Gary Marcus

Rank 19 of 47
|
Score 64
6/19/2024, 1:29:57 AM
X
In reply to:
Yaron
@yaronhadad
·
381d
@GaryMarcus How far do you think we are from the asymptote then?
Gary Marcus
@GaryMarcus
·
381d
My considered take—after five years’s observation—is exactly the same: 𝘓𝘓𝘔𝘴 𝘮𝘢𝘺 𝘣𝘦 𝘶𝘴𝘦𝘧𝘶𝘭 𝘤𝘰𝘮𝘱𝘰𝘯𝘦𝘯𝘵𝘴 𝘣𝘶𝘵 𝘰𝘯 𝘵𝘩𝘦𝘪𝘳 𝘰𝘸𝘯 𝘵𝘩𝘦𝘺 𝘸𝘪𝘭𝘭 𝘯𝘦𝘷𝘦𝘳 𝘣𝘦 𝘢𝘣𝘭𝘦 𝘵𝘰 𝘢𝘤𝘩𝘪𝘦𝘷𝘦 𝘮𝘶𝘤𝘩 𝘰𝘧 𝘸𝘩𝘢𝘵 𝘩𝘢𝘴 𝘣𝘦𝘦𝘯 𝘦𝘹𝘱𝘦𝘤𝘵𝘦𝘥 𝘰𝘧…
Sandor 🦔
@AlexAegis
·
381d
@GaryMarcus hot take: An LLM is a very low level component and a lot of work needed on top of one to create something useful. So far we just naively threw everything into it and expected good answers.

The statement 'I think GPT-5 will disappoint and that we won’t get to GPT-6 anytime soon/without some sort of breakthrough' is part of a broader discussion on the limitations and future of large language models (LLMs). The tone is speculative and critical, expressing skepticism about the immediate advancements in LLM technology. The intent appears to be to contribute to the ongoing debate about the capabilities and future development of AI models.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not cause harm and is a part of a constructive debate about AI technology. It is a critical opinion but does not engage in harmful rhetoric. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement promotes understanding by contributing to the discussion on the limitations and potential of LLMs. It encourages a realistic view of technological advancements. [+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 criticism without personal attacks. It focuses on the technology rather than individuals. [+1]