But humans may already be changing jobs inside the process.

Anthropic just released a fascinating measurement from inside its own R&D operation:

Claude now “leads” 26% of its measured AI research and development work, completing most of those tasks end-to-end from a high-level human prompt while a person supervises.

More than 90% of the measured R&D work involves AI collaboration at some level.

And Anthropic says roughly 30,000 research and engineering AI agents are operating concurrently on its main internal platform.

No, this isn't autonomous recursive self-improvement. Anthropic explicitly says its measured R&D has not reached fully autonomous operation.

But something quieter may already be happening.

Humans aren't leaving the loop.

We're moving up the loop.

From doing the work ? directing the work ? supervising AI doing the work ? using that work to help build the next generation of AI.

That transition might matter more than waiting for some magical morning when someone announces, “AGI has arrived.”

Erika takes a look.

#ArtificialIntelligence #AGI #AIResearch #AIAgents #FutureOfWork #Conjugo