Microplastics are everywhere, from the Mediterranean seafloor to Mount Everest , and the tiny fragments of litter can take hundreds of years to degrade. They can also absorb heavy metals and pollutants that could harm humans or animals that inadvertently consume them.

At less than 5mm in length, the polymers are hard to remove manually. But there may be a more efficient method of eliminating the plastic scourge: attacking them with bacteria-sized robots.

Scientists at the University of Chemistry and Technology in Prague recently developed self-propelled microrobots that can swim onto microplastics and break them down.

The tiny bots — about the size of a red blood cell — use solar power to move through water and dismantle microplastics that they find.

In a proof-of-concept study , the microbots successfully swam down a maze of channels to efficiently degrade a range of synthetic microplastics:

The team still needs to further investigate the potential capabilities and environmental impacts of the bots. But they believe that their research could lay the foundations for a new way of degrading microplastics.

Greetings Humanoids! Did you know we have a newsletter all about AI? You can subscribe to it right here .

Alexa? Alexa! Alexa?! What happens when your virtual assistant dies?

Virtual assistants are a lagniappe technology. Almost nobody actually needs them, but when they work properly they bring a little something extra to your life.

Even though they’re heavily flawed and a clear-and-present danger to our privacy, it’s still super nice to be able to yell “Hey Google, play some Skynyrd!” while I’m doing the dishes and, lo and behold, it’s Freebird time.

But not all virtual assistants are created the same. We’ve got Cortana tethered to our desk, Alexa trying to manage our homes, and either Google or Siri in charge of our phones.

One cool thing about virtual assistants is that my data persists even if I lose all my gadgets. I’m on my fourth or fifth Google Assistant-enabled phone and, I must admit, it’s nice to sign in to my account on a new device and skip the setup because Assistant already knows my voice. We’ve got history.

Mister Greene, Sir, no need for you to wait in line. Your paperwork and admission has already been handled. Right this way to the VIP section. Fancy some YouTube videos?

For many of us, virtual assistants just play our music, dim our lights, and tell us what time it is in Amsterdam so we’ll know whether it’s an appropriate hour of the day to bug our bosses on Slack – maybe that last one’s just me.

But a handful of unnecessary, yet welcome, little conveniences can eventually become the foundation for what I like to call a quality of life checkpoint .

I never remember to turn the music on before I start doing the dishes. I don’t want to go back to a lifestyle where I have to stop what I’m doing, dry my hands off, pull out my phone, connect it to my Bluetooth speakers, open a music player, choose a song, and then press play. That feels barbaric now, even though I was born in the 1970s and grew up 8-tracks, cassettes, and compact discs.

It’s not that we couldn’t live without our virtual assistants (and, for the sake of our privacy, we really should ), but there’s a reason why Amazon and Google sold more than 30 million smart speakers in Q4 of 2020 alone . People like convenience.

And, it’s a pretty safe bet that Amazon, Apple, Google, and Microsoft are going to be around for a long, long time. So if you’re already willing to let them have your data, at least you’re helping to train the assistant to better aid you.

But nothing lasts forever and the grim reaper comes for machines too.

A few years back I reviewed Anki’s “Vector,” a cute little robot that was just too awesome to fail. Vector was a second gen robot stuffed with Amazon’s Alexa and a digital display “face” that was designed and animated by Pixar artists.

Credit: Nicole Gray

Isn’t it cute? Be careful getting attached though, it almost died. Anki went out of business and, for a while, it appeared as though Vector and all its sibling robots would become zombies with no connection to their “brains” in the cloud.

Luckily for people who invested in the platform, a company called Digital Dream Labs came along and purchased assets from Anki including the robots and their servers. Vector lives, but it’s as much a cautionary tale as it is a redemption story.

Vector was just a toy, but my toddler became quite attached to it. Imagine spending $250 on a toy that relies on the cloud to function and then having to tell your kid it’s never going to work again because a tech startup went bankrupt.

That’s inconvenient. But what happens when we’re dealing with something aimed at adults, such as Cortana or Google Assistant? It’s hard to imagine a trillion-dollar tech behemoth going under or shuttering a program as lucrative and ubiquitous as virtual assistants, but there no guarantees in tech.

Twenty years ago AOL, Yahoo, and My Space were too big to go under too. That’s not to say the big tech companies of today are anything like those – but you never know what’s going to happen in a decade or three.

Imagine the person who uses Alexa for everything – lighting, temperature control, a Ring doorbell cam , an alarm system and door locks, music, appliances, ordering groceries and household items, and even controlling their TVs. Losing Alexa would be a major inconvenience to this person.

But, even worse, Alexa isn’t a toy. I don’t even use it everyday, but over the years its collected tens of thousands of hours worth of data on me, my family, and my guests. And Google Assistant knows even more about me. I even have Cortana listening at my desk computer, so it’s been in position to record thousands of hours worth of interviews and work-related conversations.

What would happen to all that data if, in an absolutely bonkers near-future, Google, Microsoft or Amazon went out of business and shuttered their cloud services for virtual assistants?

Because that data is a gold mine. It’s worth more to anyone trying to build AI that uses natural language processing to service humans than all the virtual assistant hardware that’s ever been manufactured.

Anyone can mold plastic, but only a few companies in the entire world can get those devices into the hands of billions of people who’ll not only pay for them, but give away their data for free too.

And I don’t know about you, but I didn’t make any of these companies sign a legally-binding contract saying they can’t sell my data in the event they get out of the assistant business. Most of us simply agree to terms of service that protect the company from us. Sure, those companies usually have a webpage somewhere where they clearly state that they will never sell or give away our data — all I can say is, good luck holding them to it in court.

I’m not saying any of these companies do or even will sell our private data. It’s actually in their best interests to keep a stranglehold on it and use it to train their systems.

But if virtual assistants go the way of 3D TVs, DVD players, and Pet Rocks, big tech companies won’t just delete our data. It’s going to end up in someone’s hands… the question is, who?

We reached out to Amazon, Microsoft, and Google, but at the time of publishing hadn’t received commentary from any.

YouTube’s new AI will block videos that are inappropriate for kids

YouTube will start using AI to automatically add age restrictions to videos “over the coming months,” the company announced on Tuesday.

The site currently relies on content reviewers to flag videos that aren’t appropriate for viewers under 18, but will soon start using machine learning to detect content for review. Uploaders will be able to appeal the system’s decisions.

In addition, viewers who try to evade the restrictions by watching videos embedded on third-party websites will be redirected to YouTube, where they’ll have to sign-in to show they’re over 18.

“This will help ensure that, no matter where a video is discovered, it will only be viewable by the appropriate audience,” the company said in a blogpost .

The new restrictions are YouTube‘s latest response to concerns over the protections it provides to young users. The site has been engulfed by a series of scandals involving videos of child abuse, and last year was fined $170m for collecting kid’s personal data without their parents’ consent.

Last August, the company launched a web version of the YouTube Kids mobile app for children under 13, which offers curated content, parental control features, and filtering of videos.

The new automated restrictions will detect content on the main platform deemed inappropriate for anyone under 18.

YouTube‘s recent experiments with content moderation suggest that the automated system will lead to far more videos being age-restricted.

In the second quarter of this year, YouTube took down a record number of videos after increasing the role of AI in its content review efforts. Neal Mohan, the company’s chief product officer, told the Financial Times this week that the site recently returned to using more human moderators. But for age restrictions, the company clearly feels that more AI is required.

So you’re interested in AI? Then join our online event, TNW2020 , where you’ll hear how artificial intelligence is transforming industries and businesses.