Artificial intelligence has beaten Q*bert, the iconic Atari video game

Image: aaron ontiveroz/ the Denver Post via Getty Images

Nothing is off-limits to artificial intelligence — even your favorite old video games. 

An artificial intelligence, developed by researchers from the University of Freiburg in Germany, has beaten the Q*bert arcade game by exploiting glitches in its design. 

In the game, players take the role of cartoon character Q*bert, who hops around a pyramid of 28 cubes. Every time Q*bert lands on a cube, it changes color. Players are tasked with changing every cube’s color without being captured by enemies that also roam around the pyramid. 

The AI found two sleazy ways to beat the game. First, it baited an enemy to follow it, then committed suicide by jumping off its pyramid. Though Q*bert lost a life, killing the opponent in the process left the player with enough points to reincarnate and repeat the cycle. 

Additionally, by jumping around the pyramid in a (seemingly) random fashion, the AI caused the pyramid’s tiles to begin to blink, and was granted more than one million points.  

The researchers believe that no human has ever uncovered these loopholes before, but this may not be entirely fair. The researchers tested their AI with an updated version of Q*bert — and the game’s developer claims the original version didn’t have such bugs. 

So it’s possible that humans could have found these loopholes as well. Nonetheless, the AI was able to find them after only five hours of training, which is probably less time than it would take most humans to beat the game. 

The researchers used sets of algorithms called “evolution strategies.” As the name implies, evolution strategies involve generating many algorithms and identifying those that perform the best through trial and error.  

In the paper, researchers suggest that evolution strategies can be considered “a potentially competitive approach to modern deep reinforcement learning algorithms.” Deep reinforcement learning algorithms mimic human neural networks and teach themselves effective strategy. A number of well known artificial agents fall into this category, including Alphabet Inc.’s DeepMind, which recently became one of the world’s most dominant Go players

It’s also possible that these algorithms could end up working together. “Since evolution strategies have different strengths and weaknesses than traditional deep reinforcement learning strategies, we also expect rich opportunities for combining the strength of both,” the researchers wrote. 

This study is a good sign for our robot overlords, which grow more dominant every day. In a recent study, AI outperformed lawyers in interpreting legal contracts. A Google-trained algorithm has trained itself to recognize patients at risk of heart disease — it doesn’t yet outperform existing medical approaches, but it’s on its way. 

It’s a serious, but exciting reminder to all humans. When it comes to skilled AI, nothing is out of reach — not even your childhood arcade games. 

Read more: https://mashable.com/2018/03/01/ai-beats-qbert/

New Google-powered AI scans eyes to detect heart disease, stroke risk

Image: ben brain/Digital Camera Magazine via getty images

The secret to identifying certain health conditions may be hidden in our eyes. 

Researchers from Google and its health-tech subsidiary Verily announced on Monday that they have successfully created algorithms to predict whether someone has high blood pressure or is at risk of a heart attack or stroke simply by scanning a person’s eyes, the Washington Post reports

Google’s researchers trained the algorithm with images of scanned retinas from more than 280,000 patients. By reviewing this massive database, Google’s algorithm trained itself to recognize the patterns that designated people as at-risk. 

This algorithm’s success is a sign of exciting developments in healthcare on the horizon. As Google fine-tunes the technology, it could one day help patients quickly and cheaply identify health risks. 

But don’t get too excited yet. The algorithm didn’t outperform existing medical approaches, such as blood tests according to the Washington Post report. The algorithms were able to pick out the patient at risk 70 percent of the time. That’s impressive, but it’s far from perfect. 

The procedure also hasn’t been replicated or validated to the point where it can be broadly accepted in the scientific community. 

And experts don’t think it will be necessary for Google’s technology to replace conventional, human-powered care in the near future. 

Maulik Majmudar, associate director of the Healthcare Transformation Lab at Massachusetts General Hospital, told the Post that age and gender are already good predictors of risk for such disease. While Google’s algorithm is an improvement, its improvement to current healthcare practices would be marginal at best. 

That said, it’s clear that artificial intelligence and machine learning have the potential to bring added convenience and affordability to the healthcare industry, even in areas as small as our eyes. 

Read more: https://mashable.com/2018/02/20/google-ai-detects-heart-disease-risk/

The FDA just cleared an Apple Watch band that measures your heart’s electrical activity

The AliveCor Kardia Band
Image: Alivecor

The Federal Drug Administration just cleared a new band for the Apple Watch that monitors the electrical rhythms in your heart. 

After a two-year process to satisfy the FDA’s stringent requirements, AliveCor announced today that the Kardia Band is now available for purchase for $199. It’s a mobile electrocardiogram (EKG), which measures a heart’s electrical activity and has traditionally been used by doctors to identify abnormal cardiac rhythms. 

“It’s a regulated measure of physiology by the FDA. Doctors can recognize over 100 conditions when they see an EKG,” AliveCor CEO Vic Gundotra told Mashable

But unlike conventional EKGs, this is one that you wear on your wrist. The heart rhythms display on the Apple Watch, and it takes about 30 seconds to collect the electrical signals. What’s more, the watch will notify the user when they might potentially be experiencing a common heart arrhythmia called Atrial fibrillation, or AFib. 

“It used to take a cardiologist, and you’d have to take off your shirt,” said Gundotra. “Now you can do it while at your kid’s soccer game. You can know in half a minute what’s going on in your heart.”

After being scrutinized by the FDA, Gundotra is confident the device accurately and reliably measures electrical heart activity.

“We had to prove that our algorithms are effective as a board of human cardiologists,” he said, noting that the stack of FDA regulatory papers “might be taller than me.”

We had to prove that our algorithms are effective as a board of human cardiologists

The Kardia Band can be especially useful for patients who suffer from AFib, Gordon Tomaselli, chief of the division of cardiology at John Hopkins University, told Mashable. “AFib may not be dangerous, however, it is associated with increased risk of blood clots in hearts,” he explained. “It also turns out to be a common cause of stroke that occurs for no obvious reason.”

“I think it’s a useful adjunct and useful for some patients,” he added. “And there are some people that want to know [they’re experiencing AFib] — and for those folks this is a great solution.”

Still, Tomaselli cautioned the Kardia Band might not be the best option for every single patient. 

“I think if you have a serious cardiac disease you should not be completely reliant on this to make a diagnosis,” said Tomaselli. “It is always best to make sure that your health care provider has a look at the rhythms.”

Most of all, he emphasized,If you don’t feel well, you need to see someone. Symptoms trump any recording.”

Besides watching for AFib, the Kardia Band is also infused with AI technology, a deep-learning tool the company calls “SmartRhythm Monitoring.” This program, however, doesn’t monitor heart rhythms. Instead, it watches your heart rate. The AI system checks your heart rate every five seconds, and will, based up your activity over the previous 30 minutes, notify you when your heart rate is pumping abnormally fast. 

For instance, the Kardia Band will alert you if your heart rate spikes, even though you might be idly sitting in a car.

“It worked on me when a cop pulled me over,” said AliveCor’s CEO, Gundotra. “We’re looking for things that don’t match our predictions.”

So for anyone that’s interested in better understanding their heart — both it’s rhythms and heart rate behavior — the Kardia Band can serve as a helpful tool. 

“It’s neat technology and there’s going to be more and more of this,” said Tomaselli. “I think some of this is going to be very useful not just in diagnosing things but helping, for those interested, to maintain their best cardiovascular health and fitness.”

And AliveCor’s Gundotra, after working through a tedious FDA approval process, feels the system validates this new technology. “We are very lucky to have such a strict regulatory body,” he said. “You can’t make ridiculous claims in this country.”

Every editorial product is independently selected by Mashable journalists. If you buy something featured, we may earn an affiliate commission which helps support our journalism.

Read more: http://mashable.com/2017/11/30/fda-clears-alivcor-kardia-band-for-apple-watch/