DevAI Research12 min reading time

Import AI 464: Fables writes GPU kernels; AI automation; and analog computation

Import AI
Read full post
Fable developed a highly efficient GPU kernel achieving an 18.71X speedup on an RTX PRO 6000 Blackwell, outperforming other AI-generated kernels. This advancement signals progress in AI's ability to automate core research tasks like kernel design, which is crucial for recursive self-improvement. Additionally, AI systems have improved their success rate on the Remote Labor Index from 2.5% to 16.1%, indicating enhanced capability in automating online freelance work.

More in Dev

Introducing the Agents API

Covered by 4 sources
Dev6 min read

Building The New Digital Experience With AI Starts With Engineering Quality Into It

Forbes
Dev12 min read

5 Python Techniques for Efficient Resource Orchestration

KDnuggets