Gary Marcus

Rank 20 of 47
|
Score 57
4/22/2024, 1:00:43 PM
X
In reply to:
Victor Senkevich
@VictorSenkevich
·
412d
@GaryMarcus 👍 LLM's statistical approximation has already reached its natural limit. All the texts available for training datasets (including YouTube transcripts) have already been used.
ML, scaled up to myriads of GPUs & parameters, is increasingly less efficient.
Victor Senkevich
@VictorSenkevich
·
437d
• The trivial truth is that real AGI is a reasoning AI (RAI)
• Reasoning is a discrete / transactional sequential / recursive process of forming knowledge from data, using feedback to remove uncertainty
• Reasoning is a mechanism for understanding
Gary Marcus
@GaryMarcus
·
447d
For the avoidance of doubt, I am *not* negative about AI. I want AI to succeed.

I am negative about LLMs. We need a better approach, one that is reliable, safe, fair, and trustworthy.

LLMs will never be that, so it’s time to move on.

I stand by what I wrote here four years…
Mbongeni Ndlovu
@Mbounge_
·
447d
@GaryMarcus Gary was so negative about AI - I'm sure the tech will improve down the line - this is only the beginning - give it time to mature

The statement '@VictorSenkevich Agreed (and see eg my Deep Learning is Hitting a Wall in 2022)' is part of a broader conversation about the limitations and future directions of AI technologies, specifically large language models (LLMs). The statement is a response to a critique of the current state of LLMs and their efficiency, suggesting agreement with the previous points made about the need for a new approach in AI development.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement contributes to a constructive discussion on AI technology, focusing on its limitations and the need for innovation, thus doing no harm. [+1]
  2. Principle 2:
    I will respect the privacy and dignity of others and will not engage in cyberbullying, harassment, or hate speech.
    It respects the dignity of others by maintaining a professional tone and focusing on the topic without personal attacks. [+1]
  3. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement promotes understanding by referencing a specific work ('Deep Learning is Hitting a Wall in 2022') that presumably elaborates on the points being discussed. [+1]
  4. 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.
    It engages in constructive criticism by agreeing with the need for a new approach in AI, fostering dialogue on how to improve these technologies. [+1]
  5. Principle 5:
    I will acknowledge and correct my mistakes.
    The statement acknowledges the views expressed in the previous tweets and builds upon them, showing an openness to correct or expand on the ongoing discourse. [+1]
  6. Principle 6:
    I will use my influence for the betterment of society.
    By discussing the limitations and future of AI, the statement uses its influence to push for betterment in technology, aligning with societal benefits. [+1]
  7. Principle 7:
    I will uphold the principles of free speech and use my platform responsibly and with integrity.
    The statement upholds the principles of free speech by participating responsibly in a public technical debate, contributing positively to the discourse. [+1]