Gary Marcus

Rank 15 of 47
|
Score 102
1/20/2025, 8:07:11 AM
X
In reply to:
Arctic
@neco_arctic
·
19h
@dioteos @outlander9133 @DaveShapi @GaryMarcus I think that OpenAI has a very good track record and hindsight has revealed that all the "baseless hype" has been fully substantiated in 100% of cases.
51st Unusual Anton
@dioteos
·
22h
@outlander9133 @DaveShapi @GaryMarcus Although I also think OpenAI might be manipulating things a bit, maybe it’s needed at this stage. We see quite a bit of optimism from various companies - there’s no smoke without fire, right?

Still, I’m firmly against any dishonest tactics if the rumors turn out to be true. To…
Killian, private detective. 🍊/acc
@outlander9133
·
23h
@dioteos @DaveShapi @GaryMarcus my experiences with o1-Pro suggest that, while benchmarks may not have been as clean-room as claimed, o3-Pro performance will probably meet the hype

you have to see how capable these things are firsthand for problems in your own specialty

that being said, Altman is a weasel
51st Unusual Anton
@dioteos
·
1d
@DaveShapi @GaryMarcus Are you open to the possibility that math tests have been compromised? There's too much at stake
David Shapiro ⏩
@DaveShapi
·
1d
@GaryMarcus @GaryMarcus honest question, don't you think that LLMs crushing benchmarks like GPQA and FrontierMath are like... signs of something really huge?
Gary Marcus
@GaryMarcus
·
1d
In other news, Axios reports that Elon Musk will be briefing a closed-door cabinet committee on a reported next-level breakthrough in interstellar warp drive, which is now at the prototype level and will be commercially available for short galaxy hops “very soon”.

SpaceX…

The statement engages in public discourse by discussing the capabilities and limitations of generative video models in understanding physical principles, which is a relevant topic in the field of AI and technology. It references a benchmark developed to assess these models, contributing to the broader conversation about AI's role and effectiveness in understanding and replicating human-like cognition and perception.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement does not cause harm and is focused on sharing information about a scientific benchmark. [+1]
  2. Principle 3:
    I will use my words and actions to promote understanding, empathy, and compassion.
    The statement promotes understanding by providing a resource to assess AI models' physics understanding. [+2]
  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 contributing to a discussion about AI capabilities without personal attacks. [+1]
  4. Principle 6:
    I will use my influence for the betterment of society.
    The statement uses influence to inform others about a tool that can improve AI model assessment, contributing to societal knowledge. [+2]
  5. Principle 7:
    I will uphold the principles of free speech and use my platform responsibly and with integrity.
    The statement upholds free speech by responsibly sharing information about a scientific development. [+1]