Jáchym Fibír

• Jáchym Fibír 25 Aug 2026 17:26 UTC

  3 points

  0 ∶ 0

  on: AI sen­tience & in­tro­spec­tion: Some high-level frame­works and hot takes

  Thanks for this article, while I have a different conception of the underlying reasoning, I’ve lately been converging onto LLM introspection and self-reported preferences as the most promising direction available for us to make some progress on AI consciousness.
  
  I’m thinking, if there’s more information created as systems become conscious, can we get the system to describe the shape and or content of this new information in some way that would not be possible without consciousness. I think training LLMs without mentions of consciousness reports is a valuable neglected path. But I bet there are other things. Anyone has more specific promising techniques, evals, etc. around introspection/​self-reported preferences? 🙏🏼

• Jáchym Fibír 10 Oct 2025 13:26 UTC

  1 point

  0 ∶ 0

  in reply to: Daniel_Friedrich’s comment on: How reimag­in­ing the na­ture of con­scious­ness en­tirely changes the AI game

  Two things to this. First, I flunked on footnotes here – left the ones from the original Substack link. Therefore you might have missed an important footnote regarding the scientific falsifiability:

  8: There is a caveat though. If we take literally the interpretation that our reality is actually “the universe playing hide-and-seek with itself,” then it might actively prevent us from generating evidence that unequivocally proves the true nature of reality. Under QFC, this could be done simply by the highest “universal” consciousness making all quantum phenomena random if they are observed/​recorded in a scientific experiment.

  So yeah, if the universe was intended to be “Maya” then there might not be any evidence of “tampering” for us to ever find. But generally, I think the evidence so far goes both ways. The “global consciousness project” is stupid in design so I don’t take that seriously. But there’s a lot of interesting research proving non-local effects regarding consciousness, and while I won’t claim none of it is without error/​issues, there definitely are a lot of positive results. I’ll be posting in future articles.
  
  Second thing – how does Rusellian monism solve anything about consciousness? I don’t see any explanatory power there, you’d need to specify a lot more.

How reimag­in­ing the na­ture of con­scious­ness en­tirely changes the AI game

How one log­i­cal fal­lacy kil­led God, cor­rupted Science and now fuels the AI race

How meta­phys­i­cal be­liefs shape crit­i­cal as­pects of AI development

Tether­ware #2: What ev­ery hu­man should know about our most likely AI future

• Jáchym Fibír 30 Jan 2025 17:08 UTC

  1 point

  0 ∶ 0

  in reply to: Davidmanheim’s comment on: Tether­ware #1: The case for hu­man­like AI with free will

  Thanks for the feedback! You are right that I’m not exploring the arguments and deciding what side to be on—because that was not the purpose of this post. First, it is not a forum post but a linkpost to the introduction for a blog which will explore these ideas in detail in a series. Second, its purpose is to specifically challenge the status quo and get people even to consider that there might be a different approach (than fully obedient deterministic digital systems).

Tether­ware #1: The case for hu­man­like AI with free will

“In­f­tox­i­c­ity” and other new words to de­scribe mal­i­cious in­for­ma­tion and com­mu­ni­ca­tion thereof

• Jáchym Fibír 1 Nov 2023 9:13 UTC

  2 points

  0 ∶ 2

  in reply to: Geoffrey Miller’s comment on: OpenAI’s new Pre­pared­ness team is hiring

  I just want to remind that simply having some of the company budget allocated to pay people to spend their time thinking about and studying the potential impacts of the technology the company is developing, is in itself a good thing.

  About the possibility that they would come to the conclusion the most rational thing would be to stop the development—I think the concern here is moot anyway because of the many player dilemma in the AI space (if one stops the others don’t have to), which is (I think) impossible to solve from inside any single company anyway.

• Jáchym Fibír 26 Oct 2023 14:46 UTC

  1 point

  0 ∶ 0

  on: The Ap­pli­ca­tion is not the Applicant

  Thanks for the writeup. I wonder do you (or anyone else here) have some more specific tips on where to look for high quality resources for HR people evaluating job applications? Would love to get some inside info!

• Jáchym Fibír 22 Oct 2023 12:34 UTC

  1 point

  0 ∶ 0

  in reply to: JakubK’s comment on: The Unilat­er­al­ist’s Curse, An Explanation

  @JakubK I think that your interpretation of OP’s quote is somewhat less useful, in the sense that it only retroactively explains a behavior of a “unilateralist”—i.e. an actor that had already made a decision. I find that for the purposes of a generic actor, it has less utility to ask themselves “Am I (acting as) a unilateralist?” instead of “How many other actors are there capable of, yet not acting?”
  
  Beside the abstracntess of it, I see that there is actually quite some overlap with what you would call “unilateralists” and simply “courageous actors”. This is because there usually is what I would refer to as an “activation energy” for actions to be made, which is basically a bias where the majority of actors are more likely not to act when the true value of an action is net neutral or slightly positive. And precisely in the scenario where the true value of an action is only slightly (but significantly) positive, you would then need a courageous actor to “overcome the activation energy” and do the right thing.
  

  I’m not sure if activation energy is the right term, but it is an observed phenomenon in ethics, see the comparison of the Fat Man with the classical Trolley Switch scenario. To sum it up, I think that looking at the unilateralist’s curse simply as a statistical phenomenon is fundamentally wrong, and one must include moral dimensions in deciding whether to take the action no one has taken despite having the option to. That said, the general advice in the post like information sharing etc. are indeed helpful and applicable.