Issue 51 7th May 2024

Highlights

7 takeaways from a year of building generative AI responsibly and at scale - Source

Microsoft released its first annual Responsible AI Transparency Report to help people understand how they approach responsible AI (RAI).

Stack Overflow signs deal with OpenAI to supply data to its models

OpenAI is collaborating with Stack Overflow, the Q&A forum for software developers, to improve its generative AI models’ performance on programming-related tasks. As a result of the partnership, announced Monday, OpenAI’s models, including models served through its ChatGPT chatbot platform, should get better over time at answering programming-related questions, the two companies say.

ChatGPT’s chatbot rival Claude to be introduced on iPhone

Challenger to market leader OpenAI says it wants to ‘meet users where they are’ and become part of users’ everyday life OpenAI’s ChatGPT is facing serious competition, as the company’s rival Anthropic brings its Claude chatbot to iPhones. Anthropic, led by a group of former OpenAI staff who quit over differences with chief executive Sam Altman, have a product that already beats ChatGPT on some measures of intelligence, and now wants to win over everyday users.

We’re bringing the Financial Times’ world-class journalism to ChatGPT

We will also collaborate on new AI experiences for FT readers.

Video

AI Show: On Demand | Announcing Phi-3 AI Models

Watch now to learn about Phi-3, the most capable and cost-effective small language models (SLMs) available, outperforming models of the same size and next size up across a variety of language, reasoning, coding, and math benchmarks.

Articles

Turn Llama 3 into an Embedding Model with LLM2Vec

RAG with Llama 3 for the generation and the retrieval Continue reading on Towards Data Science »

A 10-Minute Template to Build an AI Assistant on HuggingFace

AI agents are the next big thing in the AI scene. Imagine fully self-driving cars, bots doing your taxes, and AI tutors teaching your kids. “Hey Jarvis! Walk the dog please, book a trip for next weekend, and while you’re at it, go double my income.”

Universal signs TikTok deal allowing artists back on platform

Music group pulled content amid row over issues such as deepfake AI-generated music and artists’ compensation TikTok and Universal Music Group have reached a deal that will allow songs and artists from its labels including Olivia Rodrigo and Drake to return to the video-sharing app. The world’s largest music company began pulling content from TikTok in February after falling out with it over issues including artist compensation and the use of artificial intelligence-generated music on the platform.

A Hitchhiker’s Guide to Speculative Decoding

Speculative decoding is an optimization technique for inference that makes educated guesses about future tokens while generating the current token, all within a single forward pass. It incorporates a verification mechanism to ensure the correctness of these speculated tokens, thereby guaranteeing that the overall output of speculative decoding is identical to that of vanilla decoding. Optimizing the cost of inference of large language models (LLMs) is arguably one of the most critical factors in reducing the cost of generative AI and increasing its adoption. Towards this goal, various inference optimization techniques are available, including custom kernels, dynamic batching of input requests, and quantization of large models.

Microsoft at ASPLOS 2024: Advancing hardware and software for high-scale, secure, and efficient modern applications

From AI and deep learning to innovations in infrastructure, researchers from Microsoft are bridging the gap between architecture, programming languages, and operating systems to advance the state of the art at ASPLOS 2024.

I'm writing a new vector search SQLite Extension

Alex Garcia is working on sqlite-vec, a spiritual successor to his sqlite-vss project. The new SQLite C extension will have zero other dependencies (sqlite-vss used some tricky C++ libraries) and will work using virtual tables, storing chunks of vectors in shadow tables to avoid needing to load everything into memory at once.

Why RAG won’t solve generative AI’s hallucination problem

RAG is being pitched as a solution of sorts to generative AI hallucinations. But there's limits to what the technique can do.

Upcoming events

AI Community Day - 14 May 2024

Welcome to AI Community Day, where all the AI developers come together to geek out and share what they know about artificial intelligence. It's a chance for us to connect, learn from each other, and get inspired by the latest and greatest in AI. Whether you're a pro at AI or just getting started, this event is all about building our community and shaping the future of AI development.