2016 in review: computers learn how to think
In the tech world, it was a year of robots, fake news and trying to catch a Pikachu. But the major companies have been getting excited about one trend in particular: artificial intelligence.
Imagine a world where you can order a cab on your smartphone, and it will arrive without a driver behind the wheel. Where you can ask a set of speakers to tell you the weather, your appointments, or a funny joke. Where a toy robot has an ‘emotional engine’ which causes it to sulk when it loses a game, or act ‘terrified’ when it almost rolls itself off a table.
That world, of course, is already here. All three products — Uber’s driverless cars, Amazon’s Echo, and Anki’s tiny robot Cozmo — were released either this year or last. And they all rely on artificial intelligence (AI). It is no longer a science fiction fantasy; in fact, you are probably already carrying it around in your pocket.
In the last year, major technology companies like Facebook, Google, and Amazon have all invested heavily in AI. There have been headline grabbing moments, such as when Google managed to beat the world champion at Go, a complex Chinese game.
But they have also been using AI across more everyday products: Google searches are becoming faster and more accurate. Facebook uses it to decide what shows up in your newsfeed. Amazon uses it to recommend products. And all three have AI-powered ‘virtual assistants’ which can answer your burning questions.
Why now? Much of AI’s progress this year was thanks to something called ‘machine learning’. Instead of employing people to write millions of lines of code to respond to every possible situation, machine learning allows computers to teach themselves — just like human brains.
And it is not just about cute toys or better internet searches. Experts say that machine learning could eventually solve far more complex problems. It is becoming so powerful that this year, five of the major companies joined together to create an ‘ethics board’ to make sure that AI ‘is both trustworthy and beneficial.’
AI works by collecting as much data as possible. This is how it gets to know our interests, routines, even the sound of our voice. But is it too controlling?
Yes, says the author Rachel Holmes, a former Amazon employee. Tech companies already know what we buy, who we talk to, the embarrassing questions we ask Google. Now they are trying to shape our behaviour with ‘helpful’ recommendations. ‘It’s reminiscent of East Germany under the Stasi,’ she wrote — and the only way out is to log off completely.
‘What an overreaction!’ say fans of AI. Used properly, it has the potential to make everyone’s lives better and safer — by driving us more carefully, diagnosing diseases, or even just by recommending a really good book to us. It is the most exciting technology to arrive since the internet; we should all get involved and enjoy it.
- What was the best new technology you used in 2016?
- Will artificial intelligence change the world for the better? Or will it make life worse?
- Draw your own design for a robot which uses artificial intelligence to help with chores around the house.
- Rewrite the final two paragraphs of this article, presenting two sides of a different argument about AI. For example: can it be trusted? Is it a passing fad? What if computers become conscious?
Some People Say...
“The greatest danger of AI is that people conclude too early that they understand it.”Eliezer Yudkowsky
What do you think?
Q & A
- Can computers really think?
- It’s a complicated question. In a way, yes: they can respond to questions, solve problems, and in 2013 an artificial intelligence bot called ‘Eugene’ tricked people into thinking it was human, passing a test set by the computer programmer Alan Turing back in 1950. But they do not think in the same way that humans do — they are not ‘conscious’.
- I don’t really understand AI — is it dangerous?
- You’re not the only one! Machine learning programs are so complex that even the people at the forefront of the technology don’t understand exactly how they work. But the important thing is that they do — and they are being used for all sorts of different problems. But it is not dangerous in the way that Hollywood might suggest; there is no need to fear a robot army just yet.
- Driverless cars
- The taxi company began trialling these in Pittsburgh, USA, three months ago — although at the moment there are still two ‘engineers’ who sit up front and step in if something goes wrong.
- The device was launched in 2015 in America, and 2016 in the UK. It uses Amazon’s AI assistant, Alexa, to answer questions when spoken to.
- The robot was released in the USA this year, and will be coming to the UK in 2017. It is connected to your smartphone, and its movements and characteristics were designed by a former employee of the animated film company, Pixar, to ensure maximum cuteness.
- Artificial intelligence
- In this story we are talking about simple AI, which is when programs are designed to tackle specific problems — not general AI, which would mimic a human mind. For now, that is still the stuff of science fiction.
- Ethics board
- The ‘Partnership on Artificial Intelligence to Benefit People and Society’ includes Amazon, Facebook, Google, Microsoft and IBM.
- The Ministry for State Security, or secret police, which operated in Communist East Germany until 1990.