Google has made some great strides in artificial intelligence over the past few years. From refining it’s search function through AI inputs to open sourcing some of its most sophisticated AI code, the tech giant seems intent on leading the charge in machine learning.

View Companies Related to Your Industry

Explore companies, products, and service providers on TechEmergence:

Even in the last few weeks Google has conquered some AI milestones surprisingly soon. 

One of these milestones was Fan Hui, the European Go champion. If you haven’t heard of Go you’ve probably seen it played in the background in Chinese films. The ancient boardgame – created more than 2,500 years ago – involves two opponents trying to claim more territory than the other by surrounding each others pieces with their own. Go is abstract and more complex than games like chess or checkers because there are many more potential moves in Go at any given time. There also no simple, objective way to determine if a position is advantageous, so it takes players many years of practice and a whole lot of intuition to master the game.

This difficulty in mastering Go has left many people amazed that a Google computer program called AlphaGo is apparently capable of defeating some of the world’s best Go players. The European Go champion was defeated five games to zero. Next months AlphaGo goes on to challenge Lee Sedol, who is recognized as one of the world’s best players.

AlphaGo itself is pretty sophisticated according to a report published in Nature. In fact, it used two deep-learning networks to predict the opponents next move and to predict the outcome of the game from different board arrangements. AlphaGo was developed using a unique machine-learning method by Google DeepMind, the group created upon Google’s purchase of the startup DeepMind two years ago. The machine learning method – which has proved useful for image and audio analysis – combines deep learning technology with a simulation method used to model potential future moves.

And this week, the National Highway Traffic Safety Administration (NHTSA) told Google that it regards the company’s self-driving car’s computer systems to technically be “drivers”. When California ruled that Google’s autonomous vehicles had to have a driver behind the wheel, Google’s desire to create a driverless, wheelless vehicle his a road block. But the NFTSA’s recently letter accelerates Google ahead.

In their letter to Google, the NHTSA clarifies this issue by writing that, without a human behind the wheel, the next best logical solution is to classify the systems steering the car as a driver: “If no human occupant of the vehicle can actually drive the vehicle, it is more reasonable to identify the ‘driver’ as whatever (as opposed to whoever) is doing the driving. We agree with Google its SDV [self-driving vehicles] will not have a ‘driver’ in the traditional sense that vehicles have had drivers during the last more than one hundred years.”

These two milestones validate Google’s recent efforts in the realm of AI. On one hand, AlphaGo’s success against world class Go plays shows how quickly AI programs can advance. Meanwhile, the NHTSA’s classification of self-driving systems as “drivers” marks a turning point in allowing autonomous vehicles to hit the road.