In This Newsletter
We're Opening the Launchpad Next Week!
The AI/ML Career Launchpad will be opening enrollment next week. The waitlist will hear it first, so make sure you're on that list.
We've added a ton of new things and I'm really excited to bring in a new group.
And if you join now, you'll be there in time for the holiday gift… here’s a sneak peek:
ML Foundations for AI Engineers
If you want to be an effective AI engineer, you need to understand some machine learning. Luckily the ML you actually need is pretty manageable.
AI, ML, deep learning, and LLMs get smooshed together a lot, and as an AI engineer you really don’t need to be an ML expert. But some core ideas still show up in the systems you build, and knowing them helps you build things that hold up (and get along with your ML colleagues).
After eight years in ML and coaching 400+ people working on transitioning into AI roles, I have a pretty good sense of what you need to know and what you can safely skip. A few things I cover:
Overfitting applies to you too. If you tweak a prompt against the same twenty examples and then report your eval score on those same twenty, you've overfit the prompt. Train test splits matter for AI, too.
Accuracy is not usually the best metric. If fraud is 0.1% of transactions, a model that always says "not fraud" is 99.9% accurate but 100% useless. That's why precision, recall, and ranking metrics like recall at k matter.
How an LLM becomes an assistant. Tokens, embeddings, attention, and transformers, plus the fine-tuning and RLHF steps that turn a base model (basically autocomplete) into the chatbot you use every day.
Or read the companion guide here.
Stuff I've Been Enjoying Lately
Movie: You Can See Everything. I love Nathan Fielder (from way back in his Nathan for You days) and this was great.
Technical book: The Hitchhiker's Guide to Agentic AI. Free, great depth, and a good addition to AI Engineering by Chip Huyen.
Fiction: The Fifth Season by N.K. Jemisin. The world-building is incredible, so engrossing.
Educational YouTube: Maxinomics. Great breakdowns with incredible production.
Fun YouTube: Speeed. Fun videos on everything from plants to bikes from every era. A great pick if you want to get your girl into cars.
Music: I've been working on my motivation jammers playlist.
Podcast: It's a tie between Modern Wisdom and Art of Accomplishment, so I'll share this collab.
Volunteer Opportunity
This program brings together two things we care about at NYC Data Science for Social Good: social entrepreneurship rooted in community relationships, and human-centered AI that helps nonprofits solve real problems. You'll learn with a cohort, hear from practitioners, and put the learning into a hands-on capstone.
Here's the path:
Lead with your community: four required leadership and organizing workshops, grounded in Marshall Ganz's framework. Develop your public narrative, build relationships, and plan an event or campaign with a nonprofit partner.
Build responsibly with AI: five hands-on workshops covering ethics, a trauma-informed case study, security, machine learning, and LLMs. The leadership workshops come first and are required before the AI workshops.
Put it into practice: about four weeks to build a human-centered solution with a nonprofit partner or on an internal project. Interview people about the problem, plan deployment, measure impact, and present at demo day.
You don't need a technical background. What matters is a real desire to learn and a community problem you care about. If you already work with data or AI, this is a chance to connect those skills with organizing and the needs of the people you're building for.
Applications close December 1, 2026.
Apply here
Program details: datadiplomats.org
Forwarded this email? Sign up here.
Note: This email may contain affiliate links. If you make a purchase I may make a small commission, at no cost to you. Thank you for your support!