Alex O'Connor: Transformers, Generative AI & The Deep Learning Revolution #71 

Alex O’Connor is a researcher, computer scientist, and machine learning manager.
Enjoy this conversation on generative AI - language and image models, prompt engineering, the latent space, fine-tuning, tokenization, textual inversion, adversarial attacks, GPT, Stable Diffusion, BERT, and more.
📚 Episode Notes: gettingsimple.com/alex-oconnor
🎙 Apple Podcasts: podcasts.apple.com/us/podcast...
🎙 Spotify: open.spotify.com/episode/0RSi...
CONNECT WITH ALEX
🐘 Mastodon mastodon.social/@uberalex
CONNECT WITH NONO
🎥 All Episodes: • 🎙 Getting Simple Podcast
🍃 Getting Simple: gettingsimple.com
🎙 Podcast: gettingsimple.com/podcast
🗣 Submit a Question: gettingsimple.com/ask
💬 Discord: / discord
🧠 Machine Learning: • 🧠 Machine Intelligence
👨🏻‍🎨 Sketches: sketch.nono.ma
✍🏻 Blog: nono.ma
🐦 Twitter: / nonoesp
📸 Instagram: / nonoesp
00:00 Introduction
00:40 Machine learning
02:36 Spam and scams
15:57 Adversarial attacks
20:50 Deep learning revolution
23:06 Transformers
31:23 Language models
37:09 Zero-shot learning
42:16 Prompt engineering
43:45 Training costs and hardware
47:56 Open contributions
51:26 BERT and Stable Diffusion
54:42 Tokenization
59:36 Latent space
01:05:33 Ethics
01:10:39 Fine-tuning and pretrained models
01:18:43 Textual inversion
01:22:46 Dimensionality reduction
01:25:21 Mission
01:27:34 Advice for beginners
01:30:15 Books and papers
01:34:17 The lab notebook
01:44:57 Thanks
Thanks to Andrea Villalón Paredes for editing this interview.
Sleep and A Loop to Kill For songs by Steve Combs under CC BY 4.0.
#ArtificialIntelligence #MachineLearning #GenerativeAI

Наука