Recursive self-improvement is not AGI hype. And it’s not just prompt tuning either. Karpathy’s autoresearch ran 700 experiments on a single GPU, improved its own training code, and kept what worked. This loop is already here. I break down how it works, where it fails, and how
AGENTS
-
Databricks Partners with OpenAI to Launch GPT-5.5 on Unity AI Gateway
By
–
Databricks is excited to partner with @OpenAI on GPT-5.5, their latest frontier model. GPT-5.5 will be available in Unity AI Gateway on launch. You can use it with coding tools such as Codex, or to power your enterprise agents.
— Databricks (@databricks) 24 avril 2026
GPT-5.5 is state-of-the-art on many benchmarks… pic.twitter.com/UMWUFJgwCVDatabricks is excited to partner with @OpenAI on GPT-5.5, their latest frontier model. GPT-5.5 will be available in Unity AI Gateway on launch. You can use it with coding tools such as Codex, or to power your enterprise agents. GPT-5.5 is state-of-the-art on many benchmarks
-
OAI researchers ask about beginner recursive self-improvements
By
–
OAI researchers: “So, what are some good recursive self-improvements for someone just getting into recursively self-improving?”
-
Managed Claude Agents with Memory Now Available
By
–
Claude Managed Agents now have Memory support! In case you missed it, Managed Agents are available on Anthropic Workbench. Conway will likely get it too.
-
Tool Use and Self-Check as Core AI Workflow Components
By
–
Excited to compare in my workflows. Tool use + self-check is basically the whole game now.
-
LangChain Startup Tier for Faster Agent Development
By
–
At a startup and want to iterate faster and win bigger with agents? Explore our Startups tiers! https://
langchain.com/startups -
AI Achieves Record 62/100 on Senior Engineer Benchmark
By
–
It scored a 62/100 on my Senior Engineer benchmark, a new record! Worked for almost a day:
-
AI Researchers and Rocket Scientists Managing Burn
By
–
ai researchers rocket scientists managing burn
-
DevTalks Ep 3: Autonomous Coding with Cline
By
–
We're a week away from DevTalks Ep 3 Join us for a chat about autonomous coding with @cline
. What changes when inference is faster, workflows are smoother, and you stay in control. April 30 | 11am PT | 2pm ET -
Codex Babysitter Keeps Opus Thread Running Successfully
By
–
I have a codex babysitter that keeps the Opus thread running 🙂 Thank you!!