Hands-On LLMs Authors Talk the Shift From LLMs to Agents at PyData
MaartenGr · x · 2026-09-16
Jay Alammar and Maarten Grootendorst met face-to-face for the first time after writing a book together, sharing the stage at PyData Amsterdam on the fundamentals of what makes an Agent.
- Using Gemma 4 from Google DeepMind, they showed the industry shift from LLMs to Agents and from Words to Actions — the Gemma 3 → 4 transition reflects how model development itself is increasingly about offloading human tasks.
- Jay highlighted that evaluation remains a key frontier, especially given the complexity of agentic systems, drawing on his work on Cohere's North Mini Code.
More from coding & agent
- AgentIR embeds agent reasoning traces into retrieval, plus a BrowseComp-Plus benchmark — CShorten30 · 2026-09-16
- AI Evals FAQ grows to 48 Q&As: sensitive data, huge traces, and stale gold datasets — HamelHusain · 2026-09-16
- One prompt to a real object: AI agent models in Blender and 3D-prints the result — anthara_ai · 2026-09-16
- Open-source crew of 8+ AI agents manages your Obsidian vault, email triage included — tom_doerr · 2026-09-16
- AI agent gets user's $70 Amazon credit after weeks of runaround, one tap needed — armand_ruiz · 2026-09-16
- 3 weeks through Stanford CS329A: the generator has outrun the verifier — le_james94 · 2026-09-16