28/12/2017 14:50 GMT | Updated 28/12/2017 14:50 GMT

The Hype And The Hope For AI

As 2017 draws to a close it seems we are reaching a paradoxical time where many are placing AI front and centre in their 2018 trends list whilst others are finding AI to be over-hyped.  

The sad reality is that many companies are trying to make their newest product sound cutting edge by using the term artificial intelligence to mean just about anything a computer can do, take for example this child’s toy for Christmas.

Newspapers and magazines love to put out sensationalist stories on AI, buoyed by futuristic movies from the likes of Tom Cruise and Will Smith, meaning the term is thrown around casually across the world. Many are now not even clear on where we are with AI presently, never mind where it’s likely to take us in the future.

Marvin Minsky, one of the bigger names in the AI field (he co-founded MIT’s AI lab), has called “intelligence” a “suitcase word,” meaning you can stuff anything in there. In other words, it’s too broad and it doesn’t lend itself to precise definition. That’s fuelling the hype but also the confusion. The lack of precision around the term has become part of the issue.

The lessons from history

What a lot of people probably don’t know about AI is that it debutedback in 1956.  Much like today, this was a time of huge optimism and millions were invested, many believing AI would be doing the work of humans by the mid-1970s… of course, that didn’t happen.

The cautionary tale from history comes from what happened next. When expectations weren’t met with results, funding dried up and a lot of research ground to a halt leading to “The AI Winter”. It was a time when disillusionment with AI was so widespread it led venture capitalists to use the term “AI” as a synonym for “bad investment decisions.”

Now we’re at an interesting time... the sophistication and access to technology has advanced exponentially since the 1970s and there’s much more hope for AI this time around.

However, this time round, there is also a lot more hype, and as history teaches us, that can be very dangerous.

Is the hype deserved?

That depends on who you ask, of course. But consider this quote from an article about the evolution of Watson, probably still the biggest name in AI right now:

“IBM Watson is the Donald Trump of the AI industry—outlandish claims that aren’t backed by credible data,” said Oren Etzioni, CEO of the Allen Institute for AI and former computer science professor. “Everyone—journalists included—know[s] that the emperor has no clothes, but most are reluctant to say so.”

Strong comments, however, other major tech companies ― ones that have far surpassed IBM in the last 10 years ― are retooling their business models around AI, including Google, Facebook, and Microsoft. Some smart people work in those places, so the hype must have more than a kernel of truth in it, right?

Is the AI hype bad or good specifically?

You can argue either side here. If it results in a second AI winter where the entire idea collapses as a scalable concept, then the hype was likely bad; full of sound and fury, signifying nothing except businesses returning to the way they’ve always done things.

The good news: funding for AI startups by VCs has surged and shows no signs of slowing down, meaning a lot of money is being thrown at the research and development side of the equation. That is a benefit of the hype.

A drawback, however, is what if all the hype and the videos of robots doing box jumps and the articles about losing your job in five years are actually preventing real advancements in the AI field by researchers? What if the hype is creating too much of a drive for money and attention, and that’s curtailing what could be the next big advance in human evolution?

“The stock market has predicted nine of the last five recessions”

That’s a quote from Paul Samuelson, the first American winner of the Nobel Prize in Economics. Most of the recent winners of that award have been from behavioural economics, which tries to turn economics on its head and realise that people aren’t completely rational.

With this in mind, the truth is no-one knows what’s going to happen but the significant changes are unlikely to be coming in the short term.

What we do know is it’s easier to imagine a time when AI will take our jobs rather than work out how our jobs may evolve. If you tried to explain the concept of an iPhone to someone in 1975, they would have no idea why that would be necessary and would probably laugh directly in your face. Now, it seems normative.

Let’s just hope the current hype wave isn’t working against true growth for AI’s potential.