And let's contrast that with the AI hype. It's more the opposite, a kind of solution to problems we didn't really have, but are now being persuaded we do. It would be sensible to invest an equal share of the resources currently being pumped into AI with uncertain outcomes into the complex issue of climate change. And, no, AI won't solve it; unfortunately, it only makes it worse.
I don't know what to do.
> Policymakers and the public, however, remain largely unaware of the risks posed by such a practically irreversible transition
Most people still underestimate what it means for the earth system to change from the current stable state into another state, which might need many years to become stable again. And that new stable state might be a lot less favourable for us humans.
There could indeed be self-reinforcing feedbacks and tipping dynamics that make the problem worse.
There could presumably also be self-dampening, stabilising dynamics that come into play at certain levels.
Is there any science on why the former is more likely than the latter?
'A "greenhouse Earth" is a period during which no continental glaciers exist anywhere on the planet... Earth has been in a greenhouse state for about 85% of its history.
'Earth is now in an icehouse state, and ice sheets are present in both poles simultaneously... Earth's current icehouse state is known as the Quaternary Ice Age and began approximately 2.58 million years ago.'
https://en.wikipedia.org/wiki/Greenhouse_and_icehouse_Earthh...
Water vapor (clouds) is a stonger greenhouse gas than carbon dioxide. We already got measurably higher temperatures, so we also have higher water evaporation, and from the last 5 years it looks like it happens every year.
So the runaway is already happening, until something stops it near hothouse conditions or hopefully earlier than that.
* Australia's renewables generation increased from 13.7% in 2015 to 42.9% in 2025 [1]
* EIA: 99%+ of new US capacity in 2026 will be solar, wind + storage [2]
* Wind and solar overtake fossil power in the EU for the first time in 2025 [3]
1. https://explore.openelectricity.org.au/energy/nem/?range=all...
2. https://electrek.co/2026/01/28/eia-99-of-new-us-capacity-in-...
3. https://ember-energy.org/latest-insights/european-electricit...
We can't cut emissions fast enough politically, but we can race towards economically viable fusion power which would solve the problem from the supply side and would make industrial scale carbon sequestration not insane, for a century or so, until waste heat itself can't be radiated fast enough even in 250 ppm CO2 atmosphere - but that's a problem for the XXII century.
Does this mean during mid-to-late Pleistocene it used to be -6C to +2C, or does it mean it used to be 8C to 16C?
(If the former, then how did early humanoids, and many other animals, survive such cold?)
Canada has 14 tonnes carbon footprint per person. Canada is a cold country.
France has 8 tonnes carbon footprint per person. Climate is way warmer.
We can't continue adding population and the wondering what is going on.
AGI is here
If such a thing existed, we could be sure that environmentalists and leftists would have openly embraced it, rather than nip it at the bud 50 years ago. Because they are Good People™. And we should definitely listen to them now because they Follow The Science™.
>Despite decades of research and sophisticated computational climate modeling, the magnitude and pace of these events have surprised scientists, raising questions about how well current climate projections capture risk.
When this always-revise-in-one-direction phenomenon happened with the electron charge, it was considered a priori "proof" that scientists were fudging their data to match expectations. The Millikan Oil Drop Experiment is still studied in fundamentals of science class.[0]
If climate scientists are constantly revising their predictions upward, then this is equally "proof" that climate scientists are under pressure to revise their estimates downward. Far from being "alarmist," such terms are actually cudgels used to discourage climate scientists from making their data look too bad.
The result is the predictable fudging of climate data to look better than it really is.
[0] https://en.wikipedia.org/wiki/Oil_drop_experiment#Millikan's...
Unless the fake reality starts to crack with the Epstein and other current events and humanity's coming of age will happen. Even if a hundred years later than Bonhoeffer though it would.
I always wondered if we just lacked the ability to mobilize to solve big problems anymore but now I look at this 7% US GDP being allocated to AI datacentres and I realize that it isn't a lack of ability, it's a lack of desire.
Imagine if we had ram shortages because all the silicon was being diverted towards making solar panels. Imagine if we had copper shortages because it was going to the windings on wind mills. Imagine if all these economic disruptions were just temporary and for a better cause or eliminating carbon emissions and eventually moving to sequestration of carbon.
Instead we get chatbots. And funny picture makers.
Why? Because candidly looking at those risks as a society means deep collective existential dread, which automatically means an immediate civilizational collapse.
So I'm guessing some of our elite is actually ignorant and the other part is willfully shutting the hell up on this subject to let our civilization run on fumes a few more years.
It's unfortunate because a rapid civilizational collapse could give humanity as a species a better chance of survival.
Not realizing we’re gonna go extinct here in the next thousand years unless something solves it
Since humans are incapable of doing this there’s only one possible option: To create something smarter than us and give it the power to solve it because we cannot