Tutorials, news reactions, tool notes, personal miscellany, and observations outside the main topics.
Some of the LLM apps we've been experimenting with have been extremely slow, so we asked ourselves: what do GPT APIs' response times depend on? It turns out that response time mostly depends on the number of output tokens generated by the model. Why? Because LLM latency is linear in output token count. But the theory isn&
Context: first thoughts on Pause Giant AI experiments. I will refine my thinking over time. * I had not thought about AI safety much since ~2017, after thinking a lot about it in 2014-2017. In 2017, I defended my MSc thesis on an AI-safety-inspired topic (though very narrow and technical in nature), but decided to take some distance
There is a very simple, standardized way of solving the problem of too small GPT context windows. This is what to do when the context window gets full: Indexing: 1. Chunk up your context (book text, documents, messages, whatever). 2. Put each chunk through the text-embedding-ada-002 embedding model, and store the resulting vectors -- each consisting of 1536
GPT-4 came out yesterday and overshadowed announcements, each of which would have been bombshell news otherwise: * Anthropic AI announcing their ChatGPT-like API -- likely the strongest competitor to OpenAI today (only waitlist) * Google announcing PaLM API (currently it's just marketing -- no public access yet) * Adept AI announcing their $350M round (they build foundation models to interact with
"Alpaca is just $100 and competitive with InstructGPT" -- takes like this are going around Twitter, adding to the (generally justified!) hype around AI models. It is indeed a very encouraging result. Specifically, that it took so little compute to train something that achieves competitive results on benchmarks and is a relatively small model at 7B parameters. I haven&
I've been experimenting with embedding ideas into GPT-space, and using the resulting vectors to visualize. For example, you could plot different activities based on two axes: how safe they are, and how adrenaline-inducing they are. How does it work? The intuition is the following. In words, you can define an axis, e.g. "unsafe activity&
This is a talk I gave in Lisbon in October 2022, at IPFS camp. What makes it somewhat special: I generated all the illustrations for this deck using DALL-E as an experiment. It actually is a decent way of making pretty slides; I think with a better UI this would be a major improvement in my slide-making workflow.
This is a short NFTPort workshop I recorded for the ETHGlobal NY hackathon. I was actually in New York around that time but sadly had to fly out before the in-person event itself took place. Recommended over the last one because this is pre-recorded and heavily edited, so much more information-dense.
DALL-E 2 and Imagen are the hot new models everyone is sharing: the eye candy is undeniable. But since the original release of the GPT-3 large language model, OpenAI has quietly released the GPT API and playground for anyone to try. Here's a collection of my experiments with the most recent version of the model, text-
What did Estonians start paying more for in April 2022?
