Want a hyper-hot Nintendo Switch gaming system? Good luck, and bring money

Remember, the Switch does have an “off” switch

The reviews are pouring in for the new Nintendo Switch gaming system, and the general consensus? It’s a home run for Nintendo, with a few caveats. One big one: it’s sold out just about everywhere, except eBay where systems are going for over $1,000. Many gamers who ordered early and often did receive their Switch systems beginning on Friday, and for some reason were hardly heard from all weekend.

So what’s our take on the Switch? It’s solid, with a few minor complaints, so check out our review here. And as we keep playing, we’ll keep you updated on all the latest Switch news and game updates. OK, time to put those joy-cons down and go to work, or school, or to sleep…

Wake me when I’ve gotten enough exercise in my sleep

After a not-so-great financial report a couple of weeks back and layoffs that sent about 100 workers packing, fitbit has rolled out another new fitness tracker, the Alta HR.

You can be excused if you think the Alta HR looks just like some of the other thin fitbit fitness bands, but it includes two new features: a new heart rate monitor and Sleep Insights, which don’t just track your sleep, it will also use machine learning, a facet of A.I., to analyze your daily hours of unconsciousness and make recommendations on how to improve it, something we could probably all use. Pretty cool, we must admit.

Anyway, fitbit is also probably hoping it is indeed something we think we can all use, and if you’d like to improve your fitness and your shuteye, you can order an Alta HR in April. For fitbit Blaze and Charge 2 owners, the Sleep Insights feature will be coming soon in app updates from fitbit.

No, the plutonium goes over there

Speaking of machine learning – and unlearning – the smart folks at MIT have created a system that uses human brainwaves to tell robots when they’ve done something wrong. This could be very handy in the coming robot apocalypse. Anyway, the smart humans at the MIT Computer Science and Artificial Intelligence Laboratory, or CSAIL, have fitted some human subjects with a sensor cap and then told some robots what to do while they watch.

If the robots do the task wrong, the human watching immediately generates a certain kind mental signal, letting the robot know essentially in real time that it’s making a mistake. For the humans in the test, they don’t really have to do any new mental gymnastics, we generate the “you’re doing it wrong” brain signal in milliseconds automatically.

The researchers hope that their work will lead to better accuracy from robots that are overseen by humans – which we hope is all of them. Robots working on extremely dangerous tasks – say, in a nuclear plant – could have human overlords watch their work by remote, and if they mess up, they’d be able to quickly correct their actions before disaster strikes. And one kind of robot where this tech could prove invaluable: self driving robotic cars.

Of course, the human part of the equation would have to know how to drive, that would be important. Check out more details on this very interesting story here.

We’ve got more news on our Facebook page and YouTube channel, and be sure to tune in to this week’s DT podcasts: Close to the Metal (computers and such) on Tuesday, Trends with Benefits (general tech shenanigans)  on Thursdays, and Between the Streams (movie and TV topics) every Friday.