Lean into laziness
Transcript of a talk I gave on using LLMs| Speculative Decoding
It's hard to write a guide to being good at ownership.| Speculative Decoding
Once again Google has made something awesome and completely failed to market it| Speculative Decoding
By just getting Claude to write the prompts| Speculative Decoding
on the gap between available-in-theory and used-in-practice| Speculative Decoding
Crosspost of an old piece, which still holds up| Speculative Decoding
They describe the situation in detail. They share the nuances they're trying to convey. They don’t know what to say. Often, the first thing I’ll respond with is “what if you just said that?”| Speculative Decoding
Feedback is a powerful tool for shaping behaviour, but people often fail to recognize this when it comes to positive feedback. It’s easy to understand that explaining mistakes and suggesting improvements can help someone progress, but fail to notice the potential of positive feedback to do the same.| Speculative Decoding
Until recently, the best way to make talking to ChatGPT better was by using custom instructions.| Speculative Decoding
In which ChatGPT spectacularly fails to take a hint| Speculative Decoding
Alex Hey, I want to write a blog post about how I use language models from day to day, and I thought that talking to you might be a good starting point for that post.| lawsen.substack.com
