What does 20 times more powerful look like? At some point will LLM AI be constrained by datasets based on human knowledge? If it still has the hallucination problem, and no ability to reason, does it just mean its faster?

  • Anony Moose
    link
    fedilink
    English
    arrow-up
    3
    ·
    1 year ago

    Is there a paywall free version of the entire article somewhere?

  • Pons_Aelius@kbin.social
    link
    fedilink
    arrow-up
    3
    ·
    1 year ago

    no ability to reason, does it just mean its faster?

    Faster. only faster. LLMs as far as I understand have as much self-awareness and ability to reason as a Ti-83.

    We are no closer to an AGI than we were when AI research started in the '50s.

    From the link:

    “The timeline for AGI development remains a subject of ongoing debate among researchers and experts. Some argue that it may be possible in years or decades, others maintain it might take a century or longer, and a minority believe it may never be achieved.”

    The central problem is this:

    How can we create or program a machine to have self-awareness when will still do not understand how humans gain and develop?