Archives for May 20, 2019

You’ll get the slicker interface and improved health features.

You don’t have to spring for the Galaxy Watch Active to put Samsung’s latest software features on your wrist. The company is rolling out an update that brings the Active’s One UI interface, health features and other optimizations to the regular Galaxy Watch, Gear S3 and Gear Sport. You should now have a livelier, easier to parse interface that matches One UI on more recent Galaxy phones. There’s more control as well, with settings that can toggle waking on touch, the Goodnight mode and the frequency of Daily Briefings. Some of the Active’s watch faces are also available to download if you’ve been pining for them since the newer wearable’s February premiere.

On the health side, you now have faster, simpler workout tracking through both the Samsung Health app and the widget. It now tracks outdoor swimming, too. You’ll find Apple Watch-style high heart rate alerts and more detailed sleep tracking that compares your average to your age group. And when you launch Samsung Health, you’ll see the Daily Activities screen right from the start to keep tabs on your progress.

Your watch might even last longer. There are promises of better battery life, in part through more aggressive memory management that closes background apps and tweaks settings. You can fine-tune other settings, such as the brightness and screen timeout. While these features won’t be so hot if you aren’t enthused with One UI’s memory management, they could be helpful if you’re determined to get through a weekend without recharging your wristwear.

We’re tantalizingly close to AI with all five senses.

As a social media platform with global reach, Facebook leans extensively on its artificial intelligence and machine-learning systems to keep the site online and harmful content off it (at least, some of the time). Following its announcement at the start of the month regarding self-supervised learning, computer vision, and natural language processing, Facebook on Monday shared details about three additional areas of research that could eventually lead to more capable and curious AI.

“Much of our work in robotics is focused on self-supervised learning, in which systems learn directly from raw data so they can adapt to new tasks and new circumstances,” a team of researchers from FAIR (Facebook AI Research) wrote in a blog post. “In robotics, we’re advancing techniques such as model-based reinforcement learning (RL) to enable robots to teach themselves through trial and error using direct input from sensors.”

Specifically, the team has been trying to get a six-legged robot to teach itself to walk without any outside assistance. “Generally speaking, locomotion is a very difficult task in robotics and this is what it makes it very exciting from our perspective,” Roberto Calandra, a FAIR researcher, told Engadget. “We have been able to design algorithms for AI and actually test them on a really challenging problem that we otherwise don’t know how to solve.”

The hexapod begins its existence as a pile of legs with no understanding of its surroundings. Using a reinforcement-learning algorithm, the robot slowly figures out a controller that will help it meet its goal of forward locomotion. And since the algorithm utilizes a recursive self-improvement function, the robot can monitor the information it gathers and further optimize its behavior over time. That is, the more experience the robot gains, the better it performs.

This is easier said than done, given that the robot is expected to figure out not only its location and orientation in space but its balance and momentum as well — all from a series of sensors located in the machine’s knees. By optimizing the robot’s behavior and focusing on getting it walking in as few steps as possible, Facebook taught the robot how to “walk” in a matter of hours, rather than days.

But what’s a hexapod to do once it figures out how to move? Go exploring, obviously. But it’s not so easy to induce wanderlust in robots that are typically trained to achieve a narrowly defined goal. Yet this is exactly what Facebook is trying to do, with some help from its colleagues at NYU and a robotic arm.

Previous research into imparting a sense of curiosity onto AI has focused on reducing uncertainty. Facebook’s latest efforts strive for the same goal but do so in a more structured manner.

“We actually started with a model that doesn’t know much about itself,” FAIR researcher Franziska Meier told Engadget. “At this point, the robot knows how to hold its arm, but it doesn’t actually know what actions to apply to reach a certain target.” But as the robot learns which torques need to be applied to move the arm into the next target configuration, it can eventually begin to optimize its planning.

“We use this model that tells us this, to plan ahead for a number of time steps,” Meier continued. “And we try to use this planning procedure to optimize the action sequence to achieve the task.” To prevent the robot from optimizing its routines too highly and getting caught in a loop, the research team rewarded the robot for actions that resolved uncertainty.

“We do this exploration, we actually learn a better model faster, achieve the task faster, and we learn a model that generalizes better to new tasks,” Meier concluded.

Finally, Facebook has been hard at work teaching robots how to feel. Not emotionally, but physically. And it’s leveraging a predictive deep-learning model originally designed for video. “It’s essentially a technique that can predict videos from the current state, from the current image and an action,” Calandra explained.

The team trained the AI to operate directly using raw data, in this case a high-resolution tactile sensor, rather than through a model. “Our work shows that such policies may be learned entirely without rewards, through diverse unsupervised exploratory interactions with the environment,” the researchers concluded. During the experiment, the robot was able to successfully manipulate a joystick, roll a ball and identify the correct face of a 20-sided die.

“We show that we can essentially have a robot manipulating small objects in an unsupervised manner,” Calandra said. “And what it means in practice is… we can actually predict accurately what’s going to be the output of [a given] action. This allows us to start planning into the future. We can optimize for the sequence of actions that will actually yield the desired outcome.”

Combining visual and tactile inputs could greatly improve the functionality of future robotic platforms and improve learning techniques. “To build machines that can learn by interacting with the world independently, we need robots that can leverage data from multiple senses,” the team concluded. We can only imagine what Facebook has in store for this. However the company declined to comment on potential practical applications for this research in the future.

It’s hard to believe it’s only three years old.

It’s only been three years since Google first introduced the Google Assistant, the AI-powered helper through which the company wants users to access its vast stores of information. There are at least two reasons it feels like it’s been around much longer. The first is that even before the Assistant arrived, you could say “OK Google” to Android phones and get quality voice-enabled search and access to content on your device. On the surface, the Assistant felt like a logical extension of those features, not something new.

Beyond that familiarity, though, Google Assistant feels older than three years because it’s probably the most important product the company has launched in recent memory. It didn’t feel like that at first, but with the benefit of hindsight we can see that the Assistant ties into almost everything the company does — which is fitting given Google’s mission to “organize the world’s information and make it universally accessible and useful.” The Assistant is the primary way Google thinks it can make all that data “universally accessible.”

Google did take its time rolling out the Assistant, though. It was announced at the I/O developer conference alongside the Google Home smart speaker, a pair of products meant to challenge Amazon’s successful Alexa-and-Echo combo. But Google Home didn’t ship until the fall, and the Assistant’s first introduction was in the ill-fated Allo chat app. Allo never took off, so its Assistant integration seemed as much as anything like a low-pressure way for Google to start publicly testing the service.

But when the Home speaker and the first Pixel smartphone arrived in the fall with Google Assistant on board, things started to take off. The Assistant wasn’t initially as good as Alexa at running a smart home, but Google quickly closed the skills gap by getting hundreds of brands on board with its platform. And the massive amount of information Google catalogs meant that the Assistant was quite smart right off the bat, particularly at knowing what’s relevant to its users.

The Assistant has also been one of Google’s best-supported products from both a hardware and software perspective. There are now multiple smart speakers and smart displays that can run the Assistant (that’s from both Google and third-party hardware makers) and it also runs on basically any Android or iOS device. Google has added its features into more of its apps, like Google Maps, and it works with Android Auto, Android TV and Wear OS devices too. At this point, the Assistant can be found on almost any hardware you can think of.

It would take far too long to run down the many new tricks it has gained in the last few years, but suffice to say that Google’s always adding more capability. One of the more impressive newer features is interpreter mode, which lets the Assistant serve as the middle-man between two people speaking in different languages. There’s also the creepy-but-intriguing Google Duplex feature, which lets the Assistant call a restaurant and book a table for you. It can also automatically check you into a flight, give you personalized recipe recommendations. Hell, it can even start up your Instant Pot.

That’s without mentioning the improvements Google announced a few weeks ago at its I/O developer conference. Among those are a much fasterAssistant that takes up less space on your phone, a boon for lower-end devices. It’ll also be more personalized than ever, an appropriate update given that the company originally pitched the Assistant as “your own personal Google.”

Of course, one can’t mention the Google Assistant without acknowledging that it requires you to give up even more of your personal information to Google. As people finally start to push back and desire more privacy, Google’s making it easier to manage the vast trove of information it has about you. But even with privacy concerns being top of mind, don’t expect Google to pull back on prioritizing the Assistant. In just a few short years, it’s become one of the company’s defining products. If you’re still not convinced, just recall the enormous showcase the company has built in the parking lot of CES the last two years. It’s all about the Assistant, and Google doesn’t want you to forget it.

They’ve seen which way the wind is blowing.

EV infrastructure company EVgo has teamed up with Chevron to install fast chargers at select filling stations in California. The chargers, with up to 100 kW of capacity, are now running at Menlo Park and under construction at other locations in Los Angeles and Bay Area communities including Aliso Viejo, Manhattan Beach and Venice. “EVgo is helping everyone — including traditional fueling stations — take advantage of the benefits of EV adoption,” EVgo CEO Cathy Zoi said in a statement.

With an EVgo fast charger, you can get about 90 miles of range in 30 minutes, and they work with any EV that accepts DC Fast Charging. Its charging network is one of the larger ones in the US, and the company recently teamed up with GM to offer new Chevy Bolt owners free charging for a limited time. Having them at conventional gas stations is particularly convenient for highway travelers, who can take care of meals and bathroom breaks while their vehicle charges up.

Since customers won’t be buying fuel, service station owners haven’t exactly been motivated to install chargers. However, that’s slowly changing — for instance, one of the largest suppliers of gas dispenser recently invested in Tritrium, and EV charging company. And gas stations can make a cut on the electricity and sell snacks to customers, as I discovered when touring France’s charging infrastructure in a Renault Zoe.