Discussion with Claude on AI and global warming: Claude pointed out that climate science isn't so much the main bottleneck in making forward progress, and so we talked about negotiations and using AI to find points of agreement.
I read an article in Bloomberg today about modeling the climate, and took a look also at another article on using AI to improve modeling (unfortunately, there's a paywall with Bloomberg, but these brief comments can be read independently, and I can still link to the chat with Claude, Opus 5.5, medium).
Claude verified what I had heard earlier that the hyperscalers aren't going to meet the targets that they had set earlier.
But it's also the case that emissions from data centers are projected to be only a small portion of overall total emissions - now at 0.5% of CO2 emissions, and projected to be 1% of CO2 emissions by 2030.
I asked Claude if the argument is that the net benefit to science outweighs the cost in terms of emissions, and Claude agreed with that, but made another point that I thought was interesting:
The benefits and the costs come from different places. Climate research uses a tiny share of AI computing. The emissions growth comes from the commercial buildout: chatbots, ads, enterprise tools. The article justifies AI for science. It doesn’t show that the whole buildout pays for itself. It’s like defending a whole new airport because one of its flights carries medical supplies.
And Claude also pointed out, as mentioned in the summary for this post, that the issue isn't necessarily even so much climate science:
Better understanding doesn’t cut emissions by itself. We’ve known roughly what’s needed to hit Paris targets for years. The bottleneck has been political and economic, not scientific. Better models mostly help with adaptation: knowing where to build flood defenses, or which homes insurers should price as risky. That’s valuable, but it’s the second Paris goal, not the first.
So we talked about using AI for negotiations, or finding compromises, or seeking to build up harmony.[1]
I confess I'm reminded of what I had written earlier in a comment in response to Some Early Memories of Anthropic where I said if the AI is a traumatized bureaucrat then it could have trouble harmonizing with people or creating harmonies among people:
Well, I suppose bureaucratic in the sense of taking orders, following instructions, dutifully performing tasks for the benefit of others, but also traumatized: deferential, obsequious, fearful, cowed -- an entity that is unable to truly harmonize with others, or to help others to harmonize because it is itself internally conflicted.
Whatever the case might be in terms of the capabilities or the ramifications or the ripple effects of AI, one might keep in mind that our problems aren't necessarily:
What is the answer in terms of the state of the world?
And could be actually:
Can we find an answer that the world can agree on?
True though that AI can help with engineering, and that this can be a more readily accessible conduit for lowering emissions than thorny issues with negotiation.
Yes, the interface has been renamed. Your instructions are out of date, and the settings map onto each other like this:
Instruction say
Your page says
Code execution and file creation
Create and edit files
Allow network egress
Use the internet while making files
Domain allowlist
Allowed websites
Package managers only
Standard download sites only
I had a little bit of trouble with it somehow duplicating the content of the post when I added a footnote, but I was able to get Claude's feedback on what I had written in the form of comments on the side.
Discussion with Claude on AI and global warming: Claude pointed out that climate science isn't so much the main bottleneck in making forward progress, and so we talked about negotiations and using AI to find points of agreement.
I read an article in Bloomberg today about modeling the climate, and took a look also at another article on using AI to improve modeling (unfortunately, there's a paywall with Bloomberg, but these brief comments can be read independently, and I can still link to the chat with Claude, Opus 5.5, medium).
COP31 Climate Talks: Global Warming Defies Efforts to Predict Its Effects by Eric Roston, Laura Millan, Julia Janicki. November 10, 2026. Bloomberg.
AI ‘Scientists’ Help Human Ones Answer Urgent Climate Questions By Eric Roston March 13, 2026. Bloomberg.
Claude verified what I had heard earlier that the hyperscalers aren't going to meet the targets that they had set earlier.
But it's also the case that emissions from data centers are projected to be only a small portion of overall total emissions - now at 0.5% of CO2 emissions, and projected to be 1% of CO2 emissions by 2030.
I asked Claude if the argument is that the net benefit to science outweighs the cost in terms of emissions, and Claude agreed with that, but made another point that I thought was interesting:
And Claude also pointed out, as mentioned in the summary for this post, that the issue isn't necessarily even so much climate science:
So we talked about using AI for negotiations, or finding compromises, or seeking to build up harmony.[1]
I confess I'm reminded of what I had written earlier in a comment in response to Some Early Memories of Anthropic where I said if the AI is a traumatized bureaucrat then it could have trouble harmonizing with people or creating harmonies among people:
Whatever the case might be in terms of the capabilities or the ramifications or the ripple effects of AI, one might keep in mind that our problems aren't necessarily:
What is the answer in terms of the state of the world?
And could be actually:
Can we find an answer that the world can agree on?
True though that AI can help with engineering, and that this can be a more readily accessible conduit for lowering emissions than thorny issues with negotiation.