Offshore
Photo
Carlos E. Perez
RT @mcxfrank: So of course I don't know the answer! But here are a few scenarios for thinking this through. Scenario 1 is classic nativist dev psych: innate endowment plus input make the difference. You use core knowledge to bootstrap concepts from your experience. https://t.co/ElRbYLXqiB
tweet
Offshore
Photo
Carlos E. Perez
RT @mcxfrank: The take-home here is that we are off by 4-5 orders of input magnitude in the emergence of adaptive behaviors. (That's the figure from above).

The big broad cognitive science question is - which factors account for that gap? I'll think about four broad ones. https://t.co/QZfUiwJO8X
tweet
Offshore
Photo
Bojan Tunguz
I'd like to see AGI do this. https://t.co/jb7JlwNsSP
tweet
Offshore
Photo
Amjad Masad ⠕
Beyond thrilled to help bring Google's insanely powerful AI technology to developers everywhere and accelerate the adoption of AI-powered software creation, while also bringing Replit to Google Cloud customers worldwide! https://t.co/fm0TQWmDNW https://t.co/IDu2x5vFmv
We're teaming up with @googlecloud.

Replit's 20M+ developers will get Google Cloud services, infrastructure, and foundation models. Idea to live software on Replit just got even faster. https://t.co/DMu48L0gLK
- Replit ⠕
tweet
Offshore
GIF
Sylvain Filoni
RT @etienne_mineur: Petits tests techniques (et rigolos) d'une intégration de Stable Diffusion en (presque) temps réel dans une application Unity (par Julien Hognon) merci @_kzr pour ce travail (et @fffiloni pour cette info).
ps : j'ai accéléré par deux la vitesse de cette vidéo https://t.co/VlE9DQSyed
tweet
Offshore
Photo
The Right To Bear Memes
RT @MidnightMitch: @SaraGonzalesTX The last thing you see before you change your pronouns to was/were https://t.co/HPKL2A21VV
tweet
Offshore
Photo
Carlos E. Perez
I think my mission in life is to build a GPT-like system that will demolish poorly constructed arguments like this. This way all these AI celebrities will finally be put in their rightful place (i.e., like an average human being). https://t.co/24razJsKBf
I have claimed that Auto-Regressive LLMs are exponentially diverging diffusion processes.
Here is the argument:
Let e be the probability that any generated token exits the tree of "correct" answers.
Then the probability that an answer of length n is correct is (1-e)^n
1/ https://t.co/BTd9Hro7gi
- Yann LeCun
tweet
Offshore
Video
D-ID
RT @guenterexel: 🚀 AI-Innovationen 2 🤖
Eine App, mit der ihr von Angesicht zu Angesicht mit ChatGPT sprechen könnt? Das gibt es von @D_ID_:
🗣 chat.D-ID kombiniert eine Echtzeit-Gesichtsanimation mit Text-to-Speech-Technologie, um ein menschenähnliches KI-Gesprächserlebnis zu schaffen. https://t.co/3SG39ocKvu
tweet
Amjad Masad ⠕
RT @ReedAlbergotti: Lotta AI news today but one that really caught my eye was the @Replit partnership with Google. It's kind of Google's answer to Github Copilot. Join me, @amasad and others to talk about the future of coding. https://t.co/9DJfu2xC07
tweet