Quantum Computer in Space: Europe Leads the Way

Samantha: Welcome to The State of Tech, The European Edition, Sunday September twenty-seven, 2026. I'm Samantha Lawrence.

Bob: And I'm Bob Russell. Today: a quantum computer runs in space for the first time, robots that keep learning after they leave the factory, Snap pushes AI-powered smart glasses into the workplace, a North Korean operative slips into a company through remote IT work, and to close out, two things you can actually use today: the free open-source astronomy app Sky Map (Android), and the private photo library Lychee. Let's start in orbit.

A quantum processor operates in space for the first time, opening a new chapter for satellites and computing.

Samantha: A team of physicists led by Philip Walther at the University of Vienna has done something that sounded science-fictional a few years ago. They got a quantum processor to actually run in orbit. That means the fragile conditions quantum computing needs, extremely low disturbance, careful isolation, tight control, survived the launch and the harsh environment of space. This is the first time a quantum machine has demonstrably worked up there, and it moves quantum computing from lab curiosity to something with real practical possibilities beyond Earth.

Bob: The obvious question is: why bother sending a quantum chip into orbit at all? The answer is data. Modern Earth-observation satellites generate enormous amounts of imagery and sensor data, and the pipe back to the ground is the bottleneck. If a satellite can process raw data on board and only send back what actually matters, you skip the traffic jam. Quantum processors could, in theory, do certain filtering and pattern-recognition tasks far faster than classical chips, which is exactly the kind of work satellites need.

Samantha: What is striking is who pulled this off. Not a hyperscaler, not a defence giant, but a European university research group. The University of Vienna has been a heavyweight in quantum physics for years, and this puts Europe visibly at the front of space-based quantum work. That matters strategically. Quantum and space are two areas where Brussels has been trying to build sovereignty, and a demonstration like this gives European policymakers something concrete to point at when they argue for more funding rather than depending on American or Chinese platforms.

Bob: The bottom line: this is a first, not a finished product. Nobody is running a general-purpose quantum computer in orbit tomorrow. But the fact that a quantum processor survived launch and operated is the kind of milestone that unlocks follow-up experiments and, eventually, commercial services. Expect European space agencies and startups to try to build on this quickly. And for the wider industry, it sends a signal: quantum is starting to leave the perfectly controlled basement lab and step into places nobody thought it could go.

Samantha: For European industry, there is also a practical read. Satellite operators here have been complaining about downlink limits for years. If quantum processing on board becomes viable even for narrow tasks, it changes how earth-observation services are designed and priced. That is a European strength, from Airbus to smaller Earth-observation startups. This experiment gives them a research direction that is not just theoretical anymore.

Skylark Labs unveils an AI system that lets robots keep learning after they leave the factory.

Bob: Skylark Labs, together with Carnegie Mellon University and the University of California, Berkeley, has introduced something they call Continual Field-Adaptive Models. In plain language: robots that keep learning after they are deployed. Today, most robots come out of the box with a fixed set of skills, and when they meet a situation their training did not cover, they either get stuck or need to be sent back for retraining. This new approach lets them absorb new experiences in the field without forgetting what they already know.

Samantha: The technical trick that matters here is avoiding what researchers call catastrophic forgetting. If you just keep training a model on new data, it tends to overwrite the old skills. These new models are designed to add without erasing. For a warehouse robot or a delivery bot, that means it can meet a new type of package, a new layout, or an unexpected obstacle, and update its behaviour without a full retraining cycle. That is the difference between a robot that ages badly and one that gets better the longer it works.

Bob: The commercial angle is real. Companies buying robots today face a hidden cost: every time the environment changes, someone has to reconfigure or retrain the fleet. If robots can adapt on their own, the total cost of ownership drops and the range of jobs they can handle grows. That is especially interesting for logistics operators, hospitals, and public spaces, where conditions shift constantly and downtime is expensive.

Samantha: For European industry, the timing is useful. European manufacturers and warehouse operators have been cautious about robotics compared to their American and Chinese peers, partly because deployments turn out to be more brittle than the sales pitch suggests. A more resilient generation of learning robots could lower that bar. Robotics groups from Germany, France and the Nordics that focus on flexible automation now have a research direction they can plug into, rather than trying to buy stiffness out of an American platform.

Snap turns its smart glasses into an AI-powered work tool with Salesforce, NVIDIA and AWS.

Samantha: Quick interruption. If you listen to The State of Tech regularly, hit that like button and subscribe, that way you'll never miss an episode. Okay, moving on.

Bob: Snap is expanding its enterprise smart glasses business, branded SPECS, and it is going hard on generative AI. The company is teaming up with Salesforce, NVIDIA and Amazon Web Services, or Ay-Double-You-Es, to build what it calls SPECS Intelligence. The idea is that the glasses proactively suggest useful information based on a worker's schedule and what they are doing in the moment. So instead of pulling out a phone or a tablet, a frontline worker sees the next step, a customer record, or a repair instruction right in their field of view.

Samantha: The target audience here is not consumers. It is retail staff, field engineers, warehouse workers, and anyone doing hands-on work who currently loses time flipping between a screen and the physical task. Tie that to Salesforce records or an AWS-hosted knowledge base, and suddenly a maintenance technician has the manual in front of their eyes, a shop assistant sees stock information as they look at a shelf, and remote experts can guide a junior colleague through a repair from another country.

Bob: The competitive picture matters. Meta has been pushing its Ray-Ban smart glasses hard, Apple is circling with Vision, and now Snap is carving out the workplace lane with heavyweight cloud and chip partners. That is a smart position, because enterprises will pay for a device that clearly saves labour hours, whereas consumers are still deciding whether they want cameras on their face. Whether SPECS actually delivers on the demo is another matter, but the direction of travel for AI at work is clear: out of the browser and onto the body.

Samantha: For European businesses, especially in retail, logistics and industrial services, this could shift how frontline work is organised. But there is a real question about workplace surveillance and data protection. Glasses that see what a worker sees, tied to cloud AI, will land straight on the desk of European works councils and data protection authorities. Expect a slower, more negotiated rollout here than in the United States, with clear rules on recording, consent and what happens to the footage.

A North Korean operative infiltrates a New Zealand company as a remote IT contractor.

Bob: New Zealand's National Cyber Security Centre has published its Cyber Threat Report for 2026, and one case in it is a wake-up call. A North Korean operative used fake identity documents to get remote IT contract work at a major New Zealand company. When the company eventually spotted the deception and withheld payment, the operative threatened to leak commercially sensitive information they had already gathered from inside the network. It is the first time the agency has publicly confirmed that this scheme has actually landed inside a New Zealand business.

Samantha: This is not an isolated trick. Western intelligence agencies have been warning for a while that North Korea runs an organised programme to place IT workers in Western companies using fake identities, often routed through third countries. The workers do real jobs, earning real wages that flow back to the regime, and they sit inside company systems with legitimate access. The extortion twist here, threatening to leak data when caught, shows how this quietly shifts from a sanctions-evasion scheme into a straight cyber-espionage and blackmail risk.

Bob: The uncomfortable truth for HR and IT departments is that remote hiring processes, especially for contractors, were built for speed, not for counter-intelligence. Video interviews, document checks and background screening all have gaps that a determined state actor can exploit. Once someone is inside with developer or admin access, the damage they can do, or the data they can quietly copy, is significant. This case shows the extortion phase can arrive before anyone notices the initial breach.

Samantha: For European companies, the message is direct. The same scheme has been reported in the United States, the United Kingdom, and now the Asia-Pacific region, and there is no reason Europe is exempt. Security experts recommend tighter identity verification for remote contractors, especially those getting access to source code or sensitive data, and clear internal escalation paths when something feels off. It is a reminder that the biggest cybersecurity risk is not always a clever exploit; sometimes it walks in through the front door with a fake CV.

Sky Map turns any Android phone into a live window on tonight's sky.

Bob: To close out, two things you can actually use today. First one, Sky Map. It is the free Android app originally developed by Google and now maintained as an open-source project, and it does one thing beautifully: it shows you the sky above your exact location in real time. Open it, point your phone at any part of the sky, and you get a live overlay of the stars, planets, constellations and deep-sky objects overhead right now. No account required, no subscription, works on virtually any Android device.

Samantha: What makes it genuinely useful, rather than just pretty, is the search function. Type in the name of a planet or a constellation and it gives you an arrow pointing you in exactly the right direction until it is centred on your screen. Users consistently praise how well it works for parents and teachers, because you can stand outside with a child, hold up the phone, and match the screen to the real sky. There is also a night mode that switches the display to red tones so your eyes stay adjusted to the dark.

Bob: A couple of practical notes. Sky Map is free with no advertising in normal use, requires no account, and the source code is publicly available on GitHub. It does not need a data connection once installed. If you have children asking what that bright dot in the sky is, or you simply want to learn your way around the autumn constellations, Sky Map answers in about ten seconds. And with autumn evenings drawing in across Europe, the timing is right.

Lychee brings decades of family pictures back home, off the big clouds.

Samantha: The second thing to try today is Lychee. It is a free, open-source photo management tool that you run on your own web server or a small home machine. Think of it as an alternative to Google Photos, but where the pictures never leave your house. It organises your library into albums, supports tagging, and lets you share collections through password-protected links, so you stay in control of who sees what.

Bob: The setup is aimed at people who are comfortable with a bit of tinkering; installation runs through Docker or a standard web server stack, which means it can sit on a spare laptop or a Raspberry Pi. Once it is running, you upload your photos through a clean browser interface and start building albums straight away. Reviewers praise how polished the front end feels for a self-hosted tool, and how well it handles large libraries without slowing down. Because everything stays on your hardware, there are no monthly fees, no upload limits, and no ads.

Samantha: The European angle is straightforward: your photos stay under your roof, under European law by default, and out of anyone's training data. It is a genuine alternative for people who have been eyeing their cloud storage bill or worrying about what happens if a big provider changes their terms. Lychee is not the easiest option out there, but for anyone with a spare machine and an afternoon to spare, it is one of the most rewarding weekend projects you can start today.

Bob: And that is where we leave it for today.

Samantha: Today we covered: the first quantum computer running in space, robots that keep learning after deployment from Skylark Labs, Snap's AI-powered work glasses with Salesforce, NVIDIA and AWS, a North Korean operative infiltrating a New Zealand company through remote IT work, and two things to try yourself: Sky Map (Android) and Lychee.

Bob: Want to know more or react? Visit stateoftech.eu or email us at info@doorzetters.net.

Bob: State of Tech, the tech world in 15 minutes.