I wrote this post one month ago, it received minus 29 votes and 6 x’s.
Do people still feel the same way? Or are you now realising that this man is trying to turn the US into his own personal Russia? That there is a model for this that he is following—look at Turkey or Poland or Hungary or Slovakia or Brazil or Argentina or Venezuela. All slighly different, but similar in the way that an apparently stable, mature democracy was hijacked by a populist movement and eventually became an authoritarian state where the constitution and the rule of law were gradually replaced by the whims of one individual.
I spent some time in Venezuela when Chavez was in power, and it is scarily similar to the US right now. At the time, it was early in Chavez’s rule, the economy was still working, the country was rich although with a lot of terrible poverty and many people, even educated people, supported Chavez’s vision of a more equal society. But now the country has been destroyed.
I have read a wonderful novel, Europe Central, by William Vollmann, which describes what it was like to live under Stalin. So much parallels what’s happening in the US today, from punishing people for expressing the “wrong” opinions, to, for example the way Stalin was the person who decided if Shostakovich’s latest works were acceptable or not—just like the way Trump is taking over the Kennedy Center.
And this is happening to the most powerful country in the world, the country that used to be the good guys in a world where Russia and China support so much that is bad.
To me this is utterly terrifying. And I’m not sure why EA’s don’t see this as a problem.
Is it that EA’s are secretly libertarians who actually think that some of what Trump is doing is good?
Or is it that we rather focus on narrow problems that seem more tractable, and leave the global political problems to others?
I suspect it’s mostly the way you’ve written it. As a rule of thumb, always aim for high reasoning transparency. I asked ChatGPT o3 to rewrite it in a style that’s more likely to appeal to EAs and that frames it in terms of reducing the risk of stable totalitarianism. I’ve pasted its output below.
”Claim. A second Trump presidency would raise the probability that the United States drifts toward a technologically-entrenched autocracy, thereby increasing the global risk of stable totalitarianism — a scenario where an oppressive regime locks in power for centuries or more.¹
Why that matters. Even a <1 % chance of permanent totalitarian lock-in constitutes an existential risk: it would foreclose almost all future value while inflicting vast suffering.² Emerging tech — especially frontier AI, ubiquitous surveillance, and autonomous weapons — could remove the usual checks (elite defection, popular uprising, leadership succession) that historically topple dictatorships.³
Mechanisms by which Trump plausibly raises the risk: • Erosion of democratic guard-rails. Intent to purge the civil service and use federal agencies for partisan aims weakens the institutions that normally resist autocratic consolidation. • Politicised AI and surveillance. Allies have floated centralising control of federal datasets and AI models; misused, these tools could neutralise opposition and entrench rule. • Norms against power transfer. Open refusal to accept electoral defeat in 2020 signals willingness to test the limits of constitutional constraint.
Scale & neglectedness. The U.S. controls ~25 % of world GDP and a decisive share of AI R&D; trajectory changes here propagate globally. Yet only ~$70 m/yr flows to non-partisan democracy-protection charities, versus >$10 bn in partisan spend.
Tractability. Cost-effective levers include:
State-level democracy infrastructure (voter-registration, local media fact-checks) — historical cost ≈ $300–400 per net vote.
Legal defence funds for civil-service whistle-blowers.
AI-governance policy work that limits executive control over surveillance and autonomous-weapon deployment.
Next steps for funders / organisers: commission a quick Rethink Priorities dive to refine the risk delta; pilot $1–5 m to the most effective democracy-protection orgs; reassess post-election.
¹ See 80,000 Hours problem profile on risks of stable totalitarianism. 80,000 Hours ² 80k’s BOTEC puts the century-level risk at ~0.3 %, with other experts’ estimates up to 5 %. 80,000 Hours ³ Advanced AI could give a ruler decisive military, surveillance and succession advantages, removing historic failure modes for dictatorships.”
I wrote this post one month ago, it received minus 29 votes and 6 x’s.
Do people still feel the same way? Or are you now realising that this man is trying to turn the US into his own personal Russia? That there is a model for this that he is following—look at Turkey or Poland or Hungary or Slovakia or Brazil or Argentina or Venezuela. All slighly different, but similar in the way that an apparently stable, mature democracy was hijacked by a populist movement and eventually became an authoritarian state where the constitution and the rule of law were gradually replaced by the whims of one individual.
I spent some time in Venezuela when Chavez was in power, and it is scarily similar to the US right now. At the time, it was early in Chavez’s rule, the economy was still working, the country was rich although with a lot of terrible poverty and many people, even educated people, supported Chavez’s vision of a more equal society. But now the country has been destroyed.
I have read a wonderful novel, Europe Central, by William Vollmann, which describes what it was like to live under Stalin. So much parallels what’s happening in the US today, from punishing people for expressing the “wrong” opinions, to, for example the way Stalin was the person who decided if Shostakovich’s latest works were acceptable or not—just like the way Trump is taking over the Kennedy Center.
And this is happening to the most powerful country in the world, the country that used to be the good guys in a world where Russia and China support so much that is bad.
To me this is utterly terrifying. And I’m not sure why EA’s don’t see this as a problem.
Is it that EA’s are secretly libertarians who actually think that some of what Trump is doing is good?
Or is it that we rather focus on narrow problems that seem more tractable, and leave the global political problems to others?
Are those of us in Europe missing something?
Could anyone enlighten me?
I suspect it’s mostly the way you’ve written it. As a rule of thumb, always aim for high reasoning transparency. I asked ChatGPT o3 to rewrite it in a style that’s more likely to appeal to EAs and that frames it in terms of reducing the risk of stable totalitarianism. I’ve pasted its output below.
”Claim. A second Trump presidency would raise the probability that the United States drifts toward a technologically-entrenched autocracy, thereby increasing the global risk of stable totalitarianism — a scenario where an oppressive regime locks in power for centuries or more.¹
Why that matters. Even a <1 % chance of permanent totalitarian lock-in constitutes an existential risk: it would foreclose almost all future value while inflicting vast suffering.² Emerging tech — especially frontier AI, ubiquitous surveillance, and autonomous weapons — could remove the usual checks (elite defection, popular uprising, leadership succession) that historically topple dictatorships.³
Mechanisms by which Trump plausibly raises the risk:
• Erosion of democratic guard-rails. Intent to purge the civil service and use federal agencies for partisan aims weakens the institutions that normally resist autocratic consolidation.
• Politicised AI and surveillance. Allies have floated centralising control of federal datasets and AI models; misused, these tools could neutralise opposition and entrench rule.
• Norms against power transfer. Open refusal to accept electoral defeat in 2020 signals willingness to test the limits of constitutional constraint.
Scale & neglectedness. The U.S. controls ~25 % of world GDP and a decisive share of AI R&D; trajectory changes here propagate globally. Yet only ~$70 m/yr flows to non-partisan democracy-protection charities, versus >$10 bn in partisan spend.
Tractability. Cost-effective levers include:
State-level democracy infrastructure (voter-registration, local media fact-checks) — historical cost ≈ $300–400 per net vote.
Legal defence funds for civil-service whistle-blowers.
AI-governance policy work that limits executive control over surveillance and autonomous-weapon deployment.
Next steps for funders / organisers: commission a quick Rethink Priorities dive to refine the risk delta; pilot $1–5 m to the most effective democracy-protection orgs; reassess post-election.
¹ See 80,000 Hours problem profile on risks of stable totalitarianism. 80,000 Hours
² 80k’s BOTEC puts the century-level risk at ~0.3 %, with other experts’ estimates up to 5 %. 80,000 Hours
³ Advanced AI could give a ruler decisive military, surveillance and succession advantages, removing historic failure modes for dictatorships.”
That is awesome feedback, James. Thank you!
You’re welcome :)