Thread 106342961

AI can now do original mathematics. Hope you stemmies are stretching those buttholes, because pretty soon "getting dicked down" will be the only job for which you qualify.

>>106342961 (OP)
>Verified by Bubeck himself
The problem is we need to be able to review everything it does, so we can hardly use an AI for anything above intern level. Sometimes it hallucinates the whole thing, sometimes it can't count the Rs in niggerberry. Sometimes it does make good stuff. We can't let it loose to work by itself (which the intelligence explosion depends on).

>>106342961 (OP)
Was it really a requirement for you that the source of this information be A COCKSUCKING MOTHERFUCKING TWITTER SCREEN SHIT?

>>106342961 (OP)
It's over bros. AI will eat our lunch. Aichuds won, this time for real.

>>106342961 (OP)
>did new mathematics
Lmao at that retard

>>106343021
It literally did, goalpost-moving coper.

>>106342961 (OP)
>ai hallucinating new math
hold the presses

>>106343090
Cope luddite, ai is here to stay, chud. AGI is just around the corner.

>>106343039
You don't "do new mathematics"
You either discover a new field/branch in mathematics
Or you do mathematics to find a solution to a problem
"Do a new mathematics" is nonsence

>>106342961 (OP)
>new mathematics
>improving the known bound
So it's not new mathematics, it's iteration on existing math.

>>106343123
No chud, this is agi. Luddites are in the chopping block (codetroons, trannyartist, mathchuds)

>>106343039
"new mathematics" is something that requires rigorous proof of its' existence plus no contradiction to any other known axioms unless it also provides a solution to other fields if that axiom is to be considered false. What a load of shit.

>>106342961 (OP)
>Implying some autist didn't write out an outline for a proof on some Chinese frog breeding forum, from which it produced its own

>>106343123
That's what all math is, goalposter.

>>106343148
nice to assume if we can since otherwise it is original.

>>106342961 (OP)
nah

>>106343190
There is a huge difference between advancing some obscure bound in an existing field, and creating an entirely new field of mathematics.
>>106342961 (OP)
I'll admit that I don't know convex optimization, but there are many fields in math which have a shitload of "unsolved" problems with a total of 2 people who actually care about them.
Some people seem to be under the impression that doing proofs in mathematics automatically means you're a genius, which is really not the case.

[x] Language doesnt count
[x] Art doesnt count
[x] Voice doesn't count
[x] Reasoning doesn't count
[x] Math doesn't count
[x] Feel doesnt count

What is AI if not intelligence?

>>106342961 (OP)
Two more weeks

>>106344576
I want to add I'm not a goalpost mover. I actually think AI is useful, but we don't know how far it's going to advance. It's still below the level of a (good) human.

>>106344576
mathtards want to prove bounds
library devs don't care
users don't care, they want it to work

>>106344627
there are different mathtards out there. Not every field is about proving bound theorem #12873

>>106342961 (OP)
>>106342961 (OP)
I'm not impressed by your deranged ramblings you blue checkmark totoro tranime tranny shill

>>106342961 (OP)
>

>>106344829
kek fucking tranime tranny

If I can trust anyone it's a blue checkmark with the openai ghibli filter avatar

>>106342961 (OP)
i can also do original mathematics:
182498515 * 878878777411111 = 160,394,071,742,543,302,000,165

i bet you no one before me has ever multiplied those two exact numbers

>>106345687
this. "original mathematics" doesn't necessarily mean it's valuable.

>>106342961 (OP)

>Stemmies

You're definitely someone that grew up to be a neet lol

Is there actual proof of this? Because all I'm seeing is a sources: dude trust me

>>106343101
>>106343146
clankerphobes coping and seething. meat is obsolete

>>106342961 (OP)
good morning sirs

>>106345687
Holy shit I've done it as well

√(-i) * e^420 + 9 / 11

anti vax retards: 2 more weeks everyone will drop dead

AI fags: 2 more weeks everyone will lose their job

>>106342961 (OP)
I have a PhD in counting the b's in blueberry. I'm safe for now

>e/acc
>blue tick
>Ghibli slop avatar
>glazing AI
>"if you're not completely stunned by this, you're not paying attention"
faggots like this should be rounded up and launched into the sun

>>106342989
>The problem is we need to be able to review everything it does
Verification is something that needs to be done for human researchers as well. But this is solved by lean.

>>106345705
>being proud of wagecucking

Bruh