If your paper is a breakthrough, it’ll probably be rejected because the rewiewers won’t get it.
RESEARCH
-
Big Labs Still Nowhere Near AGI, Says Skeptic
By
–
I’ll happily eat my words if I’m wrong, but you can’t look at anything the big labs are rolling out and really believe “oh yeah these guys are just a couple OOMS away from birthing god.”
-
General Purpose Models for Robot Control Gaining Industry Traction
By
–
who isn’t working on a general purpose model for robot control nowadays
-
OpenAI Evals Team Publishes Benchmarks Showing Claude Superiority
By
–
I find it unimaginably based that the OAI Evals team keeps making benchmarks finding that Claude is better and publishing it anyway. they are 3 for 3 this year in acknowledging specifically how much Claude is better at tasks OAI care about. there is no sarcasm here folks. this
-
Build Your Own LLM: Practical ML Learning Roadmap
By
–
> tired of "learn ml in 5 years" energy? want the roadmap from: > "what's a token?" to
> "i built + shipped my own LLM"? phase 0: build from zero, no theory jail > math + code sanity check: > > if matrix multiplication scares you, hit 3Blue1Brown's linear algebra series
> > -

MMR1: Multimodal Reasoning with Variance-Aware Sampling
By
–
MMR1 Enhancing Multimodal Reasoning with Variance-Aware Sampling and Open Resources
-

Hunyuan3D-Omni: Unified Framework for Controllable 3D Asset Generation
By
–
Hunyuan3D-Omni A Unified Framework for Controllable Generation of 3D Assets
-
AI Tasks Show Economic Impact But Don’t Replace Human Jobs
By
–
These tasks have real-world relevance & were built for economic impact: they averaged 7 hours of expert work to do & showed expert "try, then fix" is a good workflow. But note tasks are not jobs, and this does not mean AI can just sub in for humans. http://
cdn.openai.com/pdf/d5eb7428-c
4e9-4a33-bd86-86dd4bcf12ce/GDPval.pdf
… -

HuggingFace Launches GDP Evaluation Dataset and Leaderboard
By
–
Really cool! Evaluation dataset on HF of course: https://
huggingface.co/datasets/opena
i/gdpval
… Would be interesting to create a leaderboard on http://
hf.co/spaces as well! -

AI Models Approach Human Parity on Real-World Tasks
By
–
After reading it, this does seem like a big deal Industry experts outlined important, real-world, hard tasks for AI to do. Other experts were asked to do the tasks themselves & yet others graded human & AI output Models approached parity with humans & AI is getting better fast.