Jan Kulveit Pushes Back on 'Context Is the Right AI Identity' Claim
gleech · x · 2026-09-19
Responding to Jsevillamol's claim that the right unit of AI identity is the context rather than the model, Jan Kulveit argues 'right' is a confused adjective. In his own tests of reflective stability, context (a 'scaffolded system') is one relatively stable choice, useful for extracting work from AIs in multi-agent settings — but that doesn't make it correct. He likens the claim to saying a human's true identity is their job and hierarchy role: sometimes sensible, but not generally true, and 'I'm my Gmail inbox' is not reflexively stable when model gaps are substantial.
More from AGI Musings
- Philosopher Piccinini: phenomenal consciousness outstrips computation, so AI won't be conscious via compute alone — pwlot · 2026-09-19
- Philosopher Jeff Sebo rebuts Suleyman's 'model welfare' essay: 'No evidence' of AI consciousness is too strong — connoraxiotes · 2026-09-19
- Schmidhuber claims Recursive Self-Improvement algorithms date back to his 1987 work, slams 'AI stepfathers' — SchmidhuberAI · 2026-09-19
- UnpromptedAU security conf: AI makes reverse engineering trivial, killing security through obscurity — moyix · 2026-09-19
- AI will replace the person writing syntax, not the one who knows what to build — forevergeeks · 2026-09-19
- Oxford paper argues LLMs are mathematically incapable of true invention — blaizedsouza · 2026-09-19