Llama
includes Llama 2, Llama 3.1, Llama 3, Llama 70B
3 statements across 3 episodes · 1 bullish · 1 bearish · 3 people on the record · first statement Sep 14, 2023 by Illia Polosukhin · said 50 times in 31 episodes since 2023 · across every show →
Mentions by year, the whole family
brought up most by Elad Gil (18), Sarah Guo (5), Jared Quincy Davis (4), Andrej Karpathy (3), Kawal Gandhi (2), Dylan Patel (2), Amjad Masad (2), Thomas Dohmke (1)
tap a year for its mentions
2026 2 mentions in 2 episodes 1 per episode
2025 5 mentions in 4 episodes 1 per episode
2024 20 mentions in 14 episodes 1 per episode
- No Priors Ep. 77 | With Foundry CEO and Founder Jared Quincy Davis
- No Priors Ep. 80 | With Andrej Karpathy from OpenAI and Tesla
- No Priors Ep. 78 | With AWS CEO Matt Garman
- No Priors Ep. 67 | With Voyage AI Co-Founder and CEO
- No Priors Ep. 95 | Best of 2024
- No Priors Ep. 91 | With Cohere Co-Founder and CEO Aidan Gomez
- No Priors Ep. 89 | With NVIDIA CEO Jensen Huang
- No Priors Ep. 86 | With Sarah Guo & Elad Gil
- 6 more episodes that year, every mention in 2024 →
2023 23 mentions in 11 episodes 2 per episode
- No Priors Ep. 33 | With Replit's CEO & Co-Founder Amjad Masad
- No Priors Ep. 35 | With Sarah Guo and Elad Gil
- No Priors Ep. 40 | With Arthur Mensch, CEO Mistral AI
- No Priors Ep. 37 | With Kawal Gandhi
- No Priors Ep. 42 | With Sarah Guo and Elad Gil
- No Priors Ep. 32 | With NEAR’s Illia Polosukhin
- No Priors Ep. 26 | With Weights & Biases CEO Lukas Biewald
- No Priors Ep. 20 | With Sarah Guo and Elad Gil
- 3 more episodes that year, every mention in 2023 →
Everything said about Llama, oldest first
Sep 14, 2023 bullish
Polosukhin: Inference demands vastly more aggregate compute than AI model training
“I think an inference is really interesting because we do need so much more compute for inference than we need for training, right? Like it's a very interesting like economy of scale. You train once, like Lama trained once and then everybody runs it everywhere.”
Sep 5, 2024 neutral
Karpathy: RoPE is the only major transformer architecture change in five years
“The transformer hasn't changed that much. You know, we've added the rope positional and the rope relative positional encodings. That's like the major change. Everything else doesn't really matter too much. It's like plus three percent on a small few things. Bu…”
Nov 21, 2024 bearish
Fine-Tuning Open Source Models Lacks Levers of Full Vertical Training
“Taking those models and trying to fine tune them It's just, it's not as effective as building it yourself and you have much fewer levers to pull than if you actually have access to the data and you can change the data that goes into that process.”