Google's new project IDX, Nvidia's bet on AI, and two papers on language models are discussed. The first paper from Google DeepMind explores how simple synthetic data can reduce sycophancy in large language models, while the second paper from Stanford University proposes a new algorithm called staged speculative decoding to speed up the inference of large language models in small-batch, on-device scenarios.
Contact: sergi@earkind.com
Timestamps:
00:34 Introduction
01:41 Google Unveils Project IDF
03:20 Nvidia CEO: "We bet the farm on AI and no one knew it"
04:53 Jason Wei Long Tweet on AI research (Researcher at OpenAI)
06:16 Fake sponsor
08:07 Simple synthetic data reduces sycophancy in large language models
09:34 Leveraging Few-Shot Data Augmentation and Waterfall Prompting for Response Generation
10:56 Accelerating LLM Inference with Staged Speculative Decoding
12:45 Outro
