Let’s Not Blindly Trust Tech. Just Look at the Horizon Scandal
More than 900 people were prosecuted for stealing from the Post Office because a computer generated incorrect data.
If you have not yet seen Ian Hislop mauling Sir Jake Berry over the Tories’ handling of the Horizon scandal, make haste to YouTube. The editor of Private Eye and team captain on Have I Got News For You gave the former Tory party chairman a proper dressing-down, spawning videos with names like ‘Ian Hislop BODIES Tory MP’. Hislop seems genuinely cross about the scandal, which I suppose is reasonable. After all, more than 900 people were charged for stealing from the Post Office, just because a computer system coughed up bad data. Many were ruined. Some went to prison. Others became addicts and at least a few died before their time. And all because their bosses blindly followed technology.
Well then, we have plenty of fun ahead of us. We are, after all, now a year into the ‘age of AI’, brought on by the launch of ChatGPT. Some people are practically salivating at the prospect of living in a world run by technology: they cannot wait to bring on the so-called singularity. But I am not so sure, and think it might be worth a short refresher course on why we should be wary of technology, one reason for which that it can do physical harm. I am not so much talking about the robotic vacuum that chewed up a woman’s hair in South Korea, as distressing as that sounds. (Really, I am sympathetic.) I am thinking more of how an Air France flight crashed into the Atlantic after the autopilot switched itself off.
It can do physically harm. I am not so much talking about the robotic vacuum that chewed up a woman’s hair in South Korea, as distressing as that sounds.
There is also the risk that we become over-dependent. Human beings, as my father likes to point out, are uniquely useless among animals. ‘A baby cow is on its feet within an hour of being born.’ That baby cow is on its feet because it has to be. We are not because we have people we can depend on. We can depend on technology, too — or at least, some of it. Other than for personal satisfaction, there really is no point learning skills technology renders obsolete — if you can be sure that the technology will do the job well. If you cannot be sure of that, then you should not make yourself too reliant on it. ChatGPT tried to convince me that Waltz with Bashir, a film about war trauma, was a heart-warming bildungsroman, and that its haunting last scene was uplifting. Pity those poor teenagers who think they are beating the system by having ChatGPT do their essays.
ChatGPT tried to convince me that Waltz with Bashir, a film about war trauma, was a heart-warming bildungsroman.
Of course, there are the privacy and security concerns. It would be otiose to mention all the times that a technology company has managed to mislay private data. Last year, OpenAI admitted that it had managed to lose the information of its users, potentially revealing what they had been asking ChatGPT. And though most of us are too lazy to look after our personal data (they call this, in case you care, the privacy paradox) the amount of information that a phone collects about its user is staggering. Are you not just a little bit alarmed when YouTube throws up an ad for something you have just been discussing out loud?
But the biggest problem is that technology is not human. Artificial ‘intelligence’, in fact, is a misnomer. Artificial information-processing is more exact. Machines are not like people, and people are not like machines. Despite the trend for describing nature and nurture as ‘hardware’ and ‘software’, the human brain — possibly the most complex thing in the universe — is nothing like a computer. Only the shallowest analogies work. This matters, because only human beings can understand. Only human beings, useless as we undoubtedly are, can pull together disparate ideas, half-truths, context and all the rest to make sound judgements in complex conditions. And modern societies are nothing if not complex.
