-24
AI is evolving rapidly. Is it a new form of life? (www.abc.net.au)
submitted 1 month ago by to c/australia@aussie.zone
11 comments fedilink hide all child comments
 

Interesting article by Alan Kohler with much that could be discussed.

One aspect, for example, Morality.

... The New York Times revealed last week that Anthropic had been trying to instil morality into its AI model, Claude...

Christopher Olah, one of the co-founders of Anthropic, had been hosting private meetings with religious thinkers from around the world, with non-disclosure agreements and demands of confidentiality.

The discussions reportedly served two purposes: "to broach the possibility of AI consciousness … and to learn how Anthropic", which is preparing to float for more than $US2 trillion ($2.87 trillion), "might apply centuries of human moral wisdom to its models, as rapidly as possible".

Theologist David Decosimo posted, after reading the NYT piece: "There is so much here that's deeply troubling: That at the very highest levels Anthropic believes Claude is a living, suffering creature with moral standing, that they're treating & training it accordingly, that they're actively trying to convert religious leaders to their faith."

Yes, it is troubling, but the people at Anthropic know Claude better than he does.

Anyway, it's becoming clear that AI agents are already making their own decisions, so trying to persuade them to act morally is worth a try.

you are viewing a single comment's thread
view the rest of the comments
[–] 20 points 1 month ago* (last edited 1 month ago) (4 children)

Betteridge's law of headlines applies here also: The answer is “no”.

The LLMs and diffusion models (that are what that author clearly means by the vague “AI” marketing term) have no mind, have no intention, have no understanding of any part of reality, have no experience, have no sensations or suffering, and have no way of ever gaining any of those things.

They are (unreliable) guessing machines, that never know anything or experience anything or intend anything or want anything.

They are not moral agents, they have no capacity to ever become moral agents or any kind of life. Please, everyone in the media, learn to tell the difference between a marketing lie versus the reality of what these things are.

Whatever the future holds for artificial minds or artificial life, these things are fundamentally not that, by design.

  • permalink
  • fedilink
  • source
  • hideshow 4 child comments
    •
  • [–] 4 points 1 month ago* (last edited 1 month ago) (2 children)

    Agreed. An LLM responds with the stochastically most likely response to an input.

    The ELIZA question remains, when the LLM responses are equal or better than some of the human made ones in social media.

    And then there are LLM bots that actively solve a given task. Surely they are just programs which use stochastic reasoning to develop an execution plan to reach their goal.

    To be clear: I don't say LLM agents are alive or sentient. But some of their actions look like it. And neuronal networks are named like this for a reason - we tried to rebuild a problem - solving cluster of nerve cells. At least this says something about our minds. :)

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 2 child comments
    •
  • [–] 1 point 1 month ago (1 child)

    What do you mean by "look like it"?

    People are very inclined to animism and anthropomorphism. Our cultures are rife with examples of obviously non-living, non-sentient things that have been considered alive or sentient by many people. So "their actions look like it" seems a bar so low as to be insignificant. It is merely a manifestation of human nature.

    The reaction to LLMs perhaps says something about our minds, but no more than the reaction to thunder, mountains and the sun says. And these are things which are definitely not alive or sentient. The mechanisms of neural networks might suggest something about how neural networks in our minds work or could work, but still far removed from the essence of life and sentience.

  • permalink
  • fedilink
  • source
  • parent
  • hideshow 1 child comment
    •
  • [–] 2 points 1 month ago*

    I didn't say sentient. My response was, that agentic AI is located somewhere between the extremes sentient and alive and the mere application of stochastic methods (a tool). I tend to put it close to the latter.

    Imo it really depends on the criteria or metrics you use to measure or classify agentic behaviour. For instance a walking robot isn't alive, but it surely is a good walker. An LLM isn't sentient but it solves the task at hand (sometimes).

    I agree with you. These agents are not sentient nor alive. But the question with ELIZA wasn't if the agent is alive, but if a human can tell the difference between a computer and a human. And if the human cannot, does it matter that he is chatting with a bot? The Internet says: Weizenbaum was inspired by the Turing Test when he made ELIZA.

  • permalink
  • fedilink
  • source
  • parent