Resources · Learn
Our clients are business execs who want to execute AI strategy, not just talk about it.
You don't need to be able to write code in order to make sensible decisions about where and how to use AI in your business. But you do need a mental model of how AI systems work, what they're good at, and where they fail.
Here are some resources that will help. Some are a few years old but they are still relevant today.
The Financial Times Visual Storytelling teamA visual guide to the transformer, something that's at the centre of generative AI. It's the T in ChatGPT.
open the explainer →
Grant SandersonA video explaining what happens inside a model. Grant Sanderson is a mathematician and educator, best known as the creator of the 3Blue1Brown YouTube channel.
watch on youtube →
Andrej KarpathyA three-hour video that explains how LLMs are made. Don't be put off by the length. It will give you a mental model that helps you think about what an LLM is and what it does. Try watching thirty minutes at a time. Karpathy is one of the most accomplished AI researchers in the world.
watch on youtube →Andrej KarpathyKarpathy reframes LLMs as a new kind of computer you program in plain English, and makes the case for building products that keep a human firmly in the loop.
watch on youtube →
Satya Nadella, CEO of MicrosoftAn essay arguing that the long-term value of AI will not come simply from having access to the best AI model. It will come from building an ecosystem where AI and human expertise continuously learn from each other. Satya Nadella is the CEO of Microsoft.
read the essay →
Prof. Ethan Mollick, The Wharton SchoolA newsletter that translates what AI can and can't do into business practice. Mollick is a Wharton professor and consultant.
subscribe →
Simon Willison, independent developerA hype-free running commentary on what's happening in the AI world. His end-of-year reviews are great.
read the blog →
Benedict EvansA weekly newsletter summarising what's going on in the world of AI and business. Evans used to be a tech analyst at A16Z.
subscribe →Recent Reading
Things on the web we've been reading...
Anthropic-backed Ode launches as AI labs bet that embedding forward-deployed engineers inside enterprises is the key to accelerating enterprise AI adoption.
The telecom company is staking its future on open models, both to control token cost and protect its proprietary data.
Muse Glimmer is a 30-billion-parameter open agentic model from Meta Superintelligence Labs, optimized for always-on local workflows on consumer hardware.
AI lab’s Fable 5 has met with sluggish demand from corporate clients
