|
Tuesday 23rd April’s issue is presented by Index |
 |
Index Conference: Scaling search and AI apps at Netflix, Uber, DoorDash, Linkedin
Index is a free community conference with talks on search, streaming and ML infra.
Hear about homepage personalization at Netflix, the feed infrastructure at LinkedIn and the FAISS library at Meta.
Register for the virtual and in-person event at the Computer History Museum in CA on May 16th. |
|
|
The Impact Of AI Tooling On Engineering At ANZ Bank — Abi Noda
tl;dr: “To evaluate whether Copilot should be used org-wide, the authors of this paper conducted an experiment for six weeks, and compared the tool’s impact on a test group versus a control group. They based their evaluation of the tool’s impact using measures for productivity, quality, and security.“
Leadership Management |
|
Measuring Personal Growth — Chip Huyen
tl;dr: Chip wondered how to measure 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 with a lot of friends, I found three interesting metrics: rate of change, time to solve problems, and number of future options.
CareerAdvice |
|
10 Must-Reads For Engineering Leaders — Anton Zaides
tl;dr: “Although I insist you should fully read them, I summarized my main takeaway from each book in today’s article.” Anton discusses: (1) Turn the Ship Around: building a team that doesn’t depend on you. (2) No Rules Rules: removing all controls and bureaucracy. (3) Extreme Ownership: You are the organization. And more.
Management Books |
|
How Does ChatGPT Work? As Explained By The ChatGPT Team — Gergely Orosz
tl;dr: When you ask ChatGPT a question, several steps happen: (1) Input: We take your text from the text input. (2) Tokenization: We chunk it into tokens. A token roughly maps to a couple of unicode characters. You can think of it as a word. (3) Create embeddings: We turn each token into a vector of numbers. These are called embeddings. (4) Multiply embeddings by model weights: We then multiply these embeddings by hundreds of billions of model weights. (5) Sample a prediction.
AI LLM |
|
|
“If you can't explain it to a six year old, you don't understand it yourself.”
— Albert Einstein
|
|
|
Scheduling Internals — Tony Solomonik
tl;dr: “I remember when I first learned that you can write a server handling millions of clients running on just a single thread, my mind was simply blown away. I used Node.js while knowing it is single threaded, I used async / await in Python, and I used threads, but never asked myself "How is any of this possible?". This post is written to spread the genius of concurrency and hopefully getting you excited about it too.”
Concurrency |
|
Flow State: Why Fragmented Thinking Is Worse Than Any Interruption — Nick Moore
tl;dr: In this post, Nick shares what he learned after researching the concept of "flow state." He shares where the term came from originally, what the popular usage of it misses, and how we all can stay in flow state by avoiding context switching and fragmented thinking.
Promoted by StackBlitz CareerAdvice |
|
Untangling Spaghetti: Debugging Non-Terminating Object Programs — Kent Beck
tl;dr: “I had fun today debugging an infinite loop. Some of the techniques we used build on things I've talked about recently so I thought I'd reflect on the experience a bit. There's no happy ending to my story, though: the defect remained at the end of the session. Still, the techniques are worth thinking about because infinite loops in object programs can be difficult to debug. I wish you better luck with yours.”
Debugging |
|
Scaling To Count Billions
tl;dr: Canva pays creators based on billions of content usages each month. This usage data not only includes templates but also images, videos, and so on. Building and maintaining a service to track this data for payment is challenging and must be accurate, scalable and operable. This post introduces the various architectures the team experimented with and the lessons learned along the way.
Architecture |
|
I Accidentally Built A Meme Search Engine
tl;dr: “I built a meme search engine using siglip / CLIP and vector encoding images. It was fun and I learned a lot. I have been building a lot of applied AI tools for a while. One of the components that always seemed the most magical has always been vector embeddings. Word2Vec and the like have straight blown my mind. It is like magic.” Harper describes his journey and shares the results.
AI Search |
|
Index Conference
Join engineers from Netflix, Uber and DoorDash at the free community conference for search and AI applications on May 16th. Talks include: Vector search and the FAISS library by Matthijs Douze, Co-creator of FAISS Homepage personalization at Netflix by Shriya Arora, Engineering Manager
Register Here |
Most Popular From Last Issue |
|
|
Aider: AI pair programming.
Hatchet: Distributed, fault-tolerant task queue.
Puter: OS internet operating system.
Quill: Modern rich text editor.
Reader: Convert any URL to LLM-friendly input.
|
|
Click the below and shoot me an email!
1 = Didn't enjoy it all // 5 = Really enjoyed it
1 … 2 … 3 … 4 … 5 |
|
|
|