← Workshop
Local Futures title slide

Workshop 6

Agentic Futures, Curricular Sustainability

Wednesday, July 22, 2026 · 10 AM – noon · CHDR · last in-person session

Framing

Demoing the power tools

We're going to take a tour through the Claude CLI, MCP, Hugging Face, and Ollama. Feel free to follow along if you have your own device — and remember you can also try some of these tools later on our new CHDR Spark.

The Twilight of the Chatbots

Header art from Ethan Mollick's The Twilight of the Chatbots: a robot watching people run toward a machine city at sunset

Mollick, 'The Twilight of the Chatbots' — open in a new tab ↗

Altman: 'jobs that aren't real work'

  • tomshardware.com: the CEO's framing of which jobs disappear.
  • Whose work is 'not real'? Look at the verbs.
  • Most of what humanists do is on his list.
Altman jobs article
Graeber empty-room article
The point of bullshit jobs isn't that the work is hard or the workers are lazy. It's that the work is meaningless, and the worker knows it. AI is being marketed as the cure. It is at least as likely to be a more efficient producer.

— Echoing David Graeber's I Had to Guard an Empty Room

Demo

Claude CLI

Installation directions

Easiest way to use: select Repository → Open in Command Prompt (or Terminal) from GitHub Desktop and type claude.

Commands

  • Slash commands: type / for the menu — /init, /help, /model, and friends.
  • Shift+Tab: cycles the modes — normal, auto-accept, and plan mode.
  • Plugins: /plugin installs bundles of commands, agents, and MCP servers from a marketplace.
  • Skills: reusable instructions Claude loads on demand — Superpowers is a set of these.

Installing an MCP server

  • Example one: the Hugging Face plugin — models, datasets, and Spaces inside Claude.
  • Example two: the Zotero MCP — your research library shares a context with your writing.

Live demo

Claude CLI running local models

Using Superpowers, we'll describe the project, have Claude select the best local model for audio transcription, and build an interface for our transcription tool.

Live demo

Ollama

Installation — then the Claude Code CLI with Ollama: a local model driving the same power tool (a less competent agent).

Where will this take you?

All of these tools seem to be on a trajectory of growing more powerful — and potentially more expensive. But local models are also rising in capability.