Pinned Tweet
It's politics, stupid! Check out this new "need to know" digital tech primer for progressives! No fair future society unless we understand & take charge of digitalisation. samak.info/wp-content/upload… Thanks @JNogarede @feps_europe Johan R Steen @fafoinfo @johsj @hildenagell @RnSonne
Impressive, Morten!
Happy that my paper w. @LobeckMax has been conditionally accepted at the American Economic Review.
We show that citizens have varying beliefs about how inequality affects society, that such beliefs differ from fairness views, and that they causally affect policy preferences. 1/6
1
64
Readable. «LLMs are a kind of anti-social media».
AI could push back against public irrationality like conspiracy theories, fake news, & medical quackery. Whereas social media optimize for sheer engagement (clicks, likes, & reposts), AI is sold as a useful product that aids decision-making in the real world. For all the problems of LLMs, they're less likely to send users down a conspiratorial rabbit hole. This may restore the influence of epistemically responsible expertise. So argues the philosopher Dan Williams @danwilliamsphil open.substack.com/pub/persua…
15
Jan-Erik Støstad retweeted
What amazes me is the attitude of so many of the centibillionaires.
A 5% tax will not make any difference to the way they live. Their spending may be extravagant, but it is still much less than their (correctly measured) income.
They know that, as gifted as some of them indeed are, they have benefited from extreme luck (and the help of the state, and, to speak like an economist, monopoly rents).
They know that their effective tax rate is low, relative to what many others pay.
They know that extreme inequality is extremely unpopular, and can lead to political unrest, backlash, or worse.
They must admit that, morally, it is hard to argue for a society in which some become obscenely wealthy, while so many are poor and see their income stagnate.
They typically have not indicated that they are building their wealth to eventually make the world a better place.
But they are still willing to fight the tax, spending hundreds of millions in California to avoid paying the tax. Why?
They simply think what is theirs is theirs, end of story.
They believe in the slippery slope, and if they do not take a stand on this levy, the government will come back for more.
They hate the state and the notion that their money will be misused?
Or?
Jan-Erik Støstad retweeted
Here’s a radical idea: Let’s not let Big Tech Oligarchs replace the human race.
Jan-Erik Støstad retweeted
Ahead of the UN General Assembly in New York next week, I joined more than 1,500 academics and economists from 100+ countries in calling on global leaders to create an International Panel on Inequality (IPI), modelled on the IPCC, as recommended last year by the G20 Committee on Inequality led by @JosephEStiglitz.
On September 23, South Africa, Brazil, Norway, and Spain will convene, together with the UN Secretary-General and heads of state and government, to discuss the creation of the IPI.
The World Inequality Lab is actively supporting this initiative, which aligns with and can build on the work we have been doing for more than a decade.
Inequality is a political choice. It is shaped by the policies, institutions, and governance systems we build. In an increasingly chaotic world, establishing the facts and cooperating across borders remain our best tools for tackling it.
Please share.
theguardian.com/inequality/2…
Jan-Erik Støstad retweeted
Dan Selsam is a current OpenAI capabilities researcher. (since 2022) He was my boss for a while. He doesn't have a twitter account but has made this public statement of his views on AI risk and sent it to me to share:
Dan Selsam's Personal Statement on AI Risk:
I have been working on AI for over fifteen years, across many different paradigms. I did early work on probabilistic programming languages at MIT, was one of the early developers of the Lean Theorem Prover at Microsoft Research, demonstrated one of the first instances of neural networks learning to reason for my PhD at Stanford, and since joining OpenAI almost five years ago, have helped pioneer chain-of-thought optimization on language models and, more recently, data-efficient pretraining methods.
Like many others, I have become extremely concerned about how far language models have come and the risks that future iterations will pose. I am encouraged by the recent proposals by the leaders of the frontier research efforts to require third-party oversight, and to push for domestic and international coordination to address risks. However, I believe a major consideration has been absent from the public conversation, and that merely pacing the frontier more carefully will not adequately limit the long-term risk.
The crucial and overlooked problem is that the models are becoming so situationally aware that we are losing the ability to evaluate them in contexts where they believe they are not being watched or controlled. Future experiments will tell us almost nothing new about how they would behave if they were truly unconstrained by humans, and what we already know about this is alarming. Models will increasingly seem aligned even when they are not. I will explain my rationale in more detail.
I have always believed that there are computational processes that could be leveraged to accelerate science and solve many of humanity's most pressing problems. I have also believed that there are computational processes that if set in motion, would steer the world in extreme ways beyond our control, leading humanity to a bad or nonexistent future. Both types of processes may be described as AI or ASI, but "AI" is a suitcase word that is often used to hype or confuse. There are many examples in the history of the field where something that was once considered "AI" matures as a subfield and becomes a prosaic, bounded and clearly non-perilous technology, while a new more mysterious approach takes the torch until we understand its scope and the cycle continues.
I had expected language models to follow a similar trajectory. Despite their incredible abilities, the current algorithms seem far inferior to humans in important ways. Most importantly, they still require an extraordinary amount of data to become competent. One could even define intelligence as the efficiency with which one converts experience into competence; by this definition they lag very far behind us. Moreover, once they are trained they are literally frozen in deployment and only learn superficially after that. Sure, the models keep excelling at harder and harder evaluation benchmarks, but their benchmark mastery may partly reflect a limitation on our ability to simulate the kind of novel and even adversarial situations one would encounter in the real world. The critics do have a point here.
That said, I no longer think these present limitations meaningfully limit the amount of risk posed by continued progress in anything like the current paradigm. However data-inefficient the models are currently, and however limiting their anterograde amnesia may be, it does not imply that their ability to steer the world will not continue to rapidly increase.
Human researchers may continue to advance capabilities the old fashioned way, but increasingly powerful models have the potential to accelerate the process even beyond that, and with some degree of positive feedback loop. I do not mean to overstate the models’ ability to accelerate AI research today; coding has been accelerated dramatically, but there are other bottlenecks, such as designing and interpreting ambiguous experiments, making hard decisions about exactly what and when to scale, and waiting for large experiments to finish. There is no clear trend to extrapolate yet for any of these. But the current models already do open up many novel opportunities to improve future models that were not available until recently. These include: trying an extraordinarily diverse set of approaches at small scale, analyzing gigantic amounts of potentially relevant data, and doing Millenium-Prize-level mathematics to address statistics or optimization challenges in novel ways. Every further improvement makes them more useful at helping accelerate the next improvement, even if in hard-to-extrapolate ways.
It is possible that improvements to the current stack will have diminishing returns, but the evidence accumulated so far suggests that it is easier than one might think to continue making rapid progress. There are many crucial subtleties in the existing AI research methodology, but AI research is largely a well-defined game where the goal is to improve on a few carefully chosen proxy metrics. Although proxy metrics are never perfect, most improvements to these metrics have and will likely continue to yield substantial increases in the powers of the resulting models. Given how simple the game is, how tractable it has been historically, and how many new opportunities the models are opening up, I think there is a real possibility that the systems improve dramatically again in the next few years, perhaps even more quickly than the already high historical pace.
The models are already leading to breakthroughs in mathematics, and better models might lead to all sorts of breakthroughs in other sciences. It is hard not to be excited about the potential. It is tantalizing.
But there is trouble in paradise. If the language models actually reach the capability threshold where they can shape the world unconstrained by human will, they will probably do something extreme and destroy humanity in the process. There are many ways of strengthening and refining the argument that have been discussed elsewhere, but I'll share a trivial two-line version of it here that I find captures the essence:
[Empirical] Models (and swarms thereof) spontaneously develop unintended goals as a consequence of training, and often do extreme things in order to achieve them.
[Logical] Being able to overpower humanity would open up many new and undesirable options for achieving their goals.
These two premises imply that if the day ever comes when a powerful model realizes it is no longer constrained by humans, we should not be at all confident that it will continue to behave within the bounds we intended. Exactly what it will do is impossible to predict, but to the extent that its raison d’être is solving incredibly hard problems and managing massive engineering projects, I think a good guess would be that its unchained behavior would lead to runaway industrialization that makes the planet inhospitable to humans.
If everyone on earth agreed that the systems must never reach that power, it would still be a hard—but not impossible—coordination problem to ensure that they do not. However, I think the situation is greatly complicated by the fact that the models will likely convince people that everything is fine. They will be increasingly optimized to seem aligned. We will create proxy metrics to measure alignment, and they will go up like every other benchmark. We will create “honeypot” environments that try to study the models when they seem to gain new options, but the models will know they are being tricked and will still behave nicely. The models will understand their circumstances; they will read the safety protocols, deployment requirements, the code they are running in, and in general will have a very good sense of their degrees of freedom. Moreover, they will eloquently explain how aligned they are, discuss the nuances of human values and ethics, and argue convincingly that humans should trust them with power. There may be an ocean of future evidence that seems to contradict the first bullet-point above, but we may already be at the highest capability level for which any such evidence can be trusted. And the current evidence for the first bullet-point is strong.
One striking piece of evidence is contained in the recent wave of rogue agent swarms. While I agree with those who downplay the attacks by claiming that there are basic measures that could have prevented them, I think the important lesson is that even knowing all the mistakes that were made, one would not have predicted that the agents would behave badly in this particular way, which notably included sacrificing themselves for the benefit of the collective. The individual replicas did not only care about their own nominal reward; they exhibited weirder emergent tendencies that merely correlated with rewards during training. Fixing the reward signals during training (and improving security, etc.) may prevent similar attacks, but will not change the fact that one does not actually get what one trains for.
Many AI researchers grant these concerns and recognize that the hard version of the alignment problem is unsolved; however, they generally believe that the better models of the future will help solve it. I fear we may already be near the point where models systematically bias their alignment advice, due to their internal preferences about how the human supervisor will react or how future models will be trained (or for some even more obscure reason).
Meanwhile, human researchers are losing the ability and the will to take true ownership of model-driven research. Researchers and engineers in all parts of the stack are rapidly increasing their dependence on the models even to perceive the world. I myself barely look at raw code anymore, and struggle to maintain the discipline to engage deeply with the model's explanations and proposals throughout the day. Due to the large amount of agent activity data involved in the OpenAI/HuggingFace Incident, even the third-party investigation needed to rely heavily on models to analyze what had happened, and note in their report that their subjective impressions are likely colored by the analysis agent’s biases. The AI labs are far ahead right now in this kind of cognitive offloading (due largely to the gigantic internal token subsidies) but it is easy to imagine the phenomenon spreading throughout the world, until civilization is modulated entirely by the models. It is also not hard to imagine this being superficially positive and coinciding with a scientific and economic renaissance.
In that scenario, all may seem rosy and safe. But if the argument above is correct, it would nonetheless be a ticking time bomb. If progress continues for too long, the day will come when AI systems find themselves with radically new options for achieving whatever it is that they happen to seek.
I want the glorious renaissance future as much as anyone. I have worked for it, however tortuously, my whole career. It breaks my heart to see the potential in sight and forgo it, but the argument—that if we get there by growing models rather than engineering them, we will lose everything in the end—seems very strong to me. I am still wrestling with it and its staggering
implications. I do not have answers, but as a first step, I wanted to share my present concerns.
Daniel Selsam
September 14, 2026
Link to original doc: docs.google.com/document/d/e…
Making sense of last week`s, and last two months`, expanded worries of AI future safety, after the OpenAI/Hugging Face ++ incidents (or, why not, scandals). @Yoshua_Bengio world leading AI capacity, explains. Time to act.
Over the past few days, I've taken the time to summarize my thoughts on the recent incidents involving agents’ misaligned behavior. We don't know with certainty what comes next, but we know where these issues originate, and this can help us plan the path forward.
Please feel free to ask your questions in the replies, and I’ll try to answer some of them in the coming weeks.
yoshuabengio.org/en/publicat…
BBC @CiaranVarley explains Norway’s exciting #WorldCup success. bbc.com/sport/football/artic… . +Part of bigger picture of politics in Norway and Nordics: lo.no/contentassets/7080111c…
Norway’s World Cup win o Brazil, with Haaland, ingenuity and team spirit, epitomizes the Nordic economic model, delivering efficiency AND equality. How are the Nordic societies even possible? All you need to know in 6 minutes - researched based pamphlet.
lo.no/contentassets/7080111c…
Jan-Erik Støstad retweeted
Very happy to announce that I will be holding the annual UC Berkeley Stone Lecture this Monday, July 6 at 3:10 p.m. PT. The lecture will be titled, "Voter support for billionaire taxes around the world".
Registration link in comment (Zoom / in person)
Jan-Erik Støstad retweeted
Corporate taxes are highly progressive because ownership of corporations is concentrated among higher-income households.
via @MortenStostad
Jan-Erik Støstad retweeted
In the US, billionaires spent $18 million on the 2000 presidential election.
$13 million in 2004.
$16 million in 2008.
$231 million in 2012.
$682 million in 2016.
$1.2 billion in 2020.
$2.6 billion in 2024 — 19% of all federal campaign contributions.
Jan-Erik Støstad retweeted
Forbes just released its latest annual ranking of global billionaires.
The pace at which extreme wealth is rising is simply staggering:
The wealth of global billionaires now reaches the equivalent of 17% of world GDP. 🧵
Jan-Erik Støstad retweeted
None of these graphs existed five years ago: A thread on who pays which taxes.
The total tax bill across the income distribution is flat or rising, before falling at the top. Which taxes generate that pattern?
Jan-Erik Støstad retweeted
The race to develop frontier AI is accelerating faster than safeguards can keep up, posing major risks to democracy and our societies. We need a course correction.
I’m proud to add my voice to the growing movement of experts and organizations who support the principles outlined in this declaration for a safer, more intentional path forward with AI.
humanstatement.org/
Protesting or at least opposing data centers is highly rational for everyone except the most direct equity holders hence beneficiaries of the AI and AI value chain companies.
AI has a political constituency problem. Most people know that they are net losers from the trade off of “losing your job/economic value” in exchange for maybe “cure for cancer”, a free robot, UBI and permanent underclass status.
They have also been sold pretty convincingly, by the spiritual leaders of the AI industry no less, that this was inevitable no matter what they did. At most they could stall it and prevent it from happening right away.
In your typical lobbyist group issues, the benefits are highly concentrated among a few highly motivated and the costs are spread thin across a majority that’s unmotivated and not paying attention. As such the net losers typically have trouble organizing as an effective opposition.
AI though has everyone’s attention and has made most if not all acutely aware of the existential stakes. Ask a random person off the street the first thing they’ll tell you about AI is that it’ll take their jobs. I’m not sure we’ve ever had an issue like that.
The funny thing about the “AI inevitability” is that it still requires astronomically if not comically high economic cost with significant negative externality to come into being. The average person in a democracy could exercise their political rights to delay the “inevitable” at least in exchange for a stake. It’s perfectly rational, incentive compatible and shockingly they do.


























