gpm

6 ms·

1.
by gpm 3d ago
> from 8 to 16 MB Yes
2.
by gpm 6d ago
Fascinating document - the headline (can't ready the article) talking about a bioweapon (i.e. a weird obsession of anthropic's) really buries the interesting part.
3.
by gpm 6d ago
> That seems completely at odds with "everyone will be replaced by AI" You yourself give an example of how it might not be, UBI. The assumption doesn't have to be likely here. It can be the case that you think society is p
4.
by gpm 6d ago
By making decisions under the assumption that society will continue. Even under the reduction to absurdity of everyone 's jobs being replaced in a year, there's still some system of allocating resources and they would still like
5.
by gpm 6d ago
99.5% of the time you can understand it (percentage varies with rating).
6.
by gpm 7d ago
How is the WPEngine case going anyways? Glancing at the docket [1] in an attempt to answer this myself... looks like the trial is scheduled for a bit over a year from now [2].. A ton of fighting over discovery... a motion asking for fairly
7.
by gpm 7d ago
Yeah, it's unfortunate when anyone goes off the deep end.
8.
by gpm 8d ago
Airplanes have pretty bounded damage. Generally you kill at most a few hundred people. Even weaponized a few thousand. This is a risk profile that allows risk taking with near misses and waiting until something goes wrong to fix it (though
9.
by gpm 8d ago
Yes? If we look at the math problems they're solving their just now reaching the human frontier... they weren't doing that before. And your comparison point is model released 2.5 months ago... saying for some use case you didn&#x2
10.
by gpm 8d ago
Mathematicians provide two complementary services bundled together. 1. Proving theorems - what AI can apparently replicate faster and better. 2. Creating definitions and new theorems from those definitions to prove, selecting which of the p
11.
by gpm 8d ago
> Unless [...] mathematicians are effectively useless? It's always been a bit bizarre that this isn't the case. Mathematicians are almost always working on problems that there is no good reason to expect to have utility in th
12.
by gpm 8d ago
I've actually spent time doing math and literally everyone I know who knows math learnt it by reproving things that people proved before them. I don't see why AI proving things makes this form of learning any more economically inf
13.
by gpm 8d ago
Or actually doing what they did here where it changes it to a enum Value { SimpleFloat(64bit value), ComplexNan(Heap pointer) Ptr(*const T) } You lose out on performance if you use the bit patterns in
14.
by gpm 8d ago
See also his previous thread from before the result was published (and before he knew it was coming [1]) on how a to this problem seemed increasingly likely to be solved by AI in a way that caused us to miss the insights that would traditio
15.
by gpm 8d ago
What I'd like to enable this use of LLMs more recklessly is a compiler with formal methods that lets me guarantee equivalence between the opaque optimized code and something actually understandable.
16.
by gpm 8d ago
I'd actually point at the other half of the title than the existing comment when being pedantic "Replacing [...] with a 64-Bit Word" isn't quite right, it was replaced with a manually packed 64-Bit Word and the occasion
17.
by gpm 8d ago
Eh... OpenAI spent significantly more than $1 million solving this...
18.
by gpm 8d ago
I wouldn't expect poking at millennium problems to be that rare in ChatGPT. They were uniquely successful - but it's probably not easy to check de-identified data for the presence of any of their work on the problem because it w
19.
by gpm 8d ago
I've seen it on PMC. Strangely fixed by switching to a different firefox container ("container" in this context: https://support.mozilla.org/en-US/kb/how-use-firefox-contain... )
20.
by gpm 9d ago
Oh, oops, you're right. Sorry.
21.
by gpm 9d ago
Edit: Removed incorrect claim that |B| > |A| sufficed for the counter example. It's also the definitions the book supplies though (and the standard ones). Mathematics works over definitions. Everyone is free to do math over whatever
22.
by gpm 10d ago
I like the framing in the headline - really more than the article itself. I think underappreciated in discourse like this is that of course we will still have "mathematicians" - people whose work is mathematics. The job might chan
23.
by gpm 11d ago
The sovereignty perspective here is the perspective of the sovereignty of the states that make up Europe not the sovereignty of the entity of the EU/Europe. Though not every sovereign state has an army, see Iceland, Costa Rica, theoret
24.
by gpm 13d ago
Batteries are rapidly improving (on many metrics, but the relevant one here is increasing in energy density). The range is apparently 125 miles electric, and 500 miles as a hybrid. That's not enough for Melbourne Sydney. Give it a co
25.
by gpm 13d ago
Has yet to catch up to Claude-451 from a couple of months ago.
26.
by gpm 17d ago
Yeah, I can see that. The aptera design has flexible panels molded to the body under the clear aerodynamic shell. It shouldn't suffer from aerodynamic losses but the panels probably lose efficiency from heat.
27.
by gpm 18d ago
To reduce hassle? To make it work when you live in a place where you park on the street and can't run wires to the car?
28.
by gpm 19d ago
In the Starbucks example wouldn't there theoretically be a defamation claim against the person who lied about you - just with absurdly small damages that aren't worth pursuing.
29.
by gpm 19d ago
Does Chinese hardware mean fabbed in China or designed in China and fabbed by TSMC?
30.
by gpm 29d ago
Huh, why would it shatter if the power goes out?

More ›