Chip Huyen's Blog

As we���re still in the early days of building applications with foundation models, it���s normal to make mistakes. This is a quick note with examples of some of the most common pitfalls that I���ve seen, both from public case studies and from my personal experience.

Because these pitfalls are common, if you���ve worked on any AI product, you���ve probably seen them before.

Every time there���s a new technology, I can hear the collective ...

Intelligent agents are considered by many to be the ultimate goal of AI. The classic book by Stuart Russell and Peter Norvig, Artificial Intelligence: A Modern Approach (Prentice Hall, 1995), defines the field of AI research as ���the study and design of rational agents.���

The unprecedented capabilities of foundation models have opened the door to agentic applications that were previously unimaginable. These new capabilities make it finally possible to develop autonomous, intelligent agents to...

After studying how companies deploy generative AI applications, I noticed many similarities in their platforms. This post outlines the common components of a generative AI platform, what they do, and how they are implemented. I try my best to keep the architecture general, but certain applications might deviate. This is what the overall architecture looks like.

This is a pretty complex system. This post will start from the simplest architecture and progressively add more compo...

My founder friends constantly think about growth. They think about how to measure their business growth and how to get to the next order of magnitude scale. If they���re making $1M ARR today, they think about how to get to $10M ARR. If they have 1,000 users today, they think about how to get to 10,000 users.

This made me wonder if/how people are measuring personal growth. I don���t want to use metrics like net worth or the number of followers, because that���s not what I live for. After talking...

Four years ago, I did an analysis of the open source ML ecosystem. Since then, the landscape has changed, so I revisited the topic. This time, I focused exclusively on the stack around foundation models.

The full list of open source AI repos is hosted at llama-police. The list is updated every 6 hours.

Table of contents
Data
���. How to add missing repos
The New AI Stack
���. AI stack over time
������.. Applications
������.. AI engineering
������.. Model development
������.. Infrastructure
Op...

Table of contents

Ranking Models Using Human Preference
���. How Preferential Ranking Works
���. Correctness of Chatbot Arena Ranking
������.. Eval data
������.. Results
Predicting Human Preference For Each Prompt
���. Experiment setup
���. Experiment results
������.. Domain-specific and query-specific leaderboards
Conclusion

Human preference has emerged to be both the Northstar and a powerful tool for AI model development. Human preference guides post-training techniques including RLHF an...

ML models are probabilistic. Imagine that you want to know what���s the best cuisine in the world. If you ask someone this question twice, a minute apart, their answers both times should be the same. If you ask a model the same question twice, its answer can change. If the model thinks that Vietnamese cuisine has a 70% chance of being the best cuisine and Italian cuisine has a 30% chance, it���ll answer ���Vietnamese��� 70% of the time, and ���Italian��� 30%.

This probabilistic nature makes AI ...

For a long time, each ML model operated in one data mode ��� text (translation, language modeling), image (object detection, image classification), or audio (speech recognition).

However, natural intelligence is not limited to just a single modality. Humans can read and write text. We can see images and watch videos. We listen to music to relax and watch out for strange noises to detect danger. Being able to work with multimodal data is essential for us or any AI to operate in the real world.

Never before in my life had I seen so many smart people working on the same goal: making LLMs better. After talking to many people working in both industry and academia, I noticed the 10 major research directions that emerged. The first two directions, hallucinations and context learning, are probably the most talked about today. I���m the most excited about numbers 3 (multimodality), 5 (new architecture), and 6 (GPU alternatives).

Open challenges in LLM research

1. Reduce and measure halluc...

I had a lot of fun preparing the talk: ���Leadership needs us to do generative AI. What do we do?��� for Fully Connected. The idea for the talk came from many conversations I���ve had recently with friends who need to figure out their generative AI strategy, but aren���t sure what exactly to do.

This talk is a simple framework to explore what to do with generative AI. Many ideas are still being fleshed out. I hope to convert this into a proper post when I have more time. In the meantime, I���d ...