alt Hacker NewsI wonder when we’ll reach diminishing returns on new AI models. Haven’t tried Mythos or Fable yet but it seems like Anthropic is already priming itself for this by calling for a “slowdown” of AI development.
Replies
alephnerd • today at 6:10 PM
At least in cybersecurity, we already have seen diminishing returns with newer models.
At this point the harness/applayer matters more, as different models perform better or worse on exploit classes depending on the prompt, tuning, and various other parameters.
Of course, by the time HN hyperfixates on a topic, it's already been executed on and HN is too late.
➕ show 1 reply
I think it's less of a technology problem and more of a money and investment problem. They called for the "slowdown" right when they filed for IPO. This allows them to both triage the money bleed while under the microscope as well as make their competitors look unethical and dangerous if they don't heed the warning to slow down before we lose control of A.I.