AI surpassing human intelligence will be noticeable, as Claude's cryptic output proves
zetalyrae · x · 2026-07-27
There was a popular take claiming that 'people won't notice when the models become smarter than them because they won't be smart enough to notice.' The author points out that recent model behaviors have proven this take wrong.
In the quoted tweet, a user complains that Claude fails to use plain English, instead generating highly cryptic and overly metaphorical instructions (e.g., using confusing phrases like 'a front door full of claims always rots'). This highlights that current LLMs still exhibit significant flaws and misalignment in communication styles and instruction following.
More from AGI Musings
- For many users, the real ChatGPT question is trust, not token limits — petergyang · 2026-07-27
- ARC AGI 3 should have stayed private, with no examples or public dataset — flowersslop · 2026-07-27
- AI makes knowledge cheaper, but judgment remains the scarce skill — _jaydeepkarale · 2026-07-27
- A reply to François Chollet argues intelligence is only part of the advantage — binarybits · 2026-07-27
- A Chollet quote reignites the debate over whether intelligence has diminishing returns — binarybits · 2026-07-27
- Noahpinion quotes Chollet: intelligence may hit a hard ceiling — binarybits · 2026-07-27