“We shape our tools, and thereafter our tools shape us.” — John Culkin

A few weeks ago, every electronic note I had taken vanished. The index cards survived in a drawer. The files did not. I have written before about turning that loss into this public study series — proof that a setback can become a contribution. Today’s lesson continues that work, but it turns the lens around. Instead of studying climate, we study the instrument I use to study climate.

Because here is the truth I keep arriving at: the climate analyst’s job is no longer only what you know. It is how well you can think alongside a machine that knows a great deal too.

AI Fluency Is Now Part of the Job Description

Consider what the work actually demands. A Climate Policy Analyst reads the Climate Leadership and Community Protection Act and its Scoping Plan — hundreds of pages of statute and strategy. The analyst pulls emissions figures from the EPA’s Greenhouse Gas Reporting Program and from the Energy Information Administration. The analyst then translates technical findings into language a community board, a legislator, or a neighbor can act on. That is synthesis, sourcing, and clear writing — performed at volume, under deadline.

This is exactly the territory where a capable AI partner earns its place. Not as a replacement for judgment — judgment is the one thing that cannot be delegated — but as a research co-worker that drafts, summarizes, cross-checks, and frees the analyst to do the thinking only a human can do.

I have said it across this series and I will say it again: AI is a colleague, not a substitute. But a colleague is only as useful as your skill in working with them. That skill has a name now. It is called AI fluency, and it can be learned. If you are interviewing for a policy role — as I am — expect to be asked how you use these tools. Being able to describe your AI-assisted research workflow clearly, and to name its limits honestly, is no longer a bonus. It is part of the qualification.

Thirteen Doors, One Building

In March 2026, Anthropic opened a free training catalog called Anthropic Academy. Thirteen courses. Self-paced. A certificate on completion. No paid subscription and no cost — only an email address.

Five of those courses are built for developers: building with the API, agentic coding, the Model Context Protocol, and cloud deployment. A climate analyst who researches and writes does not need to walk through those doors. Eight others teach the discipline of working with AI well — and that discipline is the curriculum for today. The full catalog, for the record:

The Path I’m Walking

Here is the sequence I am taking, and the sequence I recommend to anyone studying toward a climate research or policy role. It is a path, not a pile. Each course earns its position, and each one builds the one that follows.

  1. Claude 101 — start here, no exceptions. Before anything else, you need to know how to actually talk to an AI system in a way that produces useful work. Climate research means synthesizing dense reports, pulling key figures from long PDFs, and drafting clear prose from complex data. Claude 101 gets you doing all of that quickly. Everything later assumes this foundation.
  2. AI Fluency: Framework & Foundations — the most important course on this list for our work. It teaches you to think strategically about when AI helps and when human judgment must lead — a critical distinction when you write about science and policy, where accuracy and credibility are the whole job. This is where you learn the 4D Framework. Treat it as the core, not the warm-up.
  3. AI Fluency for Students — do not let the name deter you. The research and literature-review skills in this course map directly onto the analyst’s daily reality: academic sourcing, synthesizing long-form material, and structuring an argument that holds. The label says “student.” The skill is professional.
  4. AI Fluency for Nonprofits — climate work lives next door to advocacy. Policy writing, public communication, and persuading a non-technical audience are the exact territory this course covers. It is built for lean, mission-driven teams trying to communicate clearly and move people to act — which is the climate communicator’s challenge stated plainly.
  5. Teaching AI Fluency — the protégé effect, made formal. Once you have built your own workflow, this course helps you articulate and teach it. That matters twice over: explaining a method to others deepens your own command of it, and the ability to walk a hiring panel — or a blog readership — through your AI-assisted research process becomes part of your value as an analyst.

One bonus, no coding required. Anthropic also publishes free Jupyter notebook courses on GitHub covering prompt engineering, real-world prompting, and tool use. Even if you never run a line of the code, reading through the prompt engineering tutorial builds genuine intuition for getting sharper output when you research and draft. The full academy lives at anthropic.skilljar.com.

A Correction Worth Making

My first set of notes on this subject — the ones I am now rewriting from memory and from the index cards — contained an error. I had written down a “4E Framework” and named its four parts Effective, Efficient, Ethical, and Safe. That is wrong, and the correction is itself today’s real lesson.

The framework Anthropic teaches is the 4D Framework. Its four parts are competencies, and each begins with D:

  • Delegation — deciding whether, when, and how to engage AI on a task at all.
  • Description — communicating your goal clearly so the AI behaves usefully. Prompting lives here, and here only.
  • Discernment — evaluating the output and the process with a critical eye.
  • Diligence — taking responsibility for what you do with AI and how you do it.

So where did my four E-words come from? They are the goals, not the framework. The four D’s are the skills; effective, efficient, ethical, and safe are what those skills are meant to produce. I had mistaken the destination for the road.

I am leaving the error visible on purpose. Catching it is Discernment in practice — evaluating a source, including your own past notes, with a critical eye. An analyst who cannot audit their own work has no business auditing anyone else’s. Another card for the Mental File Cabinet, corrected and refiled.

Today’s Comprehension Check

Read it once, then come back and answer without scrolling up. Teaching the answer to someone else is the surest test that you know it.

  1. The 4D Framework names four competencies. Name them — and say which one prompt-writing belongs to.
  2. “Effective, efficient, ethical, safe” — is that the framework, or the framework’s goal? Explain the difference in one sentence.
  3. A DEC hiring panel asks how you use AI in your research. Answer in three sentences, and do it without claiming that AI replaces your judgment.
  4. Weak-area revisit, climate policy: State the CLCPA’s two headline 2030 targets — one for emissions, one for electricity.
  5. Drafting a policy memo with an AI thinking partner, in iterative back-and-forth: is that automation, augmentation, or agency?

Answer key. (1) Delegation, Description, Discernment, Diligence; prompting belongs to Description. (2) The goal — the four D’s are the skills you practice, and effective/efficient/ethical/safe is the result those skills are meant to produce. (3) Open response; a strong answer names a specific task, describes the human check you apply to the output, and states a limit. (4) A 40% reduction in statewide greenhouse gas emissions from 1990 levels by 2030, and 70% renewable electricity by 2030. (5) Augmentation — human and AI collaborating as thinking partners.

The Assignment

Take Claude 101 this week. Not next month, this week. It is free, it is short, and it is the door every other course opens off of. Then walk the path in order.

If you are studying toward a climate policy role, treat this as part of the curriculum, not a detour from it. The analyst who can read a Scoping Plan is valuable. The analyst who can read a Scoping Plan, work fluently with an AI research partner, and explain exactly how and why — that analyst is what the moment requires.

The climate and ecological emergency does not wait for us to feel ready. It does not pause while we rebuild our notes. So we learn the tools, we learn them well, and we keep the work moving — for the communities counting on the policy, and for the generations who will inherit whatever we manage to get right.

Filed publicly so the next candidate does not have to start from nothing. — Mr. Alvarez, cCcmty.com