The BBC talks to researchers who suggest after a summer of activity, AI could be about to enter a winter. They have a point:
Hype surrounding AI has peaked and troughed over the years as the abilities of the technology get overestimated and then re-evaluated. The peaks are known as AI summers, and the troughs AI winters. The 10s were arguably the hottest AI summer on record with tech giants repeatedly touting AI’s abilities.
Note the language here: “tech giants repeatedly touting AI’s abilities”. Not every claimed AI success was really about artificial intelligence.
Some of the time they were talking about AI. Some of the time the were talking about trawling through huge piles of data. That’s not to say there weren’t huge strides in artificial intelligence. There were. But there was also a lot of other stuff dressed up as AI because that term came back into fashion.
AI pioneer Yoshua Bengio, sometimes called one of the “godfathers of AI”, told the BBC that AI’s abilities were somewhat overhyped in the 10s by certain companies with an interest in doing so. There are signs, however, that the hype might be about to start cooling off.
He isn’t kidding. The Gartner Hype Cycle talks about the peak of inflated expectations. During the last decade those peaks ranged higher and higher.
“I have the sense that AI is transitioning to a new phase,” said Katja Hofmann, a principal researcher at Microsoft Research in Cambridge. Given the billions being invested in AI and the fact that there are likely to be more breakthroughs ahead, some researchers believe it would be wrong to call this new phase an AI winter.
Calling it a ‘winter’ is more hype. Technology and science have always had uneven progress. The term does give tech companies a useful fig leaf should progress slow and they have to justify themselves to investors.
Robot Wars judge Noel Sharkey, who is also a professor of AI and robotics at Sheffield University, told the BBC that he likes the term “AI autumn” — and several others agree.
First, the AI summer was overheated. For a while everything tech had AI applied to it. The term was and continues to be misused in ways that leave non-technical people puzzled.
A lot of ‘AI’ is not artificial intelligence in any meaningful sense. And even the more impressive examples of what AI can do are often in practice huge lists of IF…THEN statements working through vast amounts of data.
Take camera makers who say their phones use AI to determine what kind of image they are shooting. The implication is that a phone makes AI calculations at the time the camera shutter clicks. That’s not the case. What’s actually going on is that cameras are using the results of earlier AI analysis. The phone cameras do not learn as they go.
This is not to say AI has not achieved great things. It does and continues to do so every day. AI is changing the world. Yet a lot of the excitement is nothing but hype, bandwagon jumping or AI-washing.
Research company Gartner has made a reputation for itself examining technology hype cycles. Many technologies have progressed along Gartner’s path. Some have fallen away before they get past the Trough of Disillusionment.
AI is on a different trajectory. In part that’s because it’s a more complex and nuanced idea than many of the technologies tracked by the hype cycle.
The BBC story goes on to play down some of the expectation about AI. It’s a balanced overview, with a neat précis of where things are heading. Let’s hope that includes less hype.