AI and consciousness: from objective descriptions to 'level zero'
randomseed.ioAuthor here. This is an essay written in dialogue form about why "AI is/isn't conscious" debates often fail due to level confusion: we mix objective descriptions (models/behavior) with first-person investigation (what it's like).
I propose a stratified view of knowing. "Level zero" is the checkpoint where a thought is directly present in lived experience, independent of whether it was generated internally or externally.
Jaron Lanier - Is Consciousness an Illusion? https://www.youtube.com/watch?v=VSb_MCs9eqY
Yes – the "consciousness is an illusion" line often seems to fold back on itself: if it's an illusion, then there's still something it is like to experience that illusion, so we're back at square one.
Also interesting around 09:36. He hints that parts of the scientific community tend to avoid this territory. Intuitively you'd think exploring unknowns should be fun – so why the reluctance?
Speculation: B. Kastrup sometimes frames this as a first-person / third-person divide with historical roots: modern science won its freedom by focusing on third-person, publicly measurable claims, while bracketting first-person and "ultimate" questions.
Not a literal written deal, but a cultural boundary that still bubbles up and shapes what feels legitimate to study – i.e., culture quietly sets the constraints on what counts as rational inquiry.