Elektrine lite

← Feed

@rayckeith@techhub.social

2026-09-27 21:47 UTC

"I read Gödel, Escher, Bach in high school, and it warped my brain. In that book and later in I Am a Strange Loop, Hofstadter argued that human consciousness arises from an intricate, self-referential feedback loop. A system of symbols acquires the ability to reflect upon itself, crossing hierarchical levels until an "I" emerges from the tangle. For decades, Hofstadter seemed to suggest that if we ever built an intelligence, it would arrive through that same kind of deep, recursive, analogical architecture. Then LLMs landed, and Hofstadter has been openly devastated. In interviews—most notably in The New York Times and in video discussions like his interview on the state of AI—he’s spoken about the profound existential vertigo of watching systems like GPT-4 "scaring the daylights out of me" by producing witty, analogical, seemingly profound text without having traversed the slow, decades-long path of embodied human living. It felt to him like an unsettling shortcut, threatening to reduce the sanctity of mind and decades of his life's work to feed-forward statistical tricks." https://blog.lmorchard.com/2026/09/27/pragmatic-anthropomorphism/

Replies (1)

  • @masek@infosec.exchange 2026-09-28 08:19

    @rayckeith@techhub.social My prediction is: At the end of the discussion about AI we will be deeply disappointed by our own intelligence 🙂. There is still seems to be an expectation to find the divine or magical spark that distinguishes us from everything else. I don't expect we will find such a thing. Some years ago, I read the book "How History Gets Things Wrong: The Neuroscience of Our Addiction to Stories" by Axel Rosenberg. In the book he writes on speculations how our intelligence began to form itself. The interesting hypothesis was that it began as mechanism to predict the behavior of others in order to coordinate actions. Which is, if one tends to harsh judgement, just a kind of token prediction.

    Open ##5176221