Marco Mascorro(@Mascobot) 's Twitter Profileg
Marco Mascorro

@Mascobot

Cofounder @Fellow_AI. AI Research Partner @a16z | AI & Robotics Engineer | prev @BMW research | @MIT 35 under 35 | Opinions my own.

ID:80466845

linkhttps://www.mascobot.com calendar_today07-10-2009 02:11:33

4,8K Tweets

9,7K Followers

1,6K Following

Yann LeCun(@ylecun) 's Twitter Profile Photo

An interview of me in Wired with the unequaled Steven Levy.
'How Not to Be Stupid About AI, With Yann LeCun'
It’ll take over the world. It won’t subjugate humans. For Meta’s chief AI scientist, both things are true.

Excerpts:

- Steven Levy: In a recent talk, you said, “Machine…

account_circle
martin_casado(@martin_casado) 's Twitter Profile Photo

This is a wild use of AI in a video game. Basically a reverse turing test where the LLMs try and guess if you're human.

Check it out. One of the most innovative uses of LLMs in games I've seen.

account_circle
Andrew Ng(@AndrewYNg) 's Twitter Profile Photo

In AI, the ratio of attention on hypothetical, future, forms of harm to actual, current, realized forms of harm seems out of whack.

Many of the hypothetical forms of harm, like AI 'taking over', are based on highly questionable hypotheses about what technology that does not…

account_circle
Sharif Shameem(@sharifshameem) 's Twitter Profile Photo

The Mixtral MoE model genuinely feels like an inflection point — a true GPT-3.5 level model that can run at 30 tokens/sec on an M1.

Imagine all the products now possible when inference is 100% free and your data stays on your device!

account_circle
Matt Bornstein(@BornsteinMatt) 's Twitter Profile Photo

The big idea in the Mistral AI models is high accuracy (currently gpt 3.5 level) with very efficient inference & full open source access.

We're leading their A round, and backing this incredible team, to help them achieve that goal at scale

account_circle
Georgi Gerganov(@ggerganov) 's Twitter Profile Photo

Adding support for the new Mixtral models

Runs on CPU, CUDA and Metal with quantization support and partial GPU offloading.

Very interesting architecture to play with!

github.com/ggerganov/llam…

account_circle
Guillaume Lample(@GuillaumeLample) 's Twitter Profile Photo

Very excited to release our second model, Mixtral 8x7B, an open weight mixture of experts model.
Mixtral matches or outperforms Llama 2 70B and GPT3.5 on most benchmarks, and has the inference speed of a 12B dense model. It supports a context length of 32k tokens. (1/n)

Very excited to release our second model, Mixtral 8x7B, an open weight mixture of experts model. Mixtral matches or outperforms Llama 2 70B and GPT3.5 on most benchmarks, and has the inference speed of a 12B dense model. It supports a context length of 32k tokens. (1/n)
account_circle
Arthur Mensch(@arthurmensch) 's Twitter Profile Photo

Announcing Mixtral 8x7B mistral.ai/news/mixtral-o… and our early developer platform mistral.ai/news/la-platef…. Very proud of the team!

account_circle
Marco Mascorro(@Mascobot) 's Twitter Profile Photo

Mistral just released the details about their Mixtral MoE model, and matches or outperforms llama-2 70B and GPT3.5 on most benchmarks:

mistral.ai/news/mixtral-o…

Mistral just released the details about their Mixtral MoE model, and matches or outperforms llama-2 70B and GPT3.5 on most benchmarks: mistral.ai/news/mixtral-o…
account_circle