Artificial Intelligence could spell the End of The Human Race.

At least, that’s the scenario put forward by Professor Stephen Hawking in a BBC interview recently. Since 1962, when the first robots started work in the industrial workplace, scientists have been working to increase robotic intelligence levels. Having them perform more and more complex tasks.

The discussion was brought about, when Professor Hawking was asked about Swiftkey, a British firm which has helped upgrade his communications system. A theoretical physicist, he suffers from motor neurone disease, and Intel have improved his ability to speak with a new, more advanced, transmission system. Swiftkey is heavily involved in research work into the nuance of dialects.

He expressed concerns about the ever increasing intelligence levels of machines, and of machines developing ways to increase their own intelligence, without human interference.

The fear of AI taking over from humans is not a new one:

A recently released report by two Oxford academics, in a paper named ‘Robo-Wars: The Regulation of Robotic Weapons.’ called for guidelines into 21st century warfare being carried out by robots. Weapons such as highly sophisticated missiles. These would be able to roam an area; seek a target, and detonate without human interference.
ID-10061012
However, while agreeing Professor Hawking’s fears are justifiable, Mr Ben Medlock, chief executive and co-founder of Swiftkey, believes the day when AI equals that of humans is a long way off. He continued in the interview, “It’s our responsibility to think about all of the consequences, good and bad. We’ve had the same debate about atomic power and nanotechnology. With any technology, there’s always the dialogue about how do you use it to deliver the most benefit, and how it can be used to deliver the most harm.”

Ben Medlock is far more dubious than Professor Hawking about the speed with which this artificial intelligence technology can be produced. He continued the debate saying: “If you look at the history of AI, it has been characterised by over-optimism. The founding fathers, including Alan Turing, were overly optimistic about what we’d be able to achieve.”

While single functioning tasks, such as translating foreign languages, have already been achieved, he believes that machines which are capable of replicating the human brain are a long way in the future. Mr Medlock’s appraisal though, seems to be in the minority, could he be playing down the speed with which his own company is developing artificial intelligence?

Google join the AI club:

It’s not just British academics and computer scientists voicing concerns over the development of AI. Big bad Google, which recently bought the British artificial intelligence company, DeepMind, has already set up an ethics committee to look into such issues.

The founder of DeepMind, Demis Hassibis, says he only sold the company to Google, on the understanding the technology won’t be used for military purposes. Of course, the final outcome will depend on the findings of Google’s own ethics committee, and whether they change their findings in the years to come.

The fear is of an artificial intelligence which can learn faster than the human brain. We already have a basic computer brain in ‘Cleverbot’. A system designed to interact as a human does.ID-100216518

Cleverbot has the ability to learn from previous conversations. Its creator, Rollo Carpenter, says he believes humans will keep ahead of AI for three or four decades. His software though, has already scored highly in Turing tests, fooling many into believing their conversation had been with another human.

Mr Carpenter freely admits no-one knows what the outcome of super fast artificial intelligence will be. Whether we will be helped by it, ignored, or eventually destroyed by it. Many others in the computer and high technology industries are also voicing concerns of the ever increasing advances in AI.

The rocket manufacturer, Space X’s chief executive, Elon Musk is quoted as saying, ‘AI is the biggest existential threat to the human race.’

With quotes like that, it makes one wonder just how far down the line of super fast artificial intelligence, scientists have got, and how much they are not telling us. Will the first thing we know about it be when the latest robot has taught itself how to pick the lock of its laboratory. How to load and fire a Kalashnikov, and how to hunt down, and eliminate its maker.

Images care of: http://www.freedigitalphotos.net/images/search.php?search=Robots&cat=


Share with your friends
Facebooktwittergoogle_plusredditpinterestlinkedinmail
To report this post you need to login first.

3 Comments

Leave a Reply

Your email address will not be published. Required fields are marked *