By Ginette Methot, Staff Writer, Vault Analytics
02/02/2018 | 12:00 p.m.
Every week we send out a newsletter of the most interesting artificial intelligence applications we found that particular week, and every week our newsletter readers tell us which ones they find most interesting when they select those links. Below are their top picks, in order of reader interest:
1. Narcissistic Tendencies or Simply Intrigue?
There’s an AI you can use to create your very own “digital voice” — and all it takes is a one-minute audio clip of your dulcet tones to create it. This means you can listen to “yourself” read books, text messages, and maybe even the latest news.
As narcissistic as listening to your own voice all the time sounds, you have to admit, it’s intriguing and may bring to mind these questions: How realistic of a voice impression can it create? Listen to examples here. Can someone else take a one-minute audio clip of me talking and then use my digital voice for nefarious purposes? The company says no, but the concept is troubling. Would Alexa take commands from my “digital voice”? If any of you experiment with this, send the results our way! We’d love to know how it goes.
2. “Google Search” for Surveillance Video
Say you’re a law enforcement officer who is searching surveillance videos for a suspicious black car. You could watch hundreds of hours of surveillance video, or you could simply use a video search engine called Ella. Ella employs deep learning to easily detect objects and find you every frame that has a black car in it. The impressive thing here is that you can use normal search language, like you would in Google search, to find important footage quickly.
3. AI Can (Sort of) Read Your Mind
One of the most intriguing AI applications we’ve seen in the past week is a new step toward mind-reading AI. Scientists are getting closer to visualizing more sophisticated human thought. In the past, researchers have been able to visualize basic images from recorded brain waves, like black-and-white letters. This new study, however, employs color and recreates more complex pictures. While showing subjects a picture (or simply having them remember a picture), scientists record brain waves and recreate colorful videos from them.
To be transparent, we still think this tech has a ways to go, but there’s no denying that scientists have made a solid step toward visualizing our thoughts. Watch it at work here:
4. Am I Really that Old? Maybe Not.
Your birthday comes around reliably every year, same day. It tells you how “old” you are, but it doesn’t tell you how biologically old you are. There’s a difference in geroscience. Your birthday may numerically bump you up every year, but if you’re, say, really fit, your biological age may be lower than your traditional age. Essentially, your body isn’t deteriorating as quickly as other peoples’ born the same year as you.
There are several different biomarkers that may indicate your biological age, and scientists are using them to develop biological aging clocks that predict this age. Deep learning helped create an aging clock that more accurately predicts this age across multiple populations — which means it created a clock that can be used around the globe, not just locally like other aging clocks.
5. Hunting the Hunters
In the cover of night, when poachers seek their fortunes by killing forbidden treasures, like rhinos and elephants and big cats, this unmanned, autonomous drone uses a combo of infrared and AI to identify, watch, and follow humans. The aim is to help authorities catch the ignominious poachers. The creators hope to get the drone into the hands of park rangers in African National Parks.
This drone tech belongs to a group called Air Shepherd, which may bring to mind the group Sea Shepherd from Whale Wars. The shared shepherd name conjures up an image of plucky conservationists hunting and confronting pirates and poachers. But the Air Shepherd org seems to be more likely to operate inside the law than the controversial Sea Shepherd.
Watch how the drone works here:
6. Kitchen Appliances Now Talk to Each Other
LG has created a community of AI appliances that communicate with each other, hopefully making your life easier. The fridge can track food items and their expiration dates, offering you recipes that use those ingredients before they go to waste. Once you pick your recipe — which your appliances will walk you through step-by-step — your fridge will preheat your oven for you. And when the oven’s done cooking your meal, it’ll set your dishwasher to a wash cycle that can handle the foodstuffs coming its way. And let’s not overlook the mundane: these appliances monitor themselves for maintenance issues so you can fix potential problems before they become actual problems.
Watch one minute of this video to see some of what the fridge can do (while you also painfully watch the LG VP of Marketing squirm when his assistant CLOi won’t respond to him).
7. Super Mario Is Possessed by Deep Learning
AI is seemingly leaving few areas untouched, which includes areas you wouldn’t think you would want AI, like letting an AI take over the game on your gaming console while you sit back and watch. But there’s a lot going on in the gaming space with AI, and here’s a prime example — this guy trained his AI to dominate Super Mario.
See how he does it here:
8. The Toothbrush Tutor
There’s an AI that tells you how to brush better. After it figures out your brushing patterns, it sends that info to your pared device which instructs you in better oral care. Your dentist would be so proud.
Note: We are not recommending or promoting this product. We include the video above just to give you an idea of how this tech works.
9. What Don’t They Know?
This AI tells you when they eat, sleep, drink, walk, rest. It also predicts their sicknesses, their fertility, and their body temperature — these bovine have no privacy. But the company that created this tech says all this data is helping dairy farmers improve milk production by 30 percent, and it’s also helping cows become healthier, and hopefully happier. It tracks cow activities with a neck sensor and then uses the collected data to inform farmers about their cows’ health. All farmers have to do is check the stats on their apps.
10. Cracking Your Lubb-Dupp Cipher
Lots of us at one point or another experience an unusual heartbeat. One second your heart is happily, quietly, and regularly lubb-dupping away (I promise I didn’t make lubb-dupp up), and the next second, it’s bizarrely off. If this irregular heartbeat bothers you enough, you’ll go get it checked out, and your doctor will likely send you home with a Holter heart monitor to wear for 24 hours — but then, of course, your heart does absolutely nothing out of the ordinary in those 24 hours, and your doctor tells you that you’re probably just fine.
But what if you could wear something that would monitor your heart all the time and also identify when something’s off? This AI tech, which is now FDA approved, does just that. It collects your data and identifies when a heartbeat is unusual for whatever activity you’re doing. It can read your heart rate every five seconds, 24/7, if need be. So when your heartbeat doesn’t make sense, it tells you to take an EKG immediately through your device. Also, on a similar front, researchers led by Andrew Ng are adjusting an AI so it can speed up and improve irregular-heartbeat diagnoses. All of this to better learn what your heart’s rhythm is telling you.
11. Helping or Hindering Child-Protection Orgs?
There are mixed opinions on whether AI is helping child-protection organizations properly identify high-risk children. But if built properly, predictive algorithms are seeing some success. When people call a child-abuse hotline to warn or suggest an organization take a closer look into a child’s home situation, some orgs use AI to comb through hundreds of data points related to that child’s situation and then predict the child’s level of in-home risk.
Sometimes the prediction is at odds with professional opinions. On one hand, people think that AI is helping the organizations reach beyond their workers’ biases. On the other hand, people think the predictive algorithms are biased themselves, which is also likely true. But likely the best solution is finding the right balance between humans and AI.
12. A Nyala Antelope–Zoo Win-Win
If you’ve ever been involved in a thermostat war, you know how hard it is to keep the temperature of your dwelling comfortable for you. The Nyala antelope feel your pain. They like a hot dwelling, and this British zoo that houses them wants them to be in the hotter temperatures they like — but it also wants to pay a lower energy bill while heating the enclosure. So they’re letting an algorithm decide when to turn on the heat lamps, instead of leaving them on all the time. It uses image processing to identify when the Nyala antelope are inside their enclosure, and then it turns up the heat. It’s right 95 percent of the time, and they hope to up the accuracy.
13. Mr. Travolta Will Dance on Your Living Room Floor
An augmented reality company uses deep learning to pull Uma Thurman and John Travolta from Pulp Fiction and project 3D versions of them into your living room (see video below). While we think this tech’s current form has a lot of glitches to work through before it’s flawless, it’s still brilliant, and it gives us a unique view into our possible future: one day you’ll be able to have anybody from a video hanging out in your living room in 3D, and your child’s imaginary friend from TrollHunters won’t be so imaginary.
14. But I Miss You Right Now
You can tell when you get home from a long day at work that your pet has surely missed you. It could be a number of cues, but you know them all. Now, though, your pet doesn’t have to wait until you get home to say hi. But now, when she misses you during the day, she can now video call you on her own volition.
15. Don’t Waste My Time with Checkouts
After a few setbacks, Amazon Go, an AI-powered grocery-like store, is live. You don’t need to go through a check-out line or scan your product to purchase it at this store. All you have to do is pick it up and walk out the door, and the store takes care of charging you for it.
16. You Say Three Years, Doc?
Deep learning can now predict the likelihood of your contracting any of six common diseases within four years. These six disease that plague adults include diabetes, lung cancer, and liver cancers. The AI’s aim is to help people proactively fight a disease before it takes hold.
Finally . . .
What application from this list was your favorite? What do you have to add to the list? Tell us in the comments below.
And if you want to be on top of the latest, most interesting AI applications but don’t want to spend your time combing through hundreds of articles each week, join our fun group of in-the-know subscribers who get our free weekly newsletter informing them about the latest applications of AI.
And if you’re into audio, have a go at our podcast, Data Crunch.
Source: Deep Learning on Medium