TheChek News Flow

Israel arrests 7 Israelis for allegedly spying on military sites for Iran

Authorities say group gathered intelligence on air base, air defences and power plant over 2 years

Israel said on Monday it had arrested seven Israeli citizens after uncovering a long-standing operation to spy on sensitive military and energy installations on behalf of Iran.

The arrests of the group — who Israeli authorities alleged carried out reconnaissance for more than two years in exchange for hundreds of thousands of dollars in cryptocurrency — came after Iran and Tehran-backed Hizbollah attacked Israeli military sites in recent months.

The seven Jewish Israelis, including two minors, carried out “hundreds” of intelligence-gathering missions for Iran, including photographing sensitive military installations such as the Nevatim air base, Iron Dome air defence batteries and the Hadera power plant, said Israeli authorities. They also gathered information on naval bases and other targets, authorities said.

Some of the sites have been targeted in missile attacks by Iran or the Lebanon-based Hizbollah militant group in recent months.

Iran has not so far responded to the latest spying claims.

The suspects are expected to be charged soon with major security offences, said police and the Shin Bet internal security agency. Israeli authorities alleged they were directed by two Iranian agents code-named “Alkhan” and “Orkhan”.

Israeli police said the alleged offences were “among the most serious that the state of Israel has known. The suspects acted while knowing about their actions and out of greed for money — and harmed the State of Israel and its citizens,” an official from the force said in a statement.

Israel has been battling Iran and its various proxy militias across the Middle East for more than a year, since Hamas’s cross-border assault from Gaza on October 7 2023 sparked the ongoing conflict.

Iran and Israel exchanged direct fire for the first time in their history last April, and again earlier this month when Tehran fired nearly 200 ballistic missiles at Israel. Israeli leaders have vowed a “severe” response against Iran.

Members of the suspected spy network “were aware that the intelligence they provided compromised national security and could potentially aid enemy missile attacks”, the Israel police and Shin Bet said in a statement.

Several of the suspects were arrested near the home of an “Israeli citizen” they were said to be monitoring, raising suspicions that Iran had also made prominent Israelis potential targets.

Israel’s intelligence services have been suspected of conducting espionage, sabotage and assassinations on Iranian soil for years, including the theft of the country’s nuclear archives in 2018 and the killing of chief Iranian nuclear scientist Mohsen Fakhrizadeh in 2020.

More recently, Iran accused Israel of carrying out a bombing in July at an Iranian military guesthouse in Tehran that killed Hamas leader Ismail Haniyeh.

Iranian efforts to recruit Israeli citizens are an ongoing concern for Israel’s intelligence services, including in recent weeks.

Last month a 72-year-old Israeli businessman was indicted on espionage charges after Shin Bet said he had travelled to Iran. Last week, an Israeli couple was arrested over police allegations that they had been recruited by Iranian intelligence.

However, the latest arrests point to a network of alleged spies that was unusual both for its scale and for the long period in which it was said to have operated.

“This occurred amidst ongoing conflicts in multiple arenas. It is assessed that these activities have inflicted security damage on the state,” said a Shin Bet official.

Google DeepMind’s Demis Hassabis on his Nobel Prize: ‘It feels like a watershed moment for AI’ 

Founder of the artificial intelligence R&D lab says scientific understanding can prevent mis-steps in developing the technology

In the 15 years since it was founded, Google DeepMind has grown into one of the world’s foremost artificial intelligence research and development labs. In October, its chief executive and co-founder Sir Demis Hassabis was one of three joint recipients of this year’s Nobel Prize in chemistry for unlocking a 50-year-old problem: predicting the structure of every known protein using AI software known as AlphaFold.

DeepMind, which was acquired by Google in 2014, was founded with the mission of “solving” intelligence — designing artificial intelligence systems that could mimic and even supersede human cognitive capabilities. In recent years, the technology has become increasingly powerful and ubiquitous and is now embedded in industries ranging from healthcare and education to financial and government services.

Last year, the London-based lab merged with Google Brain, the tech giant’s own AI lab headquartered in California, to take on stiff competition from its peers in the tech industry, in the race to create powerful AI.

DeepMind’s new positioning at the centre of Google’s AI development was spurred by OpenAI’s ChatGPT, the Microsoft-backed group’s chatbot that provides plausible and nuanced text responses to questions. Despite its commercial underpinnings, Google DeepMind has remained focused on complex and fundamental problems in science and engineering, making it one of the most consequential projects in AI globally. 

In the first interview of our new AI Exchange series, Hassabis — a child chess prodigy, designer of cult video game Theme Park, and a trained neuroscientist — spoke to FT’s Madhumita Murgia just 24 hours after being announced as a Nobel Prize winner. He talked extensively about the big puzzles he wants to crack next, the role of AI in scientific progress, his views on the path to artificial general intelligence — and what will happen when we get there. 

Madhumita Murgia: Having reflected on your Nobel Prize for a day, how are you feeling?

Demis Hassabis: To be honest with you, yesterday was just a blur and my mind was completely frazzled which hardly ever happens. It was a strange experience, almost like an out-of-body experience. And it still feels pretty surreal today. When I woke up this morning, I was like, is this real or not? It still feels like a dream, to be honest. 

MM: Protein folding is essentially solved, because of your work on AlphaFold models — an AI system that can predict the structure of all known proteins. What is your next grand challenge for AI to crack?

DH: There are several. Firstly, on the biology track — you can see where we are going with that with AlphaFold 3 — the idea is to understand [biological] interactions, and eventually to model a whole pathway. And, then, I want to maybe build a virtual cell at some point. 

With Isomorphic [DeepMind’s drug development spin-off] we are trying to expand into drug discovery — designing chemical compounds, working out where they bind, predicting properties of those compounds, absorption, toxicity and so on. We have great partners [in] Eli Lilly and Novartis . . . working on projects with them, which are going really well. I want to solve some diseases, Madhu. I want us to help cure some diseases.

MM: Do you have any specific diseases you’re interested in tackling?

DH: We do. We are working on six actual drug programmes. I can’t say which areas but they are the big areas of health. I hope we will have something in the clinic in the next couple of years — so, very fast. And, then, obviously, we will have to go through the whole clinical process, but at least the drug discovery part we will have shrunk massively.

MM: What about outside of biology? Are there areas you’re excited about working on?

DH: I’m very excited about our material design work: we published a paper in Nature last year on a tool called GNoME [an AI tool that discovered 2.2mn new crystals]. That’s AlphaFold 1-level of material design. We need to get to AlphaFold-2 level, which we are working on. 

We are going to solve some important conjectures in maths with the help of AI. We got the Olympiad silver medal over the summer. It’s a really hard competition. In the next couple of years, we will solve one of the major conjectures. 

And, then, on energy/climate, you saw our Graphcast weather modelling won a MacRobert award, a big honour on the engineering side. We’re investigating if we can use some of these techniques to help with climate modelling, to do that more accurately, which will be important to help tackle climate change, as well as optimising power grids and so on. 

MM: It seems like your focus is more on the application side — on work that translates into real world impact, rather than purely fundamental. 

DH: That’s probably true to say. There aren’t many challenges like protein folding. I used to call it Fermat’s last theorem of biology, equivalent. There’s not that many things that are that important and long-standing as a challenge. 

Obviously, I’m very focused on advancing artificial general intelligence [AGI] with agent-based systems. Probably, we are going to want to talk about Project Astra and the future of digital assistants, universal digital assistants, which I’m personally working on, as well, and which I consider to be on the path to AGI. 

MM: What does the AI double Nobel Prize in chemistry and physics [this year’s prize for physics went to Geoffrey Hinton and John Hopfield for their work on neural networks, the foundational technology for modern AI systems] say about the technology’s role and impact in science?

DH: It’s interesting. Obviously, nobody knows what the committee was thinking. But it’s hard to escape the idea that maybe it’s a statement the committee is making. It feels like a watershed moment for AI, a recognition that it can actually, is mature enough now, to help with scientific discovery. 

AlphaFold is the best example of that. And Geoff and Hopfield’s prizes were for more fundamental, underlying algorithmic work . . . interesting they decided to put that together, almost as double, related awards. 

For me, I hope we look back in 10 years and AlphaFold will have heralded a new golden era of scientific discovery in all these different domains. I hope that we will be adding to that body of work. I think we’re quite unique as one of the big labs in the world that actually doesn’t just talk about using it for science, but is doing it.

There’s so many cool things going on in academia as well. I was talking to someone in astrophysics, actually a Nobel Prize winner, who is using it to scan the skies for atmospheric signals and so on. It’s perfect. It’s being used in Cern. So maybe the committee wanted to recognise that moment. I think it’s pretty cool they’ve done that. 

MM: Where is your AlphaFold work going to take us next in terms of new discoveries? Have there been any interesting breakthroughs in other labs you’ve seen that you’re excited about?

DH: I was really impressed with the special issue of Science on the nuclear pore complex, one of the biggest proteins in the body, which opens and closes like a gateway to let nutrients in and out of the cell nucleus. Four studies found this structure at the same time. Three out of four papers found AlphaFold predictions [were] a key part of them being able to solve the overall structure. That was fundamental biology understanding. That was the thing that stuck out to me. 

Enzyme design is really interesting. People like [US biochemist and Nobel laureate] Francis Arnold have looked at combining AI with directed [protein] evolution. There’s lots of interesting combinations of things. Lots of top labs have been using it for plants, to see if they can make them better resistant to climate change. Wheat has tens of thousands of proteins. No one had investigated that because it would be experimentally too expensive to do that. It’s helped in all kinds of areas, it’s been wonderful to see. 

MM: I have a conceptual question about scientific endeavour. We originally thought predicting something was the be-all-and-end-all, and spent all this time and effort predicting, say, the structure of a protein. But now we can do that really quickly with machine learning, without understanding the ‘why’. Does that mean we should be pushing ourselves to look for more, as scientists? Does that change how we learn about scientific concepts?

DH: That’s an interesting question. Prediction is partly understanding, in some sense. If you can predict, that can lead to understanding. Now, with these new [AI] systems, they are new artefacts in the world, they don’t fit into normal classification of objects. They have some intrinsic capability themselves, which makes them a unique class of new tool.

My view on that is, if the output is important enough, for example, a protein structure, then that, in itself, is valuable. If a biologist is working on leishmaniasis, it doesn’t matter where they got protein structures from as long as they are correct for them to do their science work on top. Or, if you cure cancer, you’re not going to say: don’t give me that because we don’t understand it. It would be an amazing thing, without understanding it fully. 

Science has a lot of abstraction. The whole of chemistry is like that, right? It’s built on physics, and then biology emerges out of it. But it can be understood in its own abstract layer, without necessarily understanding all the physics below it. You can talk about atoms and chemicals and compounds, without fully understanding everything about quantum mechanics — which we don’t fully understand yet. It’s an abstraction layer. It already exists in science.

And biology, we can study life and still don’t know how life evolved or emerged. We can’t even define it properly. But these are massive fields: biology, chemistry, and physics. So it’s not unusual in a sense — AI is like an abstraction layer. The people building the programs and networks understand this at some physics level but, then, this emergent property comes out of it, in this case, predictions. But you can analyse the predictions on their own at a scientific level.

Having said all of that, I think understanding is very important. Especially as we get closer to AGI. I think it will get a lot better than it is today. AI is an engineering science. That means you have to build the artefact first and then you can study it. It’s different to a natural science, where the phenomenon is already there.

And just because it’s an artificial, engineered artefact doesn’t mean it will be any less complex than the natural phenomena we want to study. So you should expect it to be just as hard to understand and unpack and deconstruct an engineered artefact like a neural network. That’s happening now and we are making some good progress. There is a whole field called mechanistic interpretation, which is all about using neuroscience tools and ideas to analyse these virtual brains. I love this area and have encouraged this at DeepMind.

MM: I looked up a project you mentioned previously about a fruit fly connectome [brain map], made using neural networks. AI helped understand that natural phenomenon. 

DH: Exactly. That’s a perfect example of how these things can be combined, and then we slowly understand more and more about the systems. So, yes, it’s a great question, and I’m very optimistic we will make a lot of progress in the next few years on the understanding of AI systems. And, then, of course, maybe they can also explain themselves. Imagine combining an AlphaFold with a language capability system, and maybe it can explain a little bit about what it’s doing. 

MM: The competitive dynamics in the technology industry have intensified a lot in AI. How do you see that impacting and shaping progress in this field? Are you worried there will be fewer ideas and a focus on transformer-based large language models (LLMs)? 

DH: I think that actually a lot of the leading labs are getting narrower with what they are exploring — scaling transformers. Clearly, they are amazing and going to be a key component of ultimate AGI systems. But we have always been big believers in exploration and innovative research. We have kept our capabilities of doing that — we have by far the broadest and deepest research bench in terms of inventing the next transformer, if that’s what is required. That’s part of our scientific heritage, not just at DeepMind but also Google Brain. We are doubling down on that, as well as obviously matching everyone on engineering and scaling.

One has to do that — partly to see how far that could go, so you know what you need to explore. I’ve always believed in pushing exciting ideas to the maximum as well as exploring new ideas. You don’t know what breakthrough you need until you know the absolute limits of the current ideas.

You saw that with long context windows [a measure of how much text can be processed by an LLM at once]. It was a cool new innovation and no one else has been able to replicate that. That’s just one thing — you’ll see a lot more breakthroughs coming into our mainstream work. 

MM: You and others have said AGI is anywhere between 5 to 20 years away: what does the scientific approach look like to achieving that goal? What happens when we get there?  

DH: The scientific approach would mean focusing a lot more time and energy and thought on understanding and analysis tools, benchmarking, and evaluations. There needs to be 10 times more of that, not just from companies but also AI safety institutes. I think from academia and civil society, [too]. 

I think we need to understand what the systems are doing, the limits of those systems, and how to, then, control and guardrail those systems. Understanding is a big part of scientific method. I think that is missing from pure engineering. Engineering is just seeing — does it work? And, if it doesn’t, you try again. It’s all trial and error.

Science is what can you understand before all that happens. And, ideally, that understanding means you make less mistakes. The reason that’s important for AGI and AI is that it’s such a powerful technology you want to make as few mis-steps as you can. 

Of course, you want to be able to get it perfect, but it’s too new and fast moving. But we can definitely do a better job than, perhaps, we’ve done with past technologies. I think we need to do that with AI. That’s what I’d advocate.

When we get nearer to AGI, maybe a few years out, then a societal question comes, which also could be informed by the scientific method. What values do we want these systems to have? What goals do we want to set them?

So they’re sort of separate things. There’s the technical question of how do you keep the thing on track to the goal that you set? But that doesn’t help you decide what goal this should be, right? But you need both those things to be correct for a safe AGI system. 

The second one, I think, may be harder, like, what goals, what values and so on — because that’s more of a UN or geopolitical question. I think we need a broad discussion on that, with governments, with civil society, and academia, all parts of society — and social science and philosophy, even, as well. 

And I try and engage with all those types of people, but I’m a bit unusual in that sense. I am trying to encourage more people to do that or at least act as a role model, act as a conduit to bring those voices around the table. 

I think we should start now because, even if AGI is 10 years away, and some people think it could be a lot sooner, that’s not a lot of time.

Gapping down
In reaction to earnings/guidance:

Other news:

Gapping up
In reaction to earnings/guidance:

Other news:

At the 1939 World’s Fair in New York, Westinghouse Electric unveiled Elektro, a 7-foot-tall robot that walked, spoke, smoked cigarettes and gestured with its arms. It was all a show—Elektro’s witty conversation flowed from phrases recorded on a vinyl disc inside its massive body. But the robot wowed crowds that piled in to witness its apparent smarts.

Eight and a half decades later, Toyota and Tesla have joined a race to commercialize walking, talking robots that resemble humans, are powered by AI and will handle many mundane factory and household tasks. But in many ways, the current versions of the robots resemble Elektro—they still need a hidden hand.

The upshot: Versatile humanoid robots seem unlikely to be available to mainstream consumers for years, if not decades.

At an event on Oct. 10, Tesla CEO Elon Musk showcased the company’s latest version of Optimus, a humanoid robot he unveiled three years ago. As he spoke, numerous Optimus robots walked into the crowd and danced on a stage, and Musk promised that future versions would have capabilities similar to those of C-3PO, the empathetic, diplomatic and sentient robot in “Star Wars.” Tesla would sell billions of such robots, Musk predicted in July, driving the company’s valuation to tens of trillions of dollars, up from $705 billion last week.

But while the Optimus robots shuffled around on their own at the event, human workers behind the scenes appeared to be driving almost everything else they did, conversing through built-in speakers and directing the robots’ arms and hands as they mixed drinks at a bar, according to guests at the event and AI experts. When guests asked, Optimus robots themselves fessed up that they were operated remotely by humans, according to guest videos posted on X. “Regarding speech, I believe this was entirely remote controlled and done with human voices through some sort of radio,” William Stein, a tech analyst with Truist Securities, told me.

Six days after the Tesla event, the Toyota Research Institute, the Japanese carmaker’s research and development arm, announced a partnership with robot maker Boston Dynamics to develop humanoid robots for use in factories and homes. Gill Pratt, CEO of TRI, told me that new advances in AI may make it possible to train humanoid robots to handle numerous mechanical tasks, even those they haven’t explicitly been taught to do.

But that’s probably years away. Pratt said that, just as the World’s Fair visitors did in the 1930s, people still often have unrealistic expectations of how close researchers are to developing robots with near-human abilities.

“We are anthropomorphizing tremendously onto the [robots],” Pratt said. “And we are kind of fooling ourselves and fooling each other by wishful thinking about what we wish that they could do versus what they actually can do, which is very, very little right now.”

Show, Don’t Tell

Musk has not disclosed the technology behind Optimus apart from saying it shares the same motors, gearboxes, batteries and AI software as Tesla’s self-driving cars, and he did not respond to an email seeking comment about Optimus.

The Toyota–Boston Dynamics team uses a new type of AI system similar to the large language models that power ChatGPT, but instead employs large behavioral models to train robots by physically showing them a task: for example with a video. This approach, called diffusion policy, is much less work than writing mountains of line-by-line instructions in code, said Shuran Song, a robotics professor at Stanford University who collaborates with Pratt.

The field is in a primitive state: Pratt’s team has used diffusion policy to teach robots specific tasks, like folding a T-shirt. But they often must show the robot how to fold the shirt 100 times before it can do so reliably.

Song said this points up how complex it is to develop an all-purpose humanoid robot: The robot will technically know how to fold the T-shirt, but it will get confused the next time it’s asked to do so on a bed with a different blanket or sheets, or in a different position, or if the weather visible through the window is not the same.

So trainers show the robot how to fold the shirt in every possible condition they can think of. “When you collect 100 examples of doing that one task, the robot can do that task,” Song said.

And that’s only in one house. Take the robot to a different house and you would have to train it again, Song said. Multiply that process by the number of tasks you’d like a robot to do—serve drinks, answer the door, iron a dress, vacuum the floor—and it becomes clear why humanlike robots are not around the corner.

Making matters worse, unlike with AI chatbots and driverless car technology, there is no ready warehouse of training data for humanoid robots. Researchers are just beginning to create the data.

Toyota’s partnership with Boston Dynamics aims to take the development of such robots to the next stage. Pratt said the companies have assigned the partnership 10 employees each, who will work in a Boston Dynamics facility in Waltham, Mass.

The companies start with a robot called Atlas, developed by Boston Dynamics. At a glance, it appears far more agile than Optimus. In the Oct. 10 event, Optimus robots didn’t fall over or bump into anything, but they also just trundled around, stood still, and generally looked as though they were mired in molasses. By comparison, Atlas flips, leaps, rolls, dances and is generally athletic, as you can see in this video.

The next challenge, Pratt said, is training robots to piece together discrete skills picked up in different tasks to handle a situation they have never seen before. “We have found ways, and we believe it has a lot of potential to go much further, of combining different tasks into general competence at mechanical things,” he said.

One of the most difficult things to replicate is the human hand, with its five fingers, all moving dexterously. No one in robotics has managed it, and even when robot hands get pretty far, the fingers sometimes break off in usage and generally are not robust.

Song suggests equipping humanoid robots with pincers rather than hands and fingers. She picked up a lightweight device on a desk behind her. It had two big pincers and a GoPro camera on top. She called it a handheld gripper and said it could instantly train a robot.

The idea is that when you buy your humanoid robot, you would also purchase one of these grippers. At home, you would take the gripper in your hands, turn on the GoPro, and go about whatever tasks you wanted the robot to carry out, like folding your laundry. Within limits, the robot would be able to replicate what you did after just one go, because it would be custom trained for your home. “Everybody can just carry this gripper and collect a lot of data for the tasks they care about,” she said.

In his Oct. 10 event, Musk predicted that every person on Earth would eventually buy an Optimus, which would become Tesla’s biggest product, surpassing EVs and its future Robotaxis. “I think this will be the biggest product ever of any kind,” he said.

Musk didn’t say when all of this would happen. I asked Song how long she thought it would take before there were all-purpose humanoid robots everywhere. Would our grandchildren see them? “I feel probably years,” she said, “but I don’t think it’s next generation. I think within our generation. When I’m getting old, I will probably have a robot at home.”

A fledgling startup is setting out to become one of Europe’s first “AI compute” hyperscalers, with renewable energy playing a pivotal part in its pitch to prospective customers.

The AI goldrush has spurred unprecedented demand for “compute,” which refers to the processing power, infrastructure and resources needed for tasks such as running algorithms, executing machine learning models, and processing data. One of the big beneficiaries of this demand has been Nvidia, emerging as a $3 trillion powerhouse off the back of demand for its GPU (graphics processing units) and associated AI hardware.

In tandem, an industry of cloud infrastructure providers has sprung up off the back of Nvidia, raising bucket loads of cash en route. In the U.S., we’ve seen the likes of Lambda and CoreWeave hit lofty billion-dollar valuations to expand their datacenter operations. Now, Finnish startup DataCrunch is throwing its hat into the ring, touting itself as one of the “few serious players” in the space with all operations in Europe.

‘GPU-as-a-service’

Founded in 2020 by CEO Ruben Bryon, DataCrunch — like its peers — sells GPUs “as-a-service,” promising to reduce the costs for AI processing. The company today said it has raised $13 million in seed funding, constituting $7.6 million in equity financing from backers such as ByFounders, J12 Ventures, and Aiven co-founder Oskari Saarenmaa. The remaining $5.4 million debt segment hails from Local Tapiola and Nordea.

While it’s slightly unusual for a seed-stage startup to raise such a significant portion as debt, DataCrunch has done this for the exact same reason that others in the space, such as CoreWeave, have also been raising hefty amounts of debt. It’s all about using physical assets — e.g. Nvidia GPUs — as collateral to secure loans, rather than giving away more equity.

It’s also more efficient to secure large buckets of capital this way, as the banks can simply take away the GPUs if things go belly-up for DataCrunch. For those who control the purse strings, it’s much less riskier than investing in a pure-play SaaS startup, for instance.

“Given the business that we’re in, our main expenses for expansion are capex [capital expenditure] driven,” Bryon told TechCrunch. “This is the logical way to go about it, and as we grow, additional access to that financing becomes available.”

This new round takes DataCrunch’s total funding raised since inception to $18 million, and will go some way toward helping it build out its infrastructure to support Nvidia’s latest servers and clusters, including the shiny new H200 GPU. In turn, this will help it grow a customer base that not only includes corporate clients such as Sony, but individual AI researchers working at the likes of OpenAI.

“That has always been an important market for us, and I think that this ‘individual’ market has been left behind by many,” Bryon said. “For me, personally, it’s important — at the weekend, I’m often using our own services, and have been since the beginning.”

Indeed, flexible, on-demand pricing is a far more alluring proposition for independent researchers and developers who might just need a little bit of compute for personal or university projects.

“People who are studying for a Masters or a PhD — that’s a segment we want to stay connected to because it’s often people who are a few years away from doing something really great,” Bryon said.

Hook them in now, and reap the rewards later when they hit the big time. That’s the general gist.

But there’s no escaping the giant elephant in the room, one that all the cloud companies are having to reckon with: the gargantuan amount of energy required to power this AI revolution.

Green machine

Part of DataCrunch’s “advantage” is the fact that its data centers are located in the Finnish capital, Helsinki, and Iceland — a country running on 100% renewable energy for years already.

“In Helsinki, we can subscribe to green energy from the grid,” Bryon said. “And currently, in one of our two Finnish data centers, the waste heat is captured to heat up Helsinki itself. In Iceland, we have the advantage that the ambient air temperature is always low, while the energy mix on the grid is already 100% green. So Iceland is pretty much one of the greenest places in the world to have these kinds of operations.”

This will be a big focal point for the company moving forward. While it plans to offer its services to any company globally, it will mostly remain anchored in the Nordics and Iceland. “Perhaps in the future we’ll look at Canada if we can find suitable locations, where we can have a similar advantage in terms of carbon footprint of our operations,” Bryon said.

It’s these “green” credentials that DataCrunch hopes will also set it apart from other European rivals: companies like FlexAI in France, which recently exited stealth with $30 million in seed funding; and Nebius, which recently emerged from the ashes of Russian internet giant Yandex and has just become a public company again.

There is a trade-off here, though: While low latency is often one of the big selling points for AI compute providers, DataCrunch isn’t necessarily going to be in that bucket, which means it will be better suited for a particular kind of workload.

“Our strategy is such that we’re not going to be the provider with the absolute lowest latency due to being in 100 locations around the world,” Bryon said. “We are more focused on the compute that doesn’t have that strict latency requirement. We can still have a decent enough latency though, it might not be 10 milliseconds, but it will still be something like 100 milliseconds.”

It’s also worth noting that DataCrunch’s data centers are in shared “co-location” facilities for now, but the company says it’s planning to start building out its own data centers in 2025 — something it will need significantly more capital for.

“I want us to be on a path toward going public with this company, and we’ll need access to plenty more capital to keep expanding the company,” Bryon said.

Train maker ‘analysing a possible transaction’ from steelmaker Sidenor

A Spanish train maker that the Madrid government wants to keep out of the hands of a Hungarian consortium is in talks over a potential acquisition by a steelmaker from Spain.

Talgo, which was drawn into a political storm by a takeover bid backed by Hungary’s illiberal prime minister Viktor Orbán, said on Monday that it was negotiating with Sidenor, which is based in the Basque country.

In August, Spain’s government vetoed the €619mn Hungarian bid for Talgo on “public security and order” grounds, creating a new conflict between EU member states and Orbán’s Russia-friendly government.

The Spanish government had no immediate comment on the Sidenor talks. But when Sidenor first signalled its interest in Talgo last week, Carlos Cuerpo, Spain’s economy minister, said the government was ready to “accompany and help” Talgo find “a viable long-term solution”.

Talgo said that in its talks with Sidenor, it was “analysing a possible transaction that could involve the acquisition of a significant percentage of [Talgo’s] share capital or its entire share capital”.

The Hungarian consortium, known as Ganz-Mavag, has vowed to take legal action in Spain and at EU level “to defend the legitimacy” of its offer for Talgo. But there are signs that its interest in the acquisition is fading.

Spain has classified the documents explaining its veto and declined to comment on whether its concerns are linked to Orbán and his relationship with Russia, the closest of any western leader since Moscow’s full-scale invasion of Ukraine in 2022.

But a senior Spanish government official previously told the Financial Times that Madrid was concerned about the possibility of the Hungarian consortium acquiring train technology that Ukraine needs to strengthen its rail links with the EU.

The Ganz-Mavag consortium is 55 per cent owned by Hungarian trainmaker Magyar Vagon, with the other 45 per cent in the hands of Corvinus, a state-owned development finance institution that co-invests with Hungarian companies abroad.

In a sign of waning interest, the Hungarian state this month reduced Corvinus’s share capital, taking out a sum that was not far short of the entity’s planned contribution to the Talgo bid.

Talgo’s principal business problem is a lack of production capacity. It has been struggling to fulfil orders on time for new trains from clients including Deutsche Bahn and state-owned Spanish train operator Renfe.

Part of the Hungarian consortium’s pitch was that it could quickly increase Talgo’s factory capacity using the existing plants of Magyar Vagon.

As a steelmaker, Sidenor does not produce any trains itself and it is not clear how it would seek to alleviate Talgo’s production bottlenecks.

Eastern Europe is also a growing market for train sales. Last month Talgo president Carlos Palacio and the president of Polish rolling stock maker Pesa, Krzysztof Zdziarski, signed a preliminary deal for Talgo to provide its technology for high-speed trains in Poland.