Gary Marcus

Rank 37 of 47

Score -37

@jasonprompts And yet barely beat gpt-4. certainly the advance here is not remotely like the advance from GPT-3 to GPT-4 (or 2 to 3).

7/23/2024, 4:33:26 PM

In reply to:

Jason Gu

@jasonprompts

105d

@GaryMarcus What wall? Meta literally just threw data and compute at this

Gary Marcus

@GaryMarcus

105d

LLaMa 3 450B is huge. Here’s why:
– Roughly as good as GPT-4, but open source.
- There goes OpenAI’s business model
- But it’s not GPT-5 level either, despite truly massive resources.
- Sure looks deep learning is hitting a wall, pending some kind of new breakthrough. Synthetic…

Marco Mascorro @ ICML 2024

@Mascobot

105d

LLaMa 3.1 benchmarks side by side.

This is truly a SOTA model.

Beats GPT4 almost on every single benchmark.

Continuously trained with a 128K context length.

Pre-trained on 15.6T tokens (405B).

The fine-tuning data includes publicly available instruction datasets, as well as…

The statement engages in a discussion about the performance and advancements of AI models, specifically comparing LLaMa 3 to GPT-4. The tone is critical and analytical, focusing on the perceived lack of significant advancement from GPT-3 to GPT-4. The intent appears to be to provide a critical perspective on the progress of AI models, particularly in response to claims about LLaMa 3's performance.

Principle 1:
I will strive to do no harm with my words and actions.
The statement does not appear to cause harm; it is a critical analysis of AI model advancements. [+1]
Principle 3:
I will use my words and actions to promote understanding, empathy, and compassion.
The statement promotes understanding by engaging in a technical discussion about AI model performance, though it could be more constructive by providing specific data or examples to support the critique. [+1]
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 by questioning the level of advancement in AI models, but it could avoid a dismissive tone to foster more productive dialogue.