AI Outputs

Source: ChatGPT - Leveraging AI Tools to Help Solve Climate Change

The entries below describe ways the Climatographers are using AI tools to generate content that can be useful in the delivery of audience-specific actionable climate knowledge. Make sure to subscribe to our Climate Learning With AI newsletter on LinkedIn to track this work. Click on the hyperlinks to access the workproducts themselves. Note that we don’t detail the prompts we used, which can be quite detailed in their own right; drop a line to discuss if interested. In many cases we’ll continue to refine both the prompts and the workproducts, which you’ll also often see integrated into our Climate Web and Your Climate Toolbox knowledge solutions. Note that this is just one of five Knowledgebases you have access to in this specialized “mini-brain.” To see the larger collection click here, or on the “Home” icon at the top center of your screen.

3/18/2025 Bi-Monthly Climate News Update - ChatGPT Pro Deep Research

3/18/2025 Bi-Monthly Climate News Update - Google Gemini Deep Research

This is a project we’ve been working on for a while - namely how to stay up to speed on climate developments. ChatGPT Pro has done a good job (see below) building a monthly summary. With the drop of new models from Google this week, and in particular its changes to its Gemini Deep Research tool, we wanted to test the prompt we’ve used in the past to generate a summary for just the first two weeks of March. We tested with ChatGPT Pro’s Deep Research as well as Gemini’s Deep Research. Both can be accessed through the links above. Both illustrate some of the challenges involved, e.g. a news story that came out during the right timeframe, but covering a publication from a year ago. Or the use of rather obscure sources. Nevertheless, they’re both quite informative, and it turns out there’s a lot even we had missed in terms of what happened over the last two weeks! We also discovered through ChatGPT that Skeptical Science puts out a weekly list of news stories, a great resource we had never seen before! Thoughts, Suggestions, Feedback welcome!

3/17/2025 A Sample Pre-Mortem

If you follow our LinkedIn posts you’ve seen reference to Pre-Mortems, which represent a useful way of exploring the future. Instead of presenting a range of scenarios, a Pre-Mortem “explains” a particular outcome, whether good or bad. It’s a psychologically powerful way of making a potential outcome “more real.” It’s not easy to do, and with the crop of new AI models (LLMs) in the last couple of weeks we wanted to see how easy it might be do to a useful Pre-Mortem. We asked the LLM to explain why PacifiCorp (a large utility here in the PNW that we know well) went bankrupt in 2030. It first responded with a long list of variables that might possibly contribute to a utility having problem, but when asked to specifically write a pre-mortem it did a darned good job! It was generated by Gemini’s 2.0 Flash Thinking model. We’ll be doing a lot more work along these lines to leverage AI tools to better communicating climate risks. Thoughts, Suggestions, Feedback welcome!

3/17/2025 Closing Arguments By Fossil Fuel Attorneys

There is a lot of focus on climate change litigation today, but we have substantial questions about the success of such litigation, and how an attorney for fossil fuel interests might present closing arguments at such a trial. We asked four AI models to draft a 30-minute closing argument at a major trial in which fossil fuel interests are being sued for climate damages in one form or another. The models were 2 Google Gemini models, and ChatGPT’s Plus and Deep Research models. ChatGPT Deep Research did a substantially better job than the others. It is a long document, and there are all kinds of ways we’d suggest improving it, but thought we’d included it here as-is. At almost 8,000 words it really is an amazing first-cut that would otherwise have taken many hours to draft. Thoughts, Suggestions, Feedback welcome!

3/17/2025 An Anti-Free Market Manifesto

In 1971 the about to be Supreme Court Justice Lewis Powell wrote a secret memo that has come to be called the Free Market Manifesto, arguing that free markets were under radical attack from all sides and needed to be defended. You can find a lot of on-line information about the Powell Memo. We asked ChatGPT Pro to take a look at the memo, and to craft an “anti-” version of the memo, in effect suggesting that free markets need to be regulated to prevent environmental and other damage. The result is quite good, but not particularly helpful. Interestingly, most of the things being suggested, e.g. that companies need to engage in longer decision-making timeframes, have gotten a lot of attention and yielded relatively little. If anything, this “anti-” version of the Powell memo suggests just how hard it is to come up with something new and compelling with respect to correcting the kinds of free market problems, e.g. externalities, leading to climate change. Thoughts, Suggestions, Feedback welcome!

3/1/2025 February, 2025 Climate Change News Summary

Our monthly in-depth look at climate news from the previous month, generated with ChatGPT Pro.