5 Interesting AI Developments From The Past Year

Artificial intelligence (AI) tends to be discussed in fairly broad terms. The concept itself is so fascinating that we often get caught up discussing its limitations or implications, or projections for the future. In some cases, we even look to science fiction such as Westworld and find ourselves discussing the dangers and moral dilemmas associated with advanced AI. All of these are worthwhile topics of conversation – but they also keep us from focusing on the details.

That is to say, sometimes we're so focused on the big picture that we miss out on some really fascinating or even just fun AI stories. So, to play a bit of catch up, here are five such stories that caught our eye in the past year or so.

1. Boston Dynamics Will Sell SpotMini

Many who follow artificial intelligence news will be familiar with a company called Boston Dynamics. The company has become somewhat famous in recent years primarily via a string of semi-viral videos showing dog-like robots that move with incredible intelligence and sophistication. These robots have always appeared to be largely experimental, but we learned just this month that Boston Dynamics is actually going to be selling the SpotMini, which is effectively the smaller version of the dog-like robot. It's one of the more sophisticated AI devices we know of that will be available to consumers.

2. AI Assists Wildlife Conservation

It's not something the public tends to think about very much, but wildlife conservation involves a lot of observation and record-keeping. Basically, people involved in these efforts need to regularly collect and maintain huge sets of data about natural habitats. And we're starting to hear more about AI and deep learning being applied in this area. Basically, cameras can now scan environments and feed information into a system that analyzes the environment and even counts and classifies different types of wildlife. It's an easier and likely more accurate version of what teams of humans or even crowdsourcing efforts might have attempted previously.

3. Machines Beat Humans In Poker

This is actually something we started to hear about in 2017, and while it sounds light and playful it's actually a fairly significant development. Hearing about machines beating humans in games is nothing new. We have seen it happening in chess for years now. The difference here is that, as one analysis pointed out, poker is playing without all the information available. There are distinctly human aspects to the game, and unknown elements to every hand, whereas chess is complex, but played within a firm set of rules and an almost mathematical array of outcomes. AI triumphs over humans in poker are indicative of very advanced machine learning.

4. AI For Military Deterrence

Naturally this is something we don't know a whole lot about, beyond theory and speculation. It's basic logic that militaries around the world will ultimately look to use artificial intelligence, hopefully for defense. But recently we've actually started to see stories about the American military employing AI specifically to locate nuclear weapons and anticipate launches or movement. This kind of deterrence would seem to be a vital application of advanced AI technology, and it'll be interesting to see if we start to hear more about similar efforts.

5. A Robot Writes Music

One big question about AI that remains largely unanswered is how successful it can be in creative endeavors we view as uniquely human. We know that machines can string words together in a narrative, reproduce pictures, etc. – but can they really make art? We're still not entirely sure, but last summer we did learn about a machine writing and playing its own music, based on having learned over 5,000 real songs from human history. The tunes are imperfect, to say the least, but it's still a fairly extraordinary development. We can only imagine a more sophisticated version of the same bot putting together beautiful original compositions.