AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
2min snip
Pausing to think about scikit-learn & OpenAI o1
Practical AI: Machine Learning, Data Science, LLM
Reflect on Advances: It's Not Thinking, Just Generating
Recent advancements in AI, specifically in reinforcement learning from human feedback (RLHF), show a shift in how models are tuned. The focus has moved to preference tuning within the chain of thought generation rather than fundamental changes in model architecture or size. Although there have been notable innovations in mixtures of models and training techniques in the past, the current updates merely involve adjustments in the prompt sets used in the RLHF process. This highlights a reduction in the complexity of change, suggesting that the user interface is applying these updates in a novel way. Despite claims of the model 'thinking' during text generation, it's important to recognize that the model is not capable of thought; it primarily produces text outputs based on learned patterns.
00:00
Transcript
Episode notes
Get the Snipd
podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any
moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share
& Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode