Por visualizar el cuadrante inteligencia vs precio, el punto que marco es el modelo más potente que teníamos hace justo un año. Todos los demás, más baratos, más potentes… es lo que se ha logrado en los último 12 meses.
RESEARCH
-
Modern SOTA LLM Reasoning Requires Ed’s Contribution
By
–
There's no modern sota LLM reasoning without you Ed
-

Live tweeting from Centring Data in African AI workshop in Kigali
By
–
I’ll be live tweeting from the Centring Data in African AI workshop happening in Kigali at the #DLI2025
→ View original post on X — @shakir_za, 2025-08-21 07:39 UTC
-

Frugal AI as innovation lever reducing energy footprint
By
–
𝐋𝐞ç𝐨𝐧 𝟳 – 𝗘𝗽𝗶𝘀𝗼𝗱𝗲 𝟮 𝗠𝗶𝘀𝗲𝘇 𝘀𝘂𝗿 𝗹’𝗜𝗔 𝗳𝗿𝘂𝗴𝗮𝗹𝗲 Et si réduire l’empreinte IA devenait le meilleur levier d’innovation ?
On pourrait croire que limiter la consommation énergétique freinerait la puissance technologique, mais en réalité, c’est tout le -
Neuroscience Theory Bottleneck Beyond Computational Power
By
–
We can, but unfortunately the neuroscience does not pan out. Computation is not the bottleneck, modeling/theory is
-
AI Gap: Workers’ Expectations vs Current Capabilities Study
By
–
A study conducted by @StanfordHAI and @DigEconLab scholars revealed a concerning gap between what workers want from AI and the reality of its current capabilities.
-

Cool Culture Celebration Party at Deep Learning Indaba 2025 Kigali
By
–
Cool culture celebration party at #DLI2025 Deep Learning Indaba in Kigali. pic.twitter.com/sxPLPPsDIL
— Max Welling (@wellingmax) 20 août 2025Cool culture celebration party at #DLI2025 Deep Learning Indaba in Kigali.
-

Model-Driven FLOP Allocation Achieves 80% Sparsity, 4x CPU Speedups
By
–
What happens when you let a model decide how to allocate FLOPs? Introducing Compute Where It Counts, a new trainable sparsity paradigm released by @crystalAIorg that beats SOTA methods, enabling 80% sparsity and 4x+ speedups on CPU.
-
AI Showdown: ChatGPT-5, Claude 4.1, Gemini 2.5, Grok 4 Tested
By
–
ChatGPT-5 vs Claude 4.1 vs Gemini 2.5 vs Grok 4
— God of Prompt (@godofprompt) 20 août 2025
I tested the smartest AIs in the world – side by side.
The results surprised me.
(watch the video to see who actually wins) 👇 pic.twitter.com/sFUBEyVrJQChatGPT-5 vs Claude 4.1 vs Gemini 2.5 vs Grok 4 I tested the smartest AIs in the world – side by side. The results surprised me. (watch the video to see who actually wins)
-
V-JEPA 2 Learns World Models from Million Hours Video
By
–
internet videos to robot actions? @AIatMeta
's V-JEPA 2 learns a world model from over 1 million hours of video, enabling it to perform complex, real-world robotic tasks zero-shot. Join author Nicolas Ballas for a talk on V-JEPA 2 this Friday!
