As a tumultuous and exciting 2017 comes to a close, we wanted to take a few moments to reflect on the big stories and developments we were most excited about in the year before.
Regardless of your political persuasion, there’s no doubting the crazy and uncertain times we’re entering/experiencing as a country. Adding to that uncertainty are some of the things we’re most excited about — with new developments in artificial intelligence, machine learning, voice assistants, automation, etc. sweeping the global technological landscape, it’s an uncertain time across the board. While we recognize how disorienting or even disconcerting some of these technological developments may be, we are also thrilled at the possibilities they open up. So, without further ado, here are our five favorites:
One of the subjects we’ve written the most about in the last year is the Alpha Zero/Alpha Go platform over at Alphabet’s DeepMind division. By aiming a neural network (Alpha Go) at reams of top human Go games, Alphabet was able to build an A.I. that could take on the top player in the world and win — at the most complex strategy game we have, no less. DeepMind took it a step further with the Alpha Zero platform, in which the A.I. didn’t learn from human games, but rather games between Alpha Go and itself, over and over and over again. Without the human biases present in the seed games the A.I. is learning from, the Alpha Zero was able to smash Alpha Go… which had already dominated the top player in the world.
Teaching a massive supercomputer to beat top human players in strategy games may sound trivial or a novelty, but it has huge implications for teaching machines how humans think, how we reason, how we make decisions, how we weigh risk and reward, etc. As such, DeepMind and Alphabet’s progress on this front has us giddy with excitement for what these developments could mean in practical applications in the year ahead and beyond.
Typing on a full keyboard can be both fast and efficient… if you’re a native typer with fast fingers. But even the fastest keyboardists aren’t as fast as human speech. That’s why we’re pumped that Siri, Google Assistant, Cortana, and Alexa are making the leap this year. That’s not to say any of these platforms are perfect by any stretch; speech recognition still has a ways to go before it’s 100% on natural human speech. But the bigger trend is toward wider and wider voice-based user experiences on our digital devices.
We’re used to talking to Siri to ask questions or set reminders; we’ve invited Alexa and Google Assistant into our homes to make our lives easier and more enjoyable. Moving forward, even more and more resources will be devoted to perfecting natural human speech recognition, but this was the transition year where it went from novelty to necessity, and we can’t wait to see the fruits of those labors. We’re already starting to see platform expansion with Alexa skills developed for all sorts of third party companies (hint hint — we’re experts at that and would love to help you find traction on these platforms), and we don’t see that trend slowing any time soon.
Apple is rarely the first mover on any given technological scene. They usually wait until there’s a proof of concept or a pretty good prototype or first generation product on the market, design the hell out of a superior version, make it beautiful and integrate seamlessly with their own suite of products, and release a polished, amazing version of that product. What’s even more exciting, though, is when Apple opens up certain tools to developers to signal what’s coming down the pike.
When Apple released AR Kit, we knew it would be a game changer. Not so much the actual developer tool itself, mind you, but rather what that tool portends.
Augmented reality (A.R.) is nowhere near full market adoption, but the hardware many of the top device manufacturers are peddling now support some version of A.R. With Apple jumping on board in this way, they’re telling the market this technology and user interface are going to feature prominently in the future of handheld devices. Heads up displays showing realtime, geographically accurate and personally useful information could be a massively useful/interesting/fun tool for users across the spectrum, and the implications for the future of mobile devices is thrilling.
Fingerprint scanners have become nearly ubiquitous on smart phones. They’re almost assuredly unique to each user and are hard to fake. And across the board, users are becoming more comfortable with using this type of security in place of traditional passwords or passcodes. With the iPhone X, Apple has taken this a step further with Face ID. By projecting a dot matrix onto a users face, it can instantly determine if you’re the owner of that device and open said device simply by pointing the phone at your face. While some people may be uncomfortable with Apple having access to 3D mapping of one’s face, it still presages a needed upgrade in security protocols away from passwords that are simply too easy to crack.
With the sheer number of high-level, massive data breaches seen across the corporate spectrum this past year, upgrading our security measures is a necessary and welcomed societal step. Biometric security may not be the panacea that solves every one of those issues, but it’s certainly a step in the right direction.
Quantum computing has long been considered the Holy Grail of supercomputing. Instead of being limited by the binary 1 or 0 of classic computer operations, quantum computers can have information and data that exists in multiple states at once, thereby increasing the computing power per cubic inch exponentially for that computer. Companies from Google to Intel are throwing massive amounts of resources at that hardware/physics problem, and it’s closer than ever before to reality. The implications for something like this will be felt at the upper echelons of computing power first (weather forecasting models, climate change modeling, major physics experiments, next generation cryptography… the problems we point our current batch of supercomputers at), but the trickle-down effects could be huge. It could make its way into A.I. and supercharge the trend recognition and thought development of neural networks, perhaps? Whatever it ends up being used for, however, the closer we get to quantum computer reality, the more and more we’ll be able to accomplish with computers on this planet.
2017 was a helluva year in technology; we’re thrilled to see what 2018 has in store!