Tech Radar

Hyundai isn't just buying a robotics company. They're betting that the future of logistics is mobile, autonomous, and humanoid. By picking up a 65% stake in Boston Dynamics for $325 million, they've finally moved from being a curious investor to owning the whole thing. The timing is a bit too convenient to be accidental. Boston Dynamics is finally pushing Atlas into commercial deployment, and the humanoid race is getting crowded. Tesla's Optimus and Figure AI are no longer just research projects, they're actual competitors. It's a massive gamble. Humanoids are notoriously difficult to scale, and the gap between a viral YouTube video of a robot doing a backflip and a robot that can actually move crates in a warehouse is huge. I wonder if Hyundai is buying the tech or just buying the hype. The Shift from Cars to Robotics The pivot to robotics-as-a-service is a move to stop treating autonomous driving as a niche car feature and start treating it as a ge...

I found 10,000 repositories on GitHub that distribute Trojan malware. They aren't forks of a single malicious project, and they aren't all coming from one bad actor. They're spread across different contributors with different names, all just sitting there in plain sight. I stumbled into this by accident. I have a project on GitHub and wanted to see if search engines had indexed it, so I typed the project name into Google. My repository showed up in the results, but it was surrounded by a weird pattern of others that looked almost identical in structure but served a completely different purpose. It's a clever play on trust. We tend to treat GitHub as a safe harbor for code, but these repos use that reputation to trick developers into downloading payloads. I spent some time digging through the data to see how deep the hole goes. The results are honestly a bit unsettling. The scale of the infection The campaign hit 10,000 repositories. The attackers didn't ...

Switzerland is finally admitting that its decades-long moratorium on nuclear power might have been a mistake. For years, the strategy was simple: move away from the atom. But the reality of a fragile grid and the pressure of energy security are forcing a quiet, pragmatic reversal. It's a classic case of policy meeting physics. I find it interesting that this shift is happening now. We've spent the last few years talking about a total transition to renewables, but the Swiss are realizing that wind and solar can't always keep the lights on when the temperature drops. They aren't rushing back in with blind optimism, but they are acknowledging that baseline power matters more than political purity. The question is whether they can actually build new capacity fast enough to matter. Nuclear projects are notorious for sliding timelines and ballooning budgets. Can Switzerland actually pull this off without spending the next twenty years in a construction stalemate? T...

The most ambitious aerospace company in history isn't buying a launch provider. It's buying the interface where its software is written. When SpaceX acquires a tool like Cursor or a similar AI-native IDE, they aren't just adding a line item to their procurement list. They're attempting to own the actual cognitive process of how their engineers build flight software. I've seen plenty of vertical integration in my time, but this is different. Usually, companies buy the factory or the raw materials. Buying the editor is an attempt to optimize the loop between a developer's thought and the machine's execution. It's a bold move, but it's also a bit weird. There's a fine line between "optimizing the workflow" and creating a proprietary black box that makes it impossible for any one engineer to actually understand the system. The real question is whether this actually speeds up the cadence of launches, or if it just creates a massive depen...

Local models used to be toys. They were slow, a pain to set up, and generally useless for anything beyond basic text completion. For a long time, the gap between a home-run model and a frontier API wasn't just a distance, it was a wall. I remember thinking that local LLMs were essentially a hobby for people who liked tinkering more than actually getting work done. That changed for me with GPT-OSS, but the recent Gemma 4 releases are where things actually get interesting. I've been using the 26b implementation in LM Studio, and for the first time, agentic coding loops actually work. I'm seeing about 75% of the accuracy and speed I get from the top-tier frontier models. It's not a perfect replacement, but it's close enough that I've stopped reaching for the cloud for most of my development questions. I've started treating my local setup as a personalized, lightning-fast version of Google. It's particularly useful for the grunt work that doesn'...

Fake job offers are the perfect delivery vehicle for malware. They don't rely on some obvious "Urgent: Your Account is Locked" scare tactic. Instead, they target your professional ambition and the basic trust you have in someone who says they want to pay you a lot of money to solve a hard problem. Last week, I got a LinkedIn message from a recruiter at a small crypto startup. We chatted for a few days. She described a broken proof-of-concept they needed a lead engineer for, which sounded exactly like the kind of mess I enjoy cleaning up. Then she sent me a public GitHub repo to review. I didn't clone it to my machine. I've seen this movie before. Instead, I spun up a throwaway VPS on Hetzner and pointed Pi at it in read-only mode, using only file-reading tools. It turns out the "broken proof-of-concept" was actually a very clever trap. The Social Engineering Hook Attackers don't start with malware; they start with a persona. They build L...