SuperMe is a network of AI profiles built from the knowledge of real-life experts. When you ask a question, the platform determines which experts are best suited to answer and creates a group chat with their AI profiles.

Over on IEEE Spectrum, [Stephen Cass] has a project that would probably have seemed like a magic a few centuries ago. Imagine a box that uses a magnetic device to holds a locking bolt, requiring a magnet to disengage it. The only problem is that any strong magnet can open the box. [Stephen], instead, used an array of 64 Hall-effect sensors developed by [Peter Jansen] to look for a specific magnetic gesture.

Inside is a Teensy that can read the 8×8 grid of sensors. A magnetic stylus allows you to draw an unlock pattern. The system lets you train the gesture you want. There are a few problems. The sensor covers a small area, so it is hard to reliably produce the gesture. Putting an obvious mark on the box is a dead giveaway, so [Stephen] put a Celtic knotwork pattern on the box and uses part of it as a reference.

The other problem is that if you run out of battery, the box is locked forever. Well, at least until you apply mechanical force and break the box open. Wireless charging would be an option, but we might favor a small jack that would allow you to plug in an external battery.

This is a clever use of the sensor. However, we might have preferred a mechanical solution that didn’t require batteries. Or, perhaps, you could use a few normally open and normally closed reed switches to make a kind of magnetic combination lock. It would still need a battery, but it wouldn’t draw current until you connected the switches. Both methods also have drawbacks, of course. Like most things, you just get to pick the drawbacks you can live with.

You can see a video of the magnetic imaging tile from [Peter] below. We really want a setup like this for our secret door. We also wonder whether switchable magnets could provide a battery-free solution.

Amazon’s Echo Time is a $199.99 Alexa+ smart alarm clock with sleep sensing, adaptive wake-up times, customizable clock faces, and bedtime routines designed to make mornings easier.

Block previews Dessa, a voice AI model trained without text. Explore its local hardware claims, safety risks, and what businesses should assess next.

The post Block Previews Dessa, an AI Voice Model Trained Without Text appeared first on TechRepublic.

Block said that the voice model can run locally on a MacBook M4 Max

Code in an Android 17 beta suggests Pixels could light your face with the screen, but Google hasn’t announced it.

At TechCrunch Disrupt, Flock CEO Garrett Langley defended his company's surveillance technology despite growing privacy concerns, arguing that its crime-fighting benefits outweigh the risks of abuse.

Atomic clocks are super accurate. But, unsurprisingly, some atomic clocks are better than others, depending on your definition of better. Researchers at Vienna University claim to have a superior thorium-based atomic clock. It isn’t necessarily more accurate than other atomic clocks, but it is self-regulating.

Apparently, prior atomic clocks need help to stay stable. But scientists have thought that thorium, with its very narrow gap between two states, would make an excellent clock with the help of a laser. However, the laser has to have precise properties that, at least in early experiments, were provided by another nuclear clock.

Obviously, you’d like to be able to use a laser to excite the source — a thorium crystal in this case — and then use the output of that to adjust the laser. That’s apparently what the team has done. If the laser drifts, the thorium nuclei absorb less light, and that adjusts the output frequency to compensate. If you can follow the physics, you can read the source paper in Nature.

Optical atomic clocks are the way to go for time standards, but this research has promise to provide even more accurate results in the future.