georgeelliot19.bsky.social
@georgeelliot19.bsky.social
That means confronting fossil fuel industry, AND confronting fossil fuel exporting countries, AND confronting economic growth dependence, AND confronting global inequality. Etc
December 18, 2025 at 2:44 AM
However AI is not the “answer” to this problem it’s just the next step in this longterm trend. The decreasing resource productivity of research has occurred *despite* increased technical capabilities. That is: AI is like a deep sea oil rig.
December 13, 2025 at 6:30 PM
The longterm trend of decreasing resource productivity of research is very real. It’s seen across almost all technologies and sectors and by some accounts across the economy as a whole (though attempts to measure this run into intractable aggregation problems, ie apples and oranges).
December 13, 2025 at 6:30 PM
we need some new economic arrangement potentially. Depends on how fast you want to cut emissions (and how you choose to distribute those emissions reductions across different nations and populations globally)
December 12, 2025 at 3:27 PM
But yes within current economic framework growth will always take precedent over climate mitigation. Because when growth stops lots of people lose their jobs, fail to pay their loans and lose their assets, fail to recoup on savings and investments. So if we want more rapid emissions cuts...
December 12, 2025 at 3:27 PM
Two solutions:
1) decouple employment from growth with government job guarantee. Redistribute labor productivity gains as wage increases and work time reductions.
2) decouple economic security from employment. Provide basic public services and/or basic income.
December 12, 2025 at 3:23 PM
Depends on circumstances. The reason economic contraction is bad for people is not because there's less stuff but because people lose their jobs. Capitalism requires growth for stability because technical innovation creates labor productivity gains and economic security is tied to employment.
December 12, 2025 at 3:19 PM
“In the future emissions reductions will be cheaper (technology) and we will be wealthier (growth) so we should wait to cut emissions”
December 10, 2025 at 8:21 PM
SRM doesn’t buy us time it kicks the can down the road. It’s more expensive to sequester CO2 in the future than to avoid emissions today (though these higher costs are more evenly spread over time). The economic logic behind SRM is the same basic logic used to delay climate action for decades.
December 10, 2025 at 8:21 PM
And every-time we mention solar radiation management it needs to be mentioned that 1) ocean acidification will continue even with SRM, and 2) SRM either needs to be combined with large-scale deployment of negative emissions technology and/or re-afforestation or you get warming shock with termination
December 10, 2025 at 8:12 PM
it really means to take BOTH climate change AND global inequality seriously. The top 10-30% of global population by income is responsible for almost entirely responsible for overshoot of planetary boundaries. Link: www.nature.com/articles/s41...
December 10, 2025 at 3:11 PM
For an example even the most ambitious decarbonization plans maintain high levels of global inequality and leave hundreds of millions of people in poor countries "in the dark" (below minimum resource thresholds required for satisfaction of essential needs). There's almost no consideration of what...
December 10, 2025 at 3:11 PM
There's so much talk about solutions what not enough talk about constraints, tradeoffs, conflicts, inequality. People will still be saying "we have the solutions! climate change is a political problem not a technical one" even if it's 2035 and emissions are still above 2000 levels (~30% decrease).
December 10, 2025 at 3:11 PM
Yes. I think the pervasive "solution-ism" among environmental scientists, engineers, policymakers amounts to a serious form of climate denial. Everywhere you look there is talk of solutions. Every day there's a new headlines claiming progress on decarbonization (while emissions rise).
December 10, 2025 at 3:11 PM
resources faster and more efficiently and at greater resource return than we did in the past. But today oil production is less productive *despite* increase technical capabilities *due to* depletion of easy to access oil resources. Is AI in science like a deep sea oil rig?
December 10, 2025 at 1:29 PM
see centuries of progress in decades. But today new ideas have become *intrinsically* harder to find and new problems have become *intrinsically* harder to solve. By analogy if we went back in time to the Texas oil boom with modern drilling capabilities we would be able to extract those...
December 10, 2025 at 1:29 PM
we see slower progress today than in the past. This suggests decreasing resource productivity of the knowledge production subsystem. And the reason is simple: we have "depleted" the "easy to access" ideas first. If we went back in time to 1750 with modern scientific capabilities we probably WOULD...
December 10, 2025 at 1:29 PM
We have many times more researchers, with many times more technical capabilities and resource use per researcher, today than in the past. Yet we dont see such dramatically accelerated progress. Across many domains of science and technology and by some accounts across the economy as a whole...
December 10, 2025 at 1:29 PM
The idea of a technological singularity is in my opinion mathematically implausible. People who claim AI will enable "more than a century of scientific and technological progress in less than a decade" need to explain why we dont *already* see this type of accelerated progress today.
December 10, 2025 at 1:29 PM
We spent something like ~$500 billion globally on datacenters this year. At between $500-$1000 per kW we could have installed an additional 0.5-1 TERAWATTS of solar panels for that money. Obviously these are very rough numbers (and exceed current productive capacity) but they are illustrative.
December 10, 2025 at 1:12 PM
Rapid decarbonization is likely to create competition for energy and materials between sectors. This zero-sum transition economy means we need to always talk about the opportunity costs of technologies. Not just what's the environmental impacts but what are the environmental opportunity costs?
December 10, 2025 at 1:12 PM
See for example this recent article by @louis-del.bsky.social and colleagues on the dynamic energy requirements of rapid low carbon transition. Main figures attached below. Link: www.sciencedirect.com/science/arti...
December 10, 2025 at 12:52 PM