2026-01-26 20:00:46
Over the past years, the author of the cURL project, [Daniel Stenberg], has repeatedly complained about the increasingly poor quality of bug reports filed due to LLM chatbot-induced confabulations, also known as ‘AI slop’. This has now led the project to suspend its bug bounty program starting February 1, 2026.
Examples of such slop are provided by [Daniel] in a GitHub gist, which covers a wide range of very intimidating-looking vulnerabilities and seemingly clear exploits. Except that none of them are vulnerabilities when actually examined by a knowledgeable developer. Each is a lengthy word salad that an LLM churned out in seconds, yet which takes a human significantly longer to parse before dealing with the typical diatribe from the submitter.
Although there are undoubtedly still valid reports coming in, the truth of the matter is that the ease with which bogus reports can be generated by anyone who has access to an LLM chatbot and some spare time has completely flooded the bug bounty system and is overwhelming the very human developers who have to dig through the proverbial midden to find that one diamond ring.
We have mentioned before how troubled bounty programs are for open source, and how projects like Mesa have already had to fight off AI slop incidents from people with zero understanding of software development.
2026-01-26 17:00:56
If you’re reading this, you probably have some fondness for human-crafted language. After all, you’ve taken the time to navigate to Hackaday and read this, rather than ask your favoured LLM to trawl the web and summarize what it finds for you. Perhaps you have no such pro-biological bias, and you just don’t know how to set up the stochastic parrot feed. If that’s the case, buckle up, because [Rafael Ben-Ari] has an article on how you can replace us with a suite of LLM agents.
He actually has two: a tech news feed, focused on the AI industry, and a retrocomputing paper based on SimCity 2000’s internal newspaper. Everything in both those papers is AI-generated; specifically, he’s using opencode to manage a whole dogpen of AI agents that serve as both reporters and editors, each in their own little sandbox.
Using opencode like this lets him vary the model by agent, potentially handing some tasks to small, locally-run models to save tokens for the more computationally-intensive tasks. It also allows each task to be assigned to a different model if so desired. With the right prompting, you could produce a niche publication with exactly the topics that interest you, and none of the ones that don’t. In theory, you could take this toolkit — the implementation of which [Rafael] has shared on GitHub — to replace your daily dose of Hackaday, but we really hope you don’t. We’d miss you.
That’s news covered, and we’ve already seen the weather reported by “AI”— now we just need an agenetic sports section and some AI-generated funny papers. That’d be the whole newspaper. If only you could trust it.
Story via reddit.
2026-01-26 14:00:00
[Bhaskar Das] has been tinkering with one of Nintendo’s more obscure handhelds, the DSi. The old-school console has been given a new job as part of an augmented reality app called AetherShell.
The concept is straightforward enough. The Nintendo DSi runs a small homebrew app which lets you use the stylus to make simple line drawings on the lower touchscreen. These drawings are then trucked out wirelessly as raw touch data via UDP packets, and fed into a Gemini tool which transforms them into animation frames. These are then sent to an iPhone app, which uses ARKit APIs and the phone’s camera to display the animations embedded into the surrounding environment via augmented reality.
One might question the utility of this project, given that the iPhone itself has a touch screen you can draw on, too. It’s a fair question, and one without a real answer, beyond the fact that sometimes it’s really fun to play with an old console and do weird things with it. Plus, there just isn’t enough DSi homebrew out in the world. We love to see more.
2026-01-26 11:00:00
We know this one is a few years old, but unless you’re deep into the cycling scene, there’s a good chance this is the first time you’ve heard of [Ali Clarkson’s] foray into home made rope spokes.
The journey to home-made rope spoke begun all the way back in 2018, shortly after the company Berd introduced their very expensive rope spokes. Berd’s spokes are made of a hollow weaved ultrahigh molecular weight polyethylene (UHMWPE) rope with very low creep. They claim wheels stronger than steel spoke equivalents at a fraction of the weight. Naturally forum users asked themselves, “well why can’t we make our own?” As it turns out, there are a handful of problems with trying this at home.
There are a number of ways to skin this proverbial cat, but they all center around some very special nautical ropes, namely, Robline DM20. This rope has excellent wear and creep characteristics, in a hollow weave much like what Berd developed. The hubs also require the addition of a bevel around the spoke holes to prevent wear. Beyond those two similarities, there are quite a number of ways to lace the spokes between the hub and wheels.
As detailed by [Ali Clarkson], one method involves creating loops out of bike spokes, with a custom jig and some brazing. Then a length of rope is passed through the hub and a special hitch is used to keep it in place. Two loops are made in the ends of this length of rope and passed through the spoke ends made earlier. Finally everything is brought up to tension and trued much like a normal wheelset. The front wheel ended up weighing around 700g, a rather impressive feat for a 24 inch downhill wheel.
However, a number of other methods have been tried on the forum threads. Namely, a number of users have attempted to varying degrees of success putting a length of spoke inside the hollow rope weave and “Chinese finger trapping” it together. The key issue here is sourcing a glue strong enough to hold the spoke piece on at lower tensions, but flexible enough to not crack with the cyclical loading on a rim.
Ultimately, this is a great look at the properties of some extremely special rope. This also isn’t the first time we have seen strange bicycle wheels made with UHMWPE.
2026-01-26 08:00:59
If predictions hold steady, nearly half of the United States will be covered in snow by the time this post goes live, with the Northeast potentially getting buried under more than 18 inches. According to the National Weather Service, the “unusually expansive and long-duration winter storm will bring heavy snow from the central U.S. across the Midwest, Ohio Valley, and through the northeastern U.S. for the remainder of the weekend into Monday.” If that sounds like a fun snow day, they go on to clarify that “crippling to locally catastrophic impacts can be expected”, so keep that in mind. Hopefully you didn’t have any travel plans, as CNBC reported that more than 13,000 flights were canceled as of Friday night. If you’re looking to keep up with the latest developments, we recently came across StormWatch (GitHub repo), a slick open source weather dashboard that’s written entirely in HTML. Stay safe out there, hackers.
Speaking of travel, did you hear about Sebastian Heyneman’s Bogus Journey to Davos? The entrepreneur (or “Tech Bro” to use the parlance of our times) was in town to woo investors attending the World Economic Forum, but ended up spending the night in a Swiss jail cell because the authorities thought he might be a spy. Apparently he had brought along a prototype for the anti-fraud device he was hawking, and mistakenly left it laying on a table while he was rubbing shoulders. It was picked up by security guards and found to contain a very spooky ESP32 development board, so naturally he was whisked off for interrogation. A search of his hotel room uncovered more suspicious equipment, including an electric screwdriver and a soldering iron. Imagine if a child had gotten their hands on them?
But the best part of the story is when Sebastian tries to explain the gadget’s function to investigators. When asked to prove that the code on the microcontroller wasn’t malicious, he was at a loss — turns out our hero used AI to create the whole thing and wasn’t even familiar with the language it was written in. In his own words: “Look, I’m not a very good hardware engineer, but I’m a great user of AI. I was one of the top users of Cursor last year. I did 43,000 agent runs and generated 25 billion tokens.” Oof. Luckily, the Swiss brought in a tech expert who quickly determined the device wasn’t dangerous. He was even nice enough to explain the code line-by-line to Sebastian before he was released. No word on whether or not they charged him for the impromptu programming lesson.
It wasn’t hard for the Swiss authorities to see what was inside the literal black box Sebastian brought with him, but what if that wasn’t possible? Well, if you’ve got an x-ray machine handy, that could certainly help. The folks at Eclypsium recently released a blog post that describes how they compared a legit FTDI cable with a suspect knock-off by peering at their innards. What we thought was particularly interesting was how they were able to correctly guess which one was the real deal based on the PCB design. The legitimate adapter featured things like ground pours and decoupling caps, and the cheap one…didn’t. Of course, this makes sense. If you’re looking to crank something out as cheaply as possible, those would be the first features to go. (Editor’s note: sarcasm.)
It doesn’t take an x-ray machine or any other fancy equipment to figure out that the Raspberry Pi 5 is faster than its predecessors. But quantifying just how much better each generation of Pi is compared to the other members of the family does require a bit more effort, which is why we were glad to see that The DIY Life did the homework for us. It’s not much of a spoiler to reveal that the Pi 5 won the head-to-head competition in essentially every category, but it’s still interesting to read along to see how each generation of hardware fared in the testing.
Finally, Albedo has released a fascinating write-up that goes over the recent flight of their Very Low Earth Orbit (VLEO) satellite, Clarity-1. As we explained earlier this week, operating at a lower orbit offers several tangible benefits to spacecraft. One of the major ones is that such an orbit decays quickly, meaning a spacecraft could burn up just months or even days after its mission is completed. For Albedo specifically, they’re taking advantage of the lower altitude to snap closeup shots of the Earth. While there were a few hiccups, the mission was overall a success, providing another example of how commercial operators can capitalize on this unique space environment.
See something interesting that you think would be a good fit for our weekly Links column? Drop us a line, we’ve love to hear about it.
2026-01-26 05:00:00
Although jet engines are theoretically quite simple devices, in reality they tread a fine line between working as intended and vaporizing into a cloud of lethal shrapnel. The main reason for this is the high rotational speed of the rotors, with any imbalance due to poor manufacturing or damage leading to undesirable outcomes. It’s for this reason that [AlfMart CNC Garage] on YouTube decided to spend some quality time building a balancer for his DIY RC turbine project and making sure it can prevent such a disaster scenario.
In the previous part of the series the turbine disc was machined out of inconel alloy, as the part will be subjected to significant heat as well when operating. To make sure that the disc is perfectly balanced, a dynamic balancing machine is required. The design that was settled on after a few failed attempts uses an ADXL335 accelerometer and Hall sensor hooked up to an ESP32, which is said to measure imbalance down to ~1 mg at 4,000 RPM.
A big part of the dynamic balancing machine is the isolation of external vibrations using a bearing-supported free-floating structure. With that taken care of, this made measuring the vibrations caused by an imbalanced rotor much easier to distinguish. The ESP32 is here basically just to read out the sensors and output the waveforms to a connected PC via serial, with the real work being a slow and methodical data interpretation and balancing by hand.
