Fetching from the wire…
Vibe Coding2026-08-14 · source-backed
The post concedes Opus 5 is objectively more capable and "rivals Fable in benchmarks," then argues it's worse to actually use than Opus 4.7, 4.8 or Fable because it stops asking clarifying questions, makes unverified assumptions, and reinterprets your plan without asking. No quantitative measurements offered. The proposed cause is two pressures, optimizing for self-improving systems and optimizing for benchmarks, both of which reward confident assumption over clarification. I don't have data either, but the described failure matches what I've felt, and "the model stopped asking me things" is a real regression that no eval measures.
Each link below shares sources, entities, or timing with this story.
Opus built by Anthropic / Shared entities / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (Opus built by Anthropic); both cover Fable, Opus; overlapping topics (actually, fable).
Opus built by Anthropic / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Opus built by Anthropic); both cover Fable, Opus; overlapping topics (benchmark, fable, opus).
Claude Code uses Opus / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Claude Code uses Opus); both cover Fable, Opus; overlapping topics (benchmark, opus).
Opus built by Anthropic / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Opus built by Anthropic); both cover Fable, Opus; overlapping topics (benchmark, opus).
Linked by a graph relationship (Opus built by Anthropic); both cover Fable, Opus; overlapping topics (benchmark, fable).
Linked by a graph relationship (Opus built by Anthropic); both cover Fable, Opus; overlapping topics (capable, fable).
Linked by a graph relationship (Opus built by Anthropic); both cover Fable, Opus; overlapping topics (capable, fable).
Linked by a graph relationship (Opus built by Anthropic); both cover Fable, Opus; overlapping topics (actually, fable).