AI Circle · Resources

Field notes from the operational frontier.

Deep dives on evaluations, post-training, human data, infrastructure, and the systems that decide whether AI actually works.

Start here Maps Blog Newsletter ↓

The AI Learning Guide

Start here

New to AI, or trying to keep up? The AI Learning Guide is what our researchers and operators send when someone asks where to begin. Pick the path that fits you, or read it end to end.

Open the guide →

I’m new to AI

Start with the handful of papers and books that explain modern AI.

Start with the basics → 02

I want to keep up

One podcast, one newsletter, and the people worth following.

Build the habit → 03 with

I want a certificate

Courses and credentials employers recognize. Free Data Gradient certificates are included.

See courses → 04

I want to know who’s building

A filterable map of 140+ companies across the AI stack.

Explore the landscape →

From the learning guide

Three picks from each part of the guide. Every list goes deeper on the full page.

Open the full guide →

Agentic Design PatternsBy Antonio Gulli. A code-backed curriculum covering 21 design patterns for building autonomous AI agents. Includes pattern overviews, practical use cases, hands-on examples, and key takeaways. Your recipe book for implementing agentic AI systems using LangChain/LangGraph, CrewAI, and Google ADK.

Co-Intelligence: Living and Working with AINYT bestseller from Wharton professor Ethan Mollick. A practical playbook for working and learning alongside AI. Useful for anyone who wants to work well with AI while staying grounded in what makes us human.

Hands-On Large Language ModelsBy Jay Alammar and Maarten Grootendorst. A visually-driven guide with nearly 300 custom illustrations that takes you from zero to confident practitioner in LLMs. Covers transformers, tokenizers, semantic search, RAG, and fine-tuning with intuitive explanations and working Python code throughout.

All 15 books →

Latent Space Podcast: The AI Engineer PodcastCovers foundation models, code generation, multimodality, AI agents, and GPU infrastructure directly from the founders, builders, and researchers working at the frontier. Features interviews with leaders from OpenAI, Anthropic, Meta, Databricks, Modular, and more. Known for high signal-to-noise ratio, deep technical discussions, and breaking news.

Cognitive Revolution PodcastHost Nathan Labenz scouts AI from every angle. Features interviews with AI builders, researchers, and investors, plus original deep-dive "AI Scouting Reports" on critical topics. Covers a wide range from AI safety debates and global AI policy to practical applications in sales, education, healthcare, and more. Good for business, policy, and academic leaders staying current on AI developments and implications.

No Priors Podcast: AI, Machine Learning, Tech, & StartupsAI and entrepreneurship podcast hosted by investor Sarah Guo (founder of Conviction) and serial entrepreneur Elad Gil (co-founder of Color Health, author of High Growth Handbook). Features conversations with leading engineers, researchers, and founders about the AI revolution covering topics like AGI timelines, market disruption, state-of-the-art research, and how commerce and society will change. Accessible for non-technical listeners while still diving deep on industry trends.

All 17 podcasts →

Cohere LLM UniversityFree curriculum covering LLM fundamentals (embeddings, transformers, attention) alongside practical modules on semantic search, prompt engineering, RAG, and tool use. Good for learning core LLM concepts with hands-on API practice.

AWS Skill BuilderAWS's official learning platform with 1,000+ free resources plus subscription labs and certification prep. Covers AI/ML, generative AI, cloud architecture, and all AWS services.

NVIDIA Deep Learning InstituteFree and paid training in deep learning, AI, and GPU-accelerated computing with hands-on cloud GPU environments. Self-paced courses (1-8 hours) and instructor-led workshops across computer vision, NLP, robotics, and data science.

Free certificates · Data Gradient Get certified in AI data work

Free courses and certificates for AI training and evaluation work.

Browse courses ↗

All courses →

Andreas HornFounder of Human in the Loop and author of The Generative AI Career Masterplan. Lectures in the AI master's program at ZIGURAT and advises AI companies. Previously Head of AIOps for EMEA at IBM.

Eduardo OrdaxLeads go-to-market for generative AI services across EMEA at AWS. Previously led go-to-market for MLOps at AWS and was General Manager AI at Vodafone.

Dr. Han Héloir-YanLeads partner solutions architecture for EMEA at Mistral AI. Previously built and led the EMEA generative AI solutions architect team at MongoDB. Earlier worked at Murex, Huawei, and Orange.

All people →

Maps

Interactive market maps of the companies behind AI’s human data. Each one is searchable, filterable, and open to suggestions.

View all maps →

01 · Market map

Human Data for Physical AI

Companies providing human data for physical AI. Grouped by data type and limited to those with public funding above $10M.

Explore the map →

02 · Interactive atlas

Physical AI Atlas

Search, filter, and shortlist human data vendors for physical AI across egocentric, teleoperation, synthetic, and more.

Open the atlas →

03 · Interactive map

Multilinguality Map

Companies selling multilingual human data for AI, across localization, evals, speech, RLHF, and low-resource languages.

Open the map →

04 · Interactive atlas · New

Coding Atlas

Companies selling coding data for AI, across code SFT, RLHF and code evaluation, benchmarks, agentic environments, and security.

Open the atlas →

Blog

Field notes and deep dives on evaluations, post-training, human data, and the systems that decide whether AI actually works.

View the archive →

Latest · Kimi K3

Kimi K3 Just Changed The Way AI Models Are Built

Chinese AI labs just caught up in a big way. Kimi K3’s architecture achieves 2.5x training efficiency with impressive memory recall.

Read the brief →

Deep dive No. 02 · Meta FAIR

Your brain just became an LLM.

Meta’s TRIBEv2 predicts how video, audio, and text drive human brain activity. It then uses that to reproduce classic neuroscience experiments.

Read the deep dive →

Edition No. 01 · Multilingual evaluation

Your AI agent isn’t wrong. Your benchmark is.

What a dinner with LILT’s CEO revealed about a benchmark error distorting how we evaluate AI in every language but English.