see Marcus on AI at Substack earlier today
RESEARCH
-
Three reasons AI’s jagged intelligence is harder than human jaggedness
By
–
Things that make the jagged intelligence of AI harder to deal with than the jaggedness of humans:
1) Weaknesses are not always intuitive or identifiable in advanced
2) All LLMs have similar weaknesses, so you can't just hire a different one
3) Jagged frontier is moving outward -
Setting up GLM-5.1 4-bit on 4x DGX Sparks cluster
By
–
Qwen3.5-397B-A17B-FP8 via OpenCode
— Ahmad (@TheAhmadOsman) 10 avril 2026
Helping me setup GLM-5.1 (4-bit)
On the 4x DGX Sparks cluster https://t.co/OTKXZ2vzkm pic.twitter.com/somZODNXQxQwen3.5-397B-A17B-FP8 via OpenCode Helping me setup GLM-5.1 (4-bit) On the 4x DGX Sparks cluster
-
AI Should Be Nurtured, Not Trained: The Promise of Neuroevolution
By
–
"AI should not be trained, but should 'grow' on its own." Sakana AI researcher @sebastianrisi appeared on the podcast @EyeOn_AI. He discussed an overview of the Neuroevolution method, which constructs neural networks through evolutionary approaches, and talked about the current state of continual learning and artificial life (ALife) research. piped.video/pPpDxB4N_mE [Translated from EN to English]
-
AGI Not Within Striking Distance: Narrow AI Advances Instead
By
–
We are not getting to the G in Artificial General Intelligence; we are getting to (impressive) advances in particular areas where particular (verifiable) techniques can be used, on problems with advantageous economics. AGI itself is NOT “in striking distance”; inferring that
-
INSPATIO-WORLD: Real-time 4D Explorable World Simulator
By
–
INSPATIO-WORLD: Real-time 4D world simulator
— DailyPapers (@HuggingPapers) 10 avril 2026
Turn any video into an explorable, interactive 4D world you can navigate in real-time using WASD controls. Built on spatiotemporal autoregressive modeling with state-anchored world states, running at 24 FPS on NVIDIA H-series GPUs and… pic.twitter.com/BFd1l3jzMtINSPATIO-WORLD: Real-time 4D world simulator Turn any video into an explorable, interactive 4D world you can navigate in real-time using WASD controls. Built on spatiotemporal autoregressive modeling with state-anchored world states, running at 24 FPS on NVIDIA H-series GPUs and ranking #1 on WorldScore-Dynamic. [Translated from EN to English]
→ View original post on X — @scobleizer, 2026-04-10 00:28 UTC
-
Memory Consolidation and Consciousness Digitization: Scientific Skepticism
By
–
We barely understand how memory consolidation works, let alone digitizing consciousness. The confidence on this is… something haha.
-
AI Cracks Animal Communication Code: Future is Here
By
–
PEOPLE ARE USING AI TO CRACK THE CODE OF ANIMAL COMMUNICATION
— 0xMarioNawfal (@RoundtableSpace) 9 avril 2026
WELCOME TO THE FUTUREpic.twitter.com/12zKktDU4CPEOPLE ARE USING AI TO CRACK THE CODE OF ANIMAL COMMUNICATION WELCOME TO THE FUTURE
→ View original post on X — @ceobillionaire, 2026-04-09 23:45 UTC
-
OpenAI’s Path to Automated AI Researcher by 2028
By
–
Compute powers every layer of AI, and the investments we’ve made mean we can run more promising research experiments, train more capable models, and support broader access. @merettm talks about our progress building an automated AI researcher and what’s ahead as AI can take on… https://t.co/zM7iFZYAsK
— OpenAI Newsroom (@OpenAINewsroom) 9 avril 2026Compute powers every layer of AI, and the investments we’ve made mean we can run more promising research experiments, train more capable models, and support broader access. @merettm talks about our progress building an automated AI researcher and what’s ahead as AI can take on harder and harder problems. Jacob Effron (@jacobeffron) At @OpenAI, Chief Scientist @merettm helps lead the research roadmap to AGI including a research intern-level AI system by September 2026 and a fully automated AI researcher by March 2028. I sat down with Jakub to check on those timelines and ask him all of my top-of-mind AI questions including: ▪️ How OpenAI thinks about extending RL beyond code and math ▪️ The current state of alignment research as more powerful models loom ▪️ The future of continual learning ▪️ How startups should think about building their own models/harnesses And he also shared some great stories around OpenAI’s pioneering work on math. YouTube: piped.video/vK1qEF3a3WM Spotify: bit.ly/4sjUyrN Apple: bit.ly/41jAdrN 0:00 Intro 1:53 Research Intern Capability Timelines 4:59 Math Breakthroughs 7:59 RL Beyond Verifiable Tasks 12:32 RL vs In-Context 19:01 Allocating Compute Internally 28:18 AI for Science 31:40 Pattern Matching 33:23 Solving the Hardest Math Problems 37:40 Chain of Thought Monitoring 44:33 Generalization and Value Alignment in Models 47:57 Inside OpenAI 51:55 Quickfire — https://nitter.net/jacobeffron/status/2042234897134162077#m
→ View original post on X — @ceobillionaire, 2026-04-09 23:39 UTC
-
Discrete symbols and mathematics foundations of AI modeling
By
–
To make sense of the world is to model it in the simplest possible way. And simplicity requires discrete symbols. This is why we developed mathematics in the first place.