Trump may see restricting Mythos and Fable as a matter of national security. But the argument cuts both ways, and with Washington now asking if AI is too important for everyone to have access, other governments are asking whether they can afford for Washington to decide who does.
大多数人可能认为美国限制AI访问是出于国家安全考量，但作者认为，这种行为实际上促使其他国家质疑美国对AI技术的垄断控制权，并重新评估依赖美国AI技术的风险。这一观点挑战了美国单方面决定AI技术访问权的合法性。
AI is not some force of nature or hyperrational, ineffable entity. Instead, he reminds us, AI is ultimately another commercial product, one emerging at a point in history when excessive power over commerce and the wider society has amassed in a vanishingly small number of hands.
Demystifying AI as 'another commercial product' is a counter-narrative to both the techno-utopian and techno-dystopian frames that dominate public discourse. By locating AI within existing structures of capital concentration, the encyclical sidesteps the AGI debate entirely and grounds the ethical question in political economy: who owns the technology and who profits from it.
Authoritarianism in all its forms depends on people acting against their own interests, in this case seeing the abdication of agency and giving up of rights as something positive, as almost a relief. Let the leader decide for you! Let the leader, who is an agent of divine providence, decide your destiny! Getting people to believe this is one of the main functions of personality cults that depict the leader as infallible.
Very similar to how Big Tech provides us with convenience while robbing us of our agencies.
But the real power of agents comes when they can work as a team.
尽管人工智能代理的能力在单独工作时已经显现，但作者强调，它们真正的力量在于团队合作，这与通常认为的个体智能体主导的观点相悖。
We document a clear trend where Chinese models overtook their counterparts built in the U.S.
这一发现挑战了美国在AI领域的主导地位叙事，暗示全球AI权力格局正在发生根本性转变。这种转变可能对技术标准制定、数据治理和国际AI合作产生深远影响。
The U.S.-China AI model performance gap has effectively closed.
这一发现具有地缘政治意义，表明AI领域的权力平衡正在发生重大转变。中美之间的技术竞争从美国单方面领先转变为势均力敌，这可能重塑全球AI治理格局和供应链结构，引发新的国际合作与竞争模式。
If Dario is right, then he has access to such a weapon right now, with his own value system to guide it. Others may as well, or may soon follow.
这是一个令人警醒的声明，暗示AI技术的控制权已经从公共部门转移到了私人企业手中。作者暗示Anthropic等公司可能已经掌握了具有战略意义的技术，而他们的价值观将直接影响这些技术的使用方向，这挑战了传统的国家主权概念。
We estimate Google is the largest single owner of AI compute, holding about one quarter of global cumulative capacity as of Q4 2025.
全球 AI 算力的 25% 被一家公司独占——这个数字令人震惊。更值得注意的是这个数字的性质：这是「累积持有量」而非「新增采购量」，意味着 Google 多年来的硬件积累已形成近乎垄断性的算力护城河。在 AI 竞赛被描述为「群雄逐鹿」的叙事下，这个数字揭示了真正的权力集中程度。
[[The Means of Production by Maximilian Kasy]] overview in slides by author https://maxkasy.github.io/home/files/slides/meansofprediction_slides_long.pdf
authoritarian governments might use powerful AI to surveil or repress their citizens in ways that would be extremely difficult to reform or overthrow.
for - progress trap - AI - misuse of power - Palantir?
examine power as an emergent consequence of deployment and incentives, not intent.
Intent def is there too though, much of this is entrenching, and much of it is a power grab (esp US tech at the mo), to get from capital/tech concentration to coopting governance structures
AI is a tech where by design it is not lowering a participation threshold, it positions itself as bigger-than-us, like nuclear reactors, not just anyone can run with it. That only after 3 years we see a budding diy / individual agency angle shows as much. It was only designed to create and entrench power (or transform it to another form), other digital techs originate as challenge to power, this one clearly the opposite. The companies involved fight against things that push towards smaller than us ai tech, like local offline first. E.g. DMA/DSA
For instance, a recent analysis by Epoch AI of the total training cost of AI models estimated that energy was a marginal part of total cost of AI training and experimentation (less than 6% in the case of all 4 frontier AI models analyzed), and a recent analysis by Dwarkesh Patel and Romeo Dean estimated that power generation represents roughly 7% of a datacenter’s cost.
Which paper or article from Romeo Dean and Dwarkesh patel?
At the level of a hyperscale data center cluster, this can translate into requirements of up to 5 and even 10 GW of power, up from 5 MW - a 2,000 fold increase in the span of a decade [4, 11].
misled investors by exploiting the promise and allure of AI technology to build a false narrative about innovation that never existed. This type of deception not only victimizes innocent investors
The crime was misleading investors, not anyone else, which is very telling. The hype around "AI" - and actually hiring remote workers to do the job - and misleading customers/users doesn't matter.
Im Standard stellt Martin Auber mit aktuellen Daten belegt dar, warum der bloße Ausbau der Kapazitäten zur Erzeugung erneuerbarer Energien nicht zu einer Dekabonisierung führen wird. Der Energiebedarf wächst wesentlich schneller als die zur Verfügung stehende erneuerbare Energiepunkt. Durch den KI-Boom wird er noch einmal deutlich gesteigert. https://www.derstandard.at/story/3000000255154/wann-kommt-die-energiewende-oder-kommt-sie-gar-nicht
Oh, compliance moats are definitely real – think of the calls for AI companies to license their training data. AI companies can easily do this – they'll just buy training data from giant media companies – the very same companies that hope to use models to replace creative workers with algorithms. Create a new copyright over training data won't eliminate AI – it'll just confine AI to the largest, best capitalized companies, who will gladly provide tools to corporations hoping to fire their workforces: https://pluralistic.net/2023/02/09/ai-monkeys-paw/#bullied-schoolkids
Concentration of power.
https://web.archive.org/web/20231019053547/https://www.careful.industries/a-thousand-cassandras
"Despite being written 18 months ago, it lays out many of the patterns and behaviours that have led to industry capture of "AI Safety"", co-author Rachel Coldicutt ( et Anna Williams, and Mallory Knodel for Open Society Foundations. )
For Open Society Foundations by 'careful industries' which is a research/consultancy, founded 2019, all UK based. Subscribed 2 authors on M, and blog.
A Thousand Cassandras in Zotero.
One thing that should be learned from the bitter lesson is the great power of general purpose methods, of methods that continue to scale with increased computation even as the available computation becomes very great. The two methods that seem to scale arbitrarily in this way are search and learning
This is a big lesson. As a field, we still have not thoroughly learned it, as we are continuing to make the same kind of mistakes. To see this, and to effectively resist it, we have to understand the appeal of these mistakes. We have to learn the bitter lesson that building in how we think we think does not work in the long run. The bitter lesson is based on the historical observations that 1) AI researchers have often tried to build knowledge into their agents, 2) this always helps in the short term, and is personally satisfying to the researcher, but 3) in the long run it plateaus and even inhibits further progress, and 4) breakthrough progress eventually arrives by an opposing approach based on scaling computation by search and learning. The eventual success is tinged with bitterness, and often incompletely digested, because it is success over a favored, human-centric approach.