The AI hype (more like AI hysteria these days) knows no limit. On the one hand, everyone and their grandmother are racing to implement AI in their business. On the other hand, the same everyone is worried about the consequences of AI. AI will soon dominate the world according to experts (and politicians) convened in Davos recently. Seriously, let’s get real. It’s just a computer.
According to the IMF chief – on stage in Davos, AI will increase productivity (good), increase global growth (good) and increase income everywhere (good). It will also increase polarization (dangerous), spread misinformation (even more dangerous) and take away jobs (bad now, good later). World leaders, experts, pundits and heads of AI companies all sing the same song. AI is great but must be reigned in, brought under control or very bad things will happen.
This is scary. Not the technology (AI), but the attitude. And the fact that all these influential people team up and sing the song Sam Altman (of OpenAI) composed last fall. Doesn’t it occur to them that he’s playing them like a fiddle? He can do that because he’s the only one of them with a real, deep understanding of the technology. The rest, Microsoft’s Nadella included, are buying the gospel and falling in step, scared to appear out of touch or among the ‘unenlightened’.
This allows Altman and half a dozen top level AI-insiders to develop a narrative in which they remain the uncontested leaders and experts. They will be consulted – they literally hold the pen – when regulations are drafted, effectively keeping competition at bay. It’s The Emperor’s New Clothes all over again. Can someone shout out real loud that AI is just a computer? Actually many computers. And that computers have taken over the world long time ago. Even light bulbs have small computers in them. This isn’t magic.
AI is simply another turn in the spiral of technological development – enabled by ‘bigger everything’:Â More memory, more data, more CPUs, faster everything – and developers taking advantage of all these resources. It couldn’t have happened before because the resources didn’t exist. It happens now because they do. A lot of technical creativity and ingenuity delivering incredible results, lots of bugs and no (human style)Â intelligence (check out Help! AI is Hallucinating!).
Actually, the term AI or ‘Artificial Intelligence’ is a big part of the attitude problem because it creates false expectations – a falsehood conveniently maintained, sometimes even inflated by the AI companies. It creates an unfortunate image of ‘thinking machines’ in the minds of most people, which – helped by creative movies and TV shows – overshadows reality: That AI is (big) software running on (big) computers with smartness determined by the data feed to it and the programmers doing the development. When explained, most people will understand this – conceptually, but it doesn’t take away the scare, the perception of ‘thinking machines’ with ‘brains’. Which brings up another unfortunate misnomer: We’ve been calling computer chips (CPUs) ‘brains’ – as in ‘the CPU is the brain of the PC’ — for decennia.
Pointing out that digital brains have less intelligence than that of an ant doesn’t stick because ChatGPT & Co. seem so capable. They can (apparently) write essays and novels, answer complicated math and science questions, paint/create lifelike pictures, drive machines and deliver medical diagnoses.
It occurs to me that ‘2001: A Space Odyssey’ may turn out to be the most influential movie of all time. Everyone in Davos – and leaders/politicians all over the world – have seen it and think of AI like Hal in the movie. No wonder they’re scared. They’re also disconnected and – as we’ve seen – vulnerable to being played.
What is it with AI? the setting, the scare, the misunderstandings should remind us about computers 40 years ago. Something new that most people don’t understand. And what people don’t understand has always been scary – since Spinning Jenny, even long before that. People (and leaders) were worried about computers replacing them at work – at all levels, from assembly lines to research and science. They did – and it hurt, but not for long. A few years later (OK, 15-20 years), everyone had moved on and computers were part of most people’s lives, visibly or out-of-sight.
40 years later, most people – even in third world countries – carry around 10s, maybe 100s of computers all the time – in phones, tablets, pacemakers, fitness trackers, watches and thousands of other variants. While computerized (aka ‘intelligent’ or ‘smart’) drones and missiles and torpedoes and all kinds of weapons kill more people in less time than ever before.
Given this near history with computers, why is AI so scary? Partly because we’re slow. Technology evolves so much faster than we do and most of us are increasingly falling behind, losing contact with our own reality. That is definitely not good. It can be fixed, not by reigning in development and restricting use but by explaining, teaching, educating and creating understanding instead of the opposite. Eliminate the (fake) idea that this is something so complicated and incomprehensible that you need at least an academic degree to understand it. A misunderstanding that benefits the Altman Choir and their investors but no-one else.
How? After all AI IS complicated beyond comprehension, isn’t it? Technically, yes. But so is a computer. My $10 RaspberryPi Zero-W is a miracle with more than a billion small switches, lots of memory and huge software inside. So is your phone and your TV. The challenge isn’t to understand the technology but the concept – how and where it may be useful and what the limitations are. You wouldn’t use your smartphone as a hammer or an oar but you would easily use it for payments – because you understand the the concept. It’s a computer with a screen – and we’re all used to computers. It has radios that are connected to some antenna somewhere and can talk to the bank because you’ve authorized it to.
Familiar elements today that would have looked like magic 10 years ago, some of them even 5 years ago. In a few short years we’ve acquired sufficient new understanding to feel it all ‘natural’, even indispensable. How can we move AI from magic and mystery to a similar level of understanding? Is it possible?
It is – and here’s the thing: No one have even tried yet – for the same reason and in the same way that the high priests of the (old) datacenter (I was one of them) kept the layman away for decennia: We talked in code (‘tribe language’) among our brethren, felt good when impressing ‘outsiders’ with obscure technical terms and made no attempt to simplify and create understanding – more like the opposite.
IOW – the first big step towards a more general understanding of AI is a change in attitude. From everyone – the Altman disciples, the developers, big and small players, the data scientists, the leaders, the politicians, you and me. They – the technical side – must want to create understanding, the rest of us must want it. Like with the radio 70 years ago, then the TV, the hi-fi, the PC, the phone and so on.
Since many of you – my readers – belong on the technical side of the aisle and presumably understand much more of the underlying ‘AI magic’ than most people, here is my challenge to you: Become active in tearing down the ‘magic curtain’. Take the time to explain the basics, create understandable allegories, ‘pictures’ using everyday objects and situations as illustrations. It’s not done in a few minutes so take one step at a time. Remember, the goal isn’t to create engineers but to remove anxiety. E.g. don’t try to explain ‘data processing’ until the concept of ‘data’ is understood. And data – simplified – is like a bunch of documents in a drawer or a binder, right? It’s mundane but it works, and in this context, the real magic is simplicity.
I’m sure you’re getting my drift, and if I managed to arouse your interest, you’d want examples. I’ll get to that in upcoming posts. Examples using a variety of everyday objects and concepts to visualize the concepts – like light switches and drawers, databases and movies, X-rays and doctors, Shakespeare and Tom Clancy – and so on.
If you’re like some of my previous audiences, you might even find some of them entertaining. Stay tuned.