Vanishing Gradients

Hilary Mason, CEO of Hidden Door, works with agents across software development and creative projects. Drawing on Richard Hamming’s distinction between science and engineering, she describes agents as “bringing the science back into the engineering and the engineering back into the science.” When you can build and test more ideas, how do you decide which ones are worth keeping? A plausible analysis still needs scientific judgment, creative output needs taste, and a chart needs to preserve its meaning when an agent revises it. In this second episode of Show Us Your Agent Skills, Thomas Wiecki &

Agents can write the code, run the analysis, and produce a convincing explanation. How do you check that they’ve answered the right question, used the data appropriately, and reached a conclusion you can act on? I joined Gaël Varoquaux (scikit-learn, Inria, and Probabl), Shipra Arora (Bain & Company), and Luca Fiaschi (PyMC Labs) to explore how data scientists are delegating work to agents and building verification into that process: combining programmatic checks, agents critiquing other agents’ work, and human judgment about the problem and its consequences. That changes where we spend our ti

“It’s an LLM, a loop, and enough tokens.” —Nicolay Gerold, Amp Code So why does a coding agent forget your instructions, keep reading tiny pieces of a file, or struggle with an edit another model handles easily? Nicolay Gerold (Amp) joins Hugo to take that loop apart and explain the harness around it: the software that executes tools, manages context, and lets you steer the agent’s work. Nico builds this machinery for Amp, a coding agent that works across software projects. “Every component in a harness is basically an assumption the model can’t do it on its own. This expires over time.” — Nic

“If you feel that a product is hard to eval, or you feel like, ‘I don’t even know how to eval this,’ it’s a strong smell that your product isn’t good.” — Hamel Husain, on AI product design An AI data agent tells you last quarter’s net revenue. It doesn’t show the metric definition, source tables, filters, query, intermediate calculations, or assumptions. You can’t trust the answer without asking a data scientist to reproduce it. Hamel Husain argues that the eval problem is evidence of bad product design. If the user can’t inspect the work well enough to decide whether the answer is right, anot

“It would be sad if local models were not an option and there were only proprietary models. It’s good to have alternatives. Competition is good for business.” — Sebastian Raschka, on open-weight AI Kimi K3’s weights landed about an hour before Hugo Bowne-Anderson and Sebastian Raschka went live. Sebastian had already updated his architecture diagram. That speed captures his approach to the current model wave: wait until the weights exist, run the model in the harness where it will actually work, then inspect the architecture closely enough to understand what changed. The conversation arrived d

When ML/AI Engineer William Horton last joined me, Maven Assistant had reached its first external users the day before. The healthcare AI agent was available to 20 percent of Maven Clinic’s users, and the team had deliberately withheld answers about benefits. A wrong response could shape a decision involving $15,000 of fertility coverage, and the evals had not earned the right to ship it. Four months later, Maven Assistant is available to 100 percent of users, benefits answering is live, and weekly conversation volume has grown by roughly ten times. Real usage also overturned part of the roadm

Every capability in an agent needs its own evidence and release bar. A model-provider slip, an incorrect tool call, and a wrong fertility-benefits answer should not be held to the same pass rate. William Horton, Staff AI Engineer at Maven Clinic, joined us the day after Maven Assistant reached its first external users. The agent helps members inside Maven Clinic’s women’s and family healthcare platform find providers, manage appointments, navigate Maven, and get basic health information. William had spent much of launch day reading chat traces and turning the surprises into product decisions a

Nicolay Gerold works all day and night on AMP, one of the most interesting coding-agent harnesses out there. If you’re building with coding agents, this conversation will help you understand: * when to trust the model, * when to build harnesses around it, * which model is worth paying for, * which programming languages gives the agent better feedback, and * when to take the keyboard back. Coding-agent products are living inside a blender. Opus 4.8 to Fable changes what the model can be trusted with, eats a workflow, and suddenly the best product decision is to delete code. AMP had handoff beca