you don't have to be Tao to find new results with these tools. I am sure your response will be that nobody cares about these results, but unless you were predicting beforehand that we would get to this stage but no farther, it's hard to take this seriously.

I've read several posts by Teortaxes where he didn't say all his opps think Chinese people are bugmen, for your information.

Wait, Teortaxes is Dase? How did he manage to stop rageposting for his twitter persona?

Can't wait for big game hunting with FPV drones.

It wasn't a bushman. It was the dentist.

How? That seems like a self correcting problem. Wolves have been around for quite a while without causing lasting harm via overhunting.

I don't see why intention factors into the cruelty of the fact.

I'm not a hunter, but I'd expect a dentist from Ohio would have an easier time quickly killing a lion with a gun rather than with a bow and arrow as part of a paleolithic LARP.

Another thing that complicates the 'wildlife management ' story is that, as far as I can tell, Cecil had never interfered with humans. It's one thing to cull animals that are killing livestock or humans due to being near pastures or villages, quite another to kill animals that are not.

In Cecil's case, the lion was "critically wounded" with an arrow and survived another 12 hours before finally being killed. 12 hours of pain and suffering with a critical wound sounds pretty cruel to me.

Haven't seen that one, but 𝖒𝖔𝖛𝖎𝖊 𝖙𝖍𝖗𝖊𝖆𝖉.

  1. A Private Life (2025). 3/5. Jodie Foster as a psychiatrist in Paris who convinces herself that her patient was murdered. RT shows a pretty poor audience rating, but I think it was competently done, even if it was obvious from the start that she wasn't murdered. Foster's character overall feels believable and real.

  2. Le Circle Rouge (1970). 4/5. Classic French heist film. The heist sequence is simply kino, and there's several other scenes that are pretty high quality. The plan isn't revealed until it's put into action, which keeps you rapt. Perhaps a little anticlimactic at the end, and my wife fell asleep.

  3. There Will Be Blood (2007). 5/5. Probably this movie has been discussed to death but I hadn't seen it before. The opening sequence does its best to make oil extraction seem evil and almost demonic, which is a little ridiculous and overwrought, but Day-Lewis's performance is just excellent. A very uncomfortable moment for me when I hear a speech like this and find it not entirely unrelatable:

    I hate most people. There are times when I... I look at people and I see nothing worth liking. I want to earn enough money I can get away from everyone. I see the worst in people, Henry. I don't need to look past seeing them to get all I need. I've built up my hatreds over the years, little by little. Having you here gives me a second breath of life. I can't keep doing this on my own... with these, umm... people. [laughs]

Had to settle for my second choice in university (EE) because I couldn't get in to study CS (and the actually hard to get in programs would have been right out).

What were the hard programs?

and let me tell you it's really fucking weird to keep receiving fan mail about a publication for a full decade from random people who've gone to the effort of figuring out your twice changed email address just for a single message

Were these, uh, Indians?

Why is this funny? Is he supposed to be some kind of British stock character? Or is it just supposed to be random chungus holds up spork?

This, of course, is a statement of fact that there exists at least one human who has infinite genders, since there are only a finite number of humans who exist, and if each one only has a finite number of genders, then the number of genders is, at most, the sum of products of only finite numbers, which is also finite.

Have you considered that people can change their gender arbitrarily often? And that genders don't need to be instantiated in a human to exist, because they exist in the realm of platonic forms until people discover them? You chud.

Thanks. Ever think about if there's a better way? The natural impulse is to put everything under one roof but it seems there's too much here.

Sometimes the stakeholders of those projects are in other companies stacked 5 deep (and I mean 5 companies deep, each with their own hierarchies) before you get to the end client.

Why are there five companies between you and the end client? The day to day of consultants is a source of endless fascination to me because every explanation I've seen is some impossibly bizarre and incomprehensible thing like this.

Uh, why do you report to more than one person?

Two. Anyone who says anything higher is suffering from the financial equivalent of hoeflation.

Strivers who actually minmax have to know what to minmax, which is still preferable to would-be minmaxers who don't know how or what to minmax.

Rotely following Honorable Mother and Father's directions on minmaxing extracurriculars to get into Harvard may be less pedagogically useful than you believe.

Your mistake is confusing the result of Goodharting metrics to within an inch of their lives with actual development of skill and habits. Developing habits is a total waste of time when you can get good results by working the system and have more time left over to study for Bio Olympiad.

Even if the child is only doing it because the parents want that resume padding, it is still good for the children, and the future adults those children grow up to be.

Nah. If you've met any of the people who used to do this stuff as kids, they just minimax it to get maximal reward for minimal effort, use their parents' volunteering as leverage to advance within the club, then become extruded homogenized IC drone product at FAANG.

Where do you buy this stuff?