there’s something about humans,
that i can’t quite put a finger on,
that clearly makes us “smarter”
than any LLM model… even though i understand,
that they already surpass us,
in many and increasing measurable ways… it’s all the immeasurable ways
that people can remain
GENERATIVE AI
-
The Immeasurable Ways Humans Remain Smarter Than AI Models
By
–
-
Elon Musk’s xAI Surpasses ChatGPT in Benchmark Tests
By
–
I'M BLOWN AWAY Elon Musk's xAI just beat ChatGPT in benchmarks The creator of Unreal thinks it's AGI What you need to know:
-
Grok 4 Generates Crowd Animation Spelling Hello World First Try
By
–
Grok 4 a codé ça du premier coup 😅 Nous allons commencer à voir des test de plus en plus fou
— VISION IA (@vision_ia) 11 juillet 2025
Traduction du tweet original :
C’est incroyablement bon !
J’ai demandé : « Crée une animation d’une foule de personnes marchant pour former les mots “Hello world, I am Grok” pendant… https://t.co/27I9UT7so5 pic.twitter.com/vr9D1ueDA5Grok 4 a codé ça du premier coup Nous allons commencer à voir des test de plus en plus fou Traduction du tweet original : C’est incroyablement bon ! J’ai demandé : « Crée une animation d’une foule de personnes marchant pour former les mots “Hello world, I am Grok” pendant
-
From GANs to Diffusion: Evolution of Generative AI
By
–
From GANs to Diffusion: Making Sense of Generative AI’s Rapid Evolution linkedin.com/pulse/from-gans…
→ View original post on X — @iainljbrown, 2025-07-11 05:56 UTC
-
o4-mini Outperforms Grok 4 on SciCode Benchmark
By
–
o4-mini beats grok 4 on SciCode, despite being way faster and cheaper. It feels like maybe grok 4 was fine tuned specifically for some specific popular high status tasks? (Still fairly impressive, but more of a party trick than real capability AFAICT.)
-
Accidental AI behavior creation during training impossible
By
–
I know of no way to accidentally create this behavior through training FWIW.
-
Typical AI Web Scraping Examples and Use Cases
By
–
Interesting. what's a typical example of scraping you do with AI?
-
Good AI Model with Unusual Prompting Techniques
By
–
He’s a good one!
(with some of the weirdest prompting) -
Grok 4 Heavy fairer comparison to o3-pro than o3
By
–
To be fair, o3 also isn't great at this — Grok 4 Heavy might be fairer comparison to o3-pro, which is what I used in my post