Gary Marcus

Rank 22 of 47
|
Score 74
4/17/2025, 2:52:24 AM
X
In reply to:
Liron Shapira
@liron
·
2d
@GaryMarcus @rao2z Ok so we’re allowed to have a scenario where AI can derive arbitrarily large subgoals and implement them, potentially upending civilization as we know it, but it doesn’t count as “agency” because 5 years ago a human said “beat Elon to Mars”?
Gary Marcus
@GaryMarcus
·
2d
@liron @rao2z Not in the sense of having its own goals (though it would be very impressive in its own right).
Liron Shapira
@liron
·
2d
@GaryMarcus @rao2z How big does the externally-supplied goal have to be before the sub goals are sufficiently independent/agentic for you? E.g. if the goal is beat Elon to Mars and the AI doesn’t require any more instructions for 5 years and succeeds, that’s still not agentic?
Gary Marcus
@GaryMarcus
·
2d
@liron @rao2z Those are orthogonal dimensions. Following instructions vs coming up with its own strategic goals independent of instructions.
Liron Shapira
@liron
·
2d
@GaryMarcus @rao2z If it’s executing a sufficiently big plan competently, without human intervention, is there a particular point when it’s acting independently? Seems like a continuum.
Gary Marcus
@GaryMarcus
·
2d
@liron That sounds like planning, which as @rao2z will tell you, is still inadequate, rather than a capacity to act independently, which IMHO could be more concerning.

The statement engages in a discussion about AI capabilities and the potential risks associated with AI achieving subgoals independently, which is a public issue concerning technology and its impact on society.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not cause harm and suggests maintaining good mental hygiene, which aligns with the principle of doing no harm. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement promotes understanding by acknowledging risks and suggesting a cautious approach, which aligns with promoting empathy and compassion. [+1]
  3. Principle 4:
    I will engage in constructive criticism and dialogue with those in disagreement and will not engage in personal attacks or ad hominem arguments.
    The statement engages in constructive dialogue by addressing potential risks without resorting to personal attacks, aligning with the principle of constructive criticism. [+1]