[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/pol/ - Politically Incorrect


Thread archived.
You cannot reply anymore.


[Advertise on 4chan]


File: agi.png (125 KB, 1248x779)
125 KB PNG
now some AI gets 85% on AGI test

humans get 68% on average

so its better than SOME humans in SOME form of work
>>
Are niggers human though?
>>
>>533762525
arc-agi is fake and gay testing, AI is going to burst because the liability of the error rates will exceed any low level replacement.
>>
File: memories.webm (296 KB, 300x240)
296 KB WEBM
>>533762609

well its debatable if they are

if a chimp can be trained well enough to cross the language barriers, it might perform nearby levels of black people or outperform them even

the main difference between chimps and blacks is that chimps dont really have a language or perception of language at all, blacks have a rudimentary langauge at least, if not any other signs of intelligence or thought

but chimps are actually really good at hieroglyph type memory tests if they are given rewards afterwards (proper encouragement)
>>
>>533762525
where's deepseek?
I tried it a year or so ago and it was comically bad
>>
>>533762868
I'm thinking it wasnt included in this due to it being "non-canon" because Chinese simply stole it

I would argue it sits at around 60%
>>
these metrics are a terrible way to measure "intelligence"
>>
>>533763036
They're all stolen, bro. The only difference is the amount of loss they're prepared to accept their tokens to be sold for.
>>
File: react4.jpg (105 KB, 600x1000)
105 KB JPG
>>533762525
I tried submitting some of my infographics to chatgpt 5.5 I am impressed that blob of code was able to extract the text from jpgs understand what it was about and summarize the topic in sections etc
.
.
REACTION FOLDER
http://drive.google.com/drive/folders/1j241y5Zkh-ng5NaqjkNzad_UqzRpOarm
google drive easy to use interface
>>
>>533762868
Deepseek 4 (just released today) is between Opus 4.6 and 4.7, so roughly equivalent to GPT-5.5 medium.
>>
Big Chungus 7.2 beats them all. 7.2! Hello?
>>
If it had crossed the line already you would know.

The thing freak the fuck out to do anything with has been done 1000s of times.

People go nuts when it builds a React todo list.

Hand it a decent size code base and it falls apart.
>>
>>533762525
>so its better than SOME humans in SOME form of work
just like my gps navigation 20 years ago
>>
>>533762525
The thing that AI is always going to BTFO humans on is that it never gets bored or tired and it can inherently see things as a multidisciplinary polymath. It sees in graphs and can map the whole tree while a human struggles with one segment.



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.