Other Challenges
Beyond the two set challenges β a deeper menu to explore or run as alternates.
These are the rest of the menu β beyond the two set challenges we run live. Each follows the same format: a short setup, a starter artifact you can paste straight into any chatbot, a few rounds of iteration, and a quick share-out. Use them as alternates, as warm-ups, or as something to try on your own afterward.
New to AI? Start anywhere on the first grid β these assume no prior experience. Already comfortable with these models? Skip to the advanced track and wire up some agents.
Core & foundational challenges
π Picture This
Rescue an ugly data table and turn it into the most informative figure in the room. Beginner-friendly, instant payoff.
π£οΈ Plain-Language Gauntlet
One dense methods paragraph, three very different audiences.
π° Hypothesis Roulette
A surprising null result walks in. Spin up five testable mechanisms.
π Model Horserace
Same task, five different free AIs β compare what each one does well.
βοΈ The Rebuttal
Reviewer 2 was brutal. Draft a response that brings the editor around.
β±οΈ IRB Speed Round
Draft a risks & benefits section fast β and catch the AI when it bluffs.
π§ͺ Experiment Generator
From a single finding to six experiments that nail the mechanism.
π Rapid Review
A 2-page evidence brief with real references, built in minutes.
π₯ Flash Debate
Build the evidence case for a contested claim β then argue it.
π€ Meet the Speaker
Walk into a visit knowing exactly who to meet and what to say.
ποΈ Policy Wonk
The general version of the comment challenge: turn any dense rule change into an effective public comment.
Advanced track β agentic tools (Claude Code / Codex / Gemini CLI)
These three move beyond the chat box. You will drive command-line agents that read files, run code, and coordinate with each other. Come with a tool installed (see the AI Toolkit) and some curiosity about how multiple agents work together.
π§ Reviewer 4
Stand up a multi-agent grant clinic that critiques your R01 like a study section.
π Who Are My Friends?
Design a system that finds the reviewers who will appreciate your paper.
𦴠Junkyard Dog
Sniff out open pain datasets, fetch them, and wire them together with code.