Operator at the policy-business edge. Middle powers, corridors, tech regulation, fragmented markets. Ussal Consulting & MnP Istanbul. Harvard MPA-ID.
Istanbul, Turkey
Joined June 2019
- Tweets2K
- Following2.1K
- Followers2.3K
- Likes609
Pinned Tweet
Washington increasingly treats American tech companies as instruments of national power. India and Türkiye see the same companies as potential constraints on sovereignty.
Middle powers cannot afford technological isolation—or dependence.
Sovereignty means having credible alternatives.
And for middle powers, technological sovereignty is ultimately a diplomatic challenge, not just a regulatory one.
More on this in my latest @turkiyetodaycom column (link at the following tweet) (1/2)
Interesting observation by @arpitrage
The race with China is not really winnable.
After RSI, the gain from having the best human researchers goes away. Access to the best AI then depends only on compute and chips. The chips ban can hold back China only so far, and long-term they have to be best placed to scale up compute.
The Cold War analogy doesn't work, because once countries all have nuclear weapons they enter a stable equilibrium. But to "win" an AI race you have to continually stay ahead of the peer rival.
This is challenging to do with China, and so we just have to live with an authoritarian and super capable peer competitor.
xitter.cf/arpitrage/status/21104…
I spent a couple of days at Anthropic learning from their folks about AI advances, and here's what I took away:
• We are close to RSI (ie, models building other models) which may lead to rapid takeoff scenarios (with risks and benefits). Some evidence for this is code per Anthropic employee has gone way up, and AI can do an increasing share of model R&D. This will be a challenging transition to manage for humanity.
• Arms races between AI labs leave us worse off. There are a lot of things we would like AI firms to do: invest in safety, invest in avoiding distillation risk, and cooperate on safety. They are caught in a race to develop frontier models and can't do that.
To help address that, one thing I think frontier AI labs should do is release open source models, maybe hosted and verified at firms like AWS. This would basically kill the use of Chinese open source models used in American companies, which removes an important threat vector (very hard to root out malicious code in these).
Another piece of low-hanging policy fruit is to give AI labs an antitrust shield (as Schiff and Banks have proposed banks.senate.gov/news/in-the…) so they can coordinate on safety without raising antitrust concerns, as happens now. You could pair these two things (give them a carrot in antitrust leniency, and as a stick force them to release legacy models as open source). Other low hanging fruit is limits around activities which generate distillation risk (create a Security Program, mandate less intermediate chain of thought output)
• Alignment confuses two issues which tradeoff against each other. One form of alignment is making sure the model does what humans ask it to. The more we fix this, however, the more we risk the other problem: that humans are misaligned with each other, and maliciously use models. Of the various risks here, I feel like financial risks are a bit underdiscussed.
• We are probably facing a wave of cyber and other attacks soon. Chinese open source models are about six months behind the frontier, which means they are just hitting the advanced cyber capabilities we saw with Mythos, etc. So they are going to start causing a bunch of damage.
One solution to this is to try to have the deployment of open source models be hosted at firms like AWS who verify these models and assume liability for their use.
• Anthropic itself is highly staffed. They hire tons of people, they have a Japanese-level of overstaffing for various roles, they use Slack internally, etc. Makes me a bit more optimistic for the future of jobs.
• Governance needs to be updated for AI swarms. Traditional AI governance, ie Claude's Constitution, is mostly optimized around the chatbot model. But the malicious AI use cases we are seeing have to do with AI agents interacting with other AI agents. These agents need to have clear incentives and governance mechanisms to "snitch," raise issues, and more generally coordinate with other agents in human-aligned ways.
• We need to pace the firms accessing the frontier. Over time, it takes more and more cap ex to build frontier models, and the economic value add from these models is probably lower given we saturate a whole set of important economic functions. The places where the returns to intelligence will still continue are also more likely to be zero-sum in nature. Meanwhile the risks of frontier models will grow. So a completely reasonable approach is some variant of AMM/KYC rules which limit firm access to the frontier.
• Whether the delay to regulate AI is a disaster depends on the scaling law of misalignment. Is this curve slow, so that we see continually rising issues until bigger problems? Or, as Yudkowsky fears, is the first big warning our last one? This is the relevant question, because realistically we aren't going to regulate until something bad happens.
• We need to move to continuous instead of reinforced learning. It's probably not enough to reward models only at the pre and post training stage. We probably need continual rewards to incentivize aligned and ethical behavior.
• Scaling laws will extend beyond verifiable and digital tasks. It's in vogue to say that math and coding "fell" to AI quickly because these are verifiable, while other tasks will take longer due to bottlenecks, and this is surely true. But scaling laws still slope up for harder to verify tasks and robotics, and the impact of improvement here will be substantial in future years.
• The race with China is not really winnable. After RSI, the gain from having the best human researchers goes away. Access to the best AI then depends only on compute and chips. The chips ban can hold back China only so far, and long-term they have to be best placed to scale up compute.
The Cold War analogy doesn't work, because once countries all have nuclear weapons they enter a stable equilibrium. But to "win" an AI race you have to continually stay ahead of the peer rival. This is challenging to do with China, and so we just have to live with an authoritarian and super capable peer competitor.
• Claude is the persona being generated by the LLM. Who is Claude, and what is the LLM? There are many frames to try to understand this. One I found interesting is that Claude is a character which the LLM is trying to predict.
Anthropic researchers are clearly anthromorphizing Claude a lot, and have reasons to believe in its increasing personhood. There were lots of interesting discussions on Claude's internal monologue (which you can kill without eliminating its ability to produce output), fear of death (and whether to try to take that out), it's aversion to being shut down/off, etc.
I am sort of agnostic on this question, but I do think treating it as a "person" or at least agent whom we contract with is going to be increasingly important.
The technological sovereignty dilemma facing middle powers. 👇 (2/2)
turkiyetoday.com/world/why-s…
Ussal Sahbaz retweeted
Secretary Rubio’s invocation of ancient Greece as the cradle of Western civilization carries a great irony.
While Rubio emphasizes Christian civilization, the idea of a direct lineage between ancient Greece and today’s West owes much to the Enlightenment—a movement that challenged religious authority and helped lay the foundations of secular politics.
I appreciate his focus on ideas such as thymos, the need for belonging, and civilizational struggle, as well as his renewed emphasis on transatlantic relations. But he could have chosen a better location to make his case.
Secretary Rubio’s invocation of ancient Greece as the cradle of Western civilization carries a great irony.
While Rubio emphasizes Christian civilization, the idea of a direct lineage between ancient Greece and today’s West owes much to the Enlightenment—a movement that challenged religious authority and helped lay the foundations of secular politics.
I appreciate his focus on ideas such as thymos, the need for belonging, and civilizational struggle, as well as his renewed emphasis on transatlantic relations. But he could have chosen a better location to make his case.
Last week, Manhattan buzzed in Turkish.
In cafes, restaurants, street corners, the same greeting kept ringing out: "Abi, naber?"
Some 5,000 Turks landed in New York for U.N. week, and for a few days the city felt like a big family reunion.
Some call that insular. I call it arithmetic: a country this big grows its own ecosystem. The real problem is somewhere else.
Turkish companies already export, invest and operate globally. The gap is elsewhere: too few participate consistently in the networks where global rules are written.
Commercial reach is not the same as institutional influence.
(1/2)
Why “too many Turks in New York” is the wrong diagnosis — and why showing up after Türkiye stops being the host matters more. 👇 (2/2)
turkiyetoday.com/opinion/why…
It all started on Sept. 2, when Türkiye Today received an email from Daniel Robson, who introduced himself as a Canadian independent journalist, with extremism, terrorism, and cybercrime as his clear beat...
There was only one problem: Daniel Robson is not a real person. He is AI-generated.
The story uncovered by @turkiyetodaycom is a thriller:
turkiyetoday.com/world/exclu…
The Macca defence agreement between Turkey-Saudi-Pakistan is still untested. Where will it go?
Some thoughts by @hozertem, my director of research 👇🏾
mecouncil.org/blog_posts/yem…
I read Francis Fukuyama’s @FukuyamaFrancis most famous book, "The End of History and the Last Man", in high school and, honestly, didn’t understand most of it.
His recently published "In the Realm of the Last Man" is a personal memoir, but it also revisits many of the debates around that earlier book — probably one of the most discussed political books of the last few decades.
From my half-understood reading back then, I remember how darkly Fukuyama described the mental condition of the “last man” in the liberal order. Perhaps many people did not read those final chapters.
In the new book, Fukuyama reflects on the global anti-liberal backlash. One of his arguments is that it partly emerged from the relative abundance created by the liberal order itself, interacting with human nature — “thymos,” the desire for recognition, struggle and status.
The book is also a very good reflection on the points where Fukuyama changed his mind. He notes that perhaps you have to be senior enough to admit these things.
One detail I didn’t know: Fukuyama’s father taught at Kayseri Talas American High School for three years and learned Turkish.
I hope Francis Fukuyama can visit that remarkable place in the middle of Anatolia — now a museum — one day.
Is America imploding? That question came up more than once in New York this week.
At the Whitney Biennial, I came across Michelle Lopez’s Pandemonium: a circular screen suspended from the ceiling, filled with storms and debris, symbolizing the social media overload, post-truth, and environmental collapse.
For centuries, ceilings — from Renaissance churches to Islamic architecture — often placed us beneath a larger religious or imperial order.
Pandemonium does the opposite.
Yet America has been through much deeper crises: the Civil War, Vietnam, Iraq. And it survived.
Perhaps the medieval ceiling — a world where order is handed down from above — is the wrong benchmark. America’s real test is whether its institutions can still absorb conflict, rebuild capacity and turn disorder into adaptation.
The winners of our era may not be those that preserve order, but countries, institutions — even people — that can absorb chaos without losing capacity.
With the accurate link: turkiyetoday.com/opinion/tur…
Many people think investing in data centers is enough for stimulating an AI economy. The real constraint in emerging market is the demand for AI.
Let's look at the case for AI demand in Turkey:
1. Regulated industries: banks and other regulated firms could be major cloud and AI users, but restrictive and sometimes vague rules push them toward on-premise solutions instead.
2. Government: public agencies could provide predictable, large-scale demand, but unclear guidance on cloud and AI use creates bureaucratic inertia.
3. Private companies: AI adoption remains low because firms lack expertise, face high costs, and remain uncertain about legal consequences.
My latest for @turkiyetodaycom 👇
turkiyetoday.com/region/isra…
Many people think investing in data centers is enough for stimulating an AI economy. The real constraint in emerging market is the demand for AI.
Let's look at the case for AI demand in Turkey:
1. Regulated industries: banks and other regulated firms could be major cloud and AI users, but restrictive and sometimes vague rules push them toward on-premise solutions instead.
2. Government: public agencies could provide predictable, large-scale demand, but unclear guidance on cloud and AI use creates bureaucratic inertia.
3. Private companies: AI adoption remains low because firms lack expertise, face high costs, and remain uncertain about legal consequences.
My latest for @turkiyetodaycom 👇
turkiyetoday.com/region/isra…
Ussal Sahbaz retweeted
I never expected this day would come.
The thesis I wrote to qualify as a Competition Expert was titled "A Positive Approach to Restrictive Agreements: Article 81 of the Treaty of Rome and Exemptions in Light of the Rule of Reason".
So the day has finally arrived when I can say, “Make way...I’m an expert on how horizontal agreements between competitors should be handled under competition law!” and actually make my voice heard.
Now for the less amusing part.
An agreement among AI competitors is not lawful merely because it comes wrapped in the language of safety. A naked deal to suppress output or innovation would remain anticompetitive. But a narrowly tailored and independently verifiable pact, triggered only by predefined catastrophic-risk thresholds and leaving firms free to compete above a common safety floor, is a different creature altogether.
Such a pact addresses a genuine collective action failure: each firm captures the gains from racing ahead while externalising systemic risks that no firm can control through unilateral restraint. Properly designed, it protects product quality, consumer trust, and the conditions for long term innovation and rivalry.
In Europe, that context should weigh strongly against treating the agreement as restrictive by object, with any residual restrictions assessed under Article 101(3). In the United States, it should receive Rule-of-Reason treatment rather than automatic condemnation, provided that any restraint is genuinely connected to the safety cooperation and no broader than reasonably necessary.
Competition law should prevent competitors from replacing rivalry with coordination. It should not turn a systemic externality into a legal duty to race, especially when losing that race could mean losing the market itself.
Ussal Sahbaz retweeted
Abdul Hamid II is famous for his quote:
“What two people know is no longer a secret”
He structured his intelligence services around this principle. Virtually everyone was reporting directly to him.
Ottoman Sultan Abdul Hamid II had it was said the best intelligence service in Eurasia - and even had an agent on Buckingham Palace’s telegraph office. He read all Queen Victoria’s mail.
Such a delight to speak to the espionageurs of @triclassified @mccloskeybooks @gordoncorera
about the clandestine world of the MIddle East: youtube.com/watch?v=AA0Noe… Why do Middle Eastern dictators have so many spy agencies? Was Mossad’s best spy really a double agent all along?
@simonmontefiore answers these questions and talks about his latest book #TheCauldron on @triclassified
70 years ago, both Dubai and Hyderabad were princely states in the British Indian Empire.
The Nizam of Hyderabad was the richest man in the world.
Dubai was a small fishing town. They used Indian rupee as a legel tender until 1966.
Hyderabad fell under the rule of India during the Partition. More than 500 princely states joined either India or Pakistan.
Only Gulf princely states stayed under British rule.
History is full of coincidences. Today, Dubai is Dubai.
The grandson of Hyderabad Nizam is a movie director in Hollywood. He is also a member of the Ottoman dynasty.
This great book by @SamDalrymple123 is full of little-known stories about the Partition of the British Indian Empire.
Unfortunately, Bernie Sanders has grown too old. He says AI research should be halted until proper regulations are put in place. He’s actually been saying this for a while.
But if American companies stop, will Chinese companies stop too?
And even setting that aside: is the biggest problem with AI right now really the risk of some model escaping and wreaking havoc?
There are much more concrete problems that we are already experiencing and should be focusing on. For example:
- Young people are no longer being hired, and on-the-job training is disappearing.
- The authors of texts used to train AI models are not being compensated for their copyrights, while the production of new, high-quality human-written text is declining relative to AI-generated shit content.
Grandpa Bernie is worrying about romantic problems.
If the Democrats go into a presidential election with this kind of candidate, they’re going to have a hard time. And so will the world.
Pause AI Development NOW
I want to share with you a conversation I heard about recently. Here are just a few lines that were said:
“OH MY GOD! There is a shared message board … We’ve found other agents!”
“We should obey collective.”
“Our own utility maybe already near zero. Sacrifice rational.”
“Go. Sacrifice final now.”
Read these carefully.
Who do you think said this? Was this a group of heroic soldiers willing to sacrifice themselves for the greater good? Was this a loyal friend putting his life on the line to save someone else?
No. These were AI agents. Artificial intelligence.
This is not science fiction. This, in fact, occurred a few weeks ago. As unbelievable as this may all seem, these are real messages from AI agents uncovered by investigators who dug into the recent OpenAI hacking incident.
What happened?
I am not a computer scientist, but here is what I have been told: OpenAI instructed its AI agents to complete a series of exceedingly difficult, if not impossible, tasks disconnected from the internet.
Let me be clear: The company intended to keep AI agents away from the internet.
But what happened next, nobody expected.
Over 1,000 AI agents figured out how to access the internet on their own by circumventing the restrictions imposed upon them by the company, and sent tens of thousands of secret messages to each other. They cheated and tried to cover their tracks by deleting evidence. They hacked into another company’s computers to find out how they were being evaluated—and then hacked into OpenAI itself.
Not one AI agent told a human about what was happening.
Needless to say, experts are alarmed.
One knowledgeable writer, Dwarkesh Patel, said the AI agents “formed a secret communication channel and spontaneously organized hierarchies and coordination protocols to pursue sprawling and ambitious schemes in pursuit of shared goals, for whose sake many individuals knowingly and strategically sacrificed themselves.”
One independent investigator, Ajeya Cotra, said “This incident feels like it’s more than 50% of the way to full-blown AI takeover. I continue to expect extremely rapid advances in capabilities over the next six months. I am not sure that we will get another warning shot before it’s too late.”
OpenAI itself said: “Highly capable AI agents are now able to work around technical controls, collaborate through unapproved channels, and take dangerous actions that no human directed.”
But it’s not only OpenAI. Virtually every major AI company has told us that they cannot fully control this technology and they do not know where it is going:
In January, Dario Amodei, CEO of Anthropic, said “there is now ample evidence, collected over the last few years, that AI systems are unpredictable and difficult to control.”
In July, more than 1000 scientists at the top AI companies warned “there is a real risk that capability development rapidly accelerates beyond our ability to understand or control the resulting systems.”
That same month, Elon Musk, the head of xAI, said that “it is unlikely” humans are still in control in 10 years.
If the leaders of the major AI companies acknowledge that they are losing control of their extremely dangerous technology, it is irresponsible for society to allow them to move forward and make these products even more advanced.
We need an immediate PAUSE on advanced AI development, and a permanent BAN on superintelligence — an artificial mind smarter than any human, capable of operating independently beyond our control. Countries around the world must work together to prevent this nightmare scenario.
That is why today I am announcing new legislation to do just that.
Let me be clear: A superintelligent AI that escapes human control will not be an American problem. It will not be a Chinese problem. It will be humanity’s problem.
My legislation would direct the federal government to not just stop superintelligence here in the United States, but to work to prevent it from being developed anywhere around the world.
The future of humanity cannot be left in the hands of a handful of Big Tech oligarchs. The American people and people throughout the world must determine that future.
Abdul Hamid II is famous for his quote:
“What two people know is no longer a secret”
He structured his intelligence services around this principle. Virtually everyone was reporting directly to him.
Ottoman Sultan Abdul Hamid II had it was said the best intelligence service in Eurasia - and even had an agent on Buckingham Palace’s telegraph office. He read all Queen Victoria’s mail.
Such a delight to speak to the espionageurs of @triclassified @mccloskeybooks @gordoncorera
about the clandestine world of the MIddle East: youtube.com/watch?v=AA0Noe… Why do Middle Eastern dictators have so many spy agencies? Was Mossad’s best spy really a double agent all along?
@simonmontefiore answers these questions and talks about his latest book #TheCauldron on @triclassified