I keep seeing this story but there are only a few data points that go along with it. The whole narrative of what AI will do to jobs seems to be agenda-based at this point. No one knows and it's way too early to tell. Everyone who is loud about it is trying to push their own agenda or obfuscate something else. Many companies are doing layoffs or becoming draconian with their employees in the name of AI but Ai is typically just an excuse to do what they wanted to do anyways.
Interesting premise but the statement "Pick a pool of people who seem able to judge promising papers" is pretty load-bearing to the overall "game" and probably just kind of hand waved. How do you find these people? It seems like you would need a similar game just to identify which people are good at predicting impactful papers to make the exercise worth it at all.
Fable so far has only been incrementally better than Opus 4.8 in my testing. The testing so far has been coding on a large monolith. In general Claude is starting to feel like it's optimize for greenfield vibe-coding projects. I vastly prefer Codex for coding on large, established codebases.