Tooling

Octopus v4 is a new language model that uses functional tokens and a graph structure to integrate multiple open-source models, each specialized for different tasks.

via Hugging Face.

Nature of Code likely explores simulating natural phenomena, like physics or ecosystems, using computer programming.

via Nature of Code

Dave Kellogg shares slides from an event he hosted at Balderton Capital in London to discuss the impact of AI on go-to-market functions. via Kellblog.

Bruno Aziza and Matt Turck discuss the shifting data landscape in machine learning and AI, including cost considerations and the differences between SLMs and LLMs. via VentureBeat.

Simon Willison's latest blog post covers updates for working with Llama 3, AI tools for data journalism, and new projects like llm-evals and datasette-secrets. via Simon Willison's Weblog

Harvard is hosting a database with interesting biological measurements ranging from the duration of an eye blink to the mass of a fruit fly. via kottke.org

AutoCrawler is a new two-stage framework that uses the hierarchical structure of HTML to improve web automation and crawler generation for information-rich web pages.

via Hugging Face.

Researchers are proposing “Dynamic Typography,” which uses neural networks to create animated and deformed letters based on meaning for expressive, readable text. via Hugging Face.

Wired's Steven Levy details how eight Google employees invented the "transformer" concept, leading to a recent tech breakthrough. via kottke.org

Ken Kantzer shares practical lessons learned from using OpenAI's models, including a simple token estimation method and challenges with structured data extraction.

via Simon Willison's Weblog

Molly White's newsletter post explores the potential benefits and significant drawbacks of AI tools, questioning if AI's benefits outweigh its ethical and societal costs. via @citneedednews