After the Bell: Mr Ludd and the AI revolution

Many years ago, in a land far, far away, there was a man with the efficiently spelt name, Ned Ludd. At a moment of great industrial change he was able to lead people who felt that they were going to be on the losing end of a massive transformation.

I’ve always felt that he was a bit ridiculous. You can’t stop progress and if you stand up in the wrong river at the wrong time, history will just sweep you away.

I know that’s probably true of AI; that it will, or so we’re told, literally change the world. But I still just can’t get away from a feeling of slight irritation at the fact that everyone seems to talk about it all the time, while hardly anyone is using it for anything really important.

Read more: The AI surge and the magpie effect hypnotising the C-suite

Instead of it really making a meaningful difference, what I see is people just using AI for fun. Often childishly so.

It’s the same for previous breakthrough technologies. 

There is a sort of apocryphal surely-it-can’t-be-true claim that if the first book that was produced on a printing press was a Bible, the second was probably some kind of pornography.

In my experience, whenever a group of men get together to get some new communications device working – whether it’s SMS phone technology in the late Nineties, or WhatsApp in the 2010s, or even walkie-talkies on that Scout camp in Bedfordview back in the Eighties, or (perhaps) that FM radio transmitter in what is now Makhanda in 1995 – once a group of men know that it is working, this brand-new technology is put to the simple purpose of insulting each other.

We men just can’t help it. 

Now it seems that AI is often just being used to produce the worst prose known to machine-kind.

I suspect that some very clever people who have an awful lot of money are beginning to wonder about this too.

The Economist recently ran several stories questioning the incredible valuations of the American companies that are driving the AI surge.

This has huge implications. A huge amount of money, capital in its truest form, has moved into US stock markets based really on a promise. It’s obviously a big promise. Its promoters believe that AI will change the way we live more than anything else in the future (although they don’t seem to be promising that we will all have jobs…).

What seems to be happening now is a sort of rethink about AI, that the idea of just betting on one big model and hoping it wins is falling away. Instead AI will be used in lots of different, possibly smaller ways.

This means that instead of one big overarching model, or even a group of “killer app” models, there will be many different smaller models used for lots of different things.

Thus, instead of one big company controlling technology that will change the world, we will see many smaller AI products that will do different things and be used for different purposes.

Read more: The great schism - where does AI go from here?

Earlier today, BusinessLIVE published an interview with Discovery Bank CEO Hylton Kallner in which he explained how they are going to use AI to, basically, replace personal bankers.

Instead of phoning or WhatsApping a person to ask them to make a payment or sign an agreement, you would simply instruct this AI agent to do it.

I do get the upside – everything would be quicker and more efficient. It reminds me a little of when MultiChoice retrenched its call-centre workers about five years ago. But it did mean you just cannot speak to a human being to fix a problem.

They’re not the only ones. I recently had the same experience with one of the top two cellphone networks.

Now, perhaps it’s just my (temporary) spiritual affiliation with Mr Ludd, but I always find it incredibly frustrating when you have a problem and you just cannot speak to an actual person.

And if you look at some of the big complaints that people lodge very publicly against consumer-facing organisations, it’s often something simple that became very big just because the customer could not speak to a person to resolve it.

So, imagine if there is some kind of problem with your AI banker? This can happen much more easily than you think. I’ve been in situations where I’ve asked someone to do something for me and they’ve actually got in touch to check if that is really what I want to do. Because they have realised something that I missed.

Or, the English language being the slightly absurd thing that it is, there might be a misinterpretation simply because an email or an instruction is not clearly worded.

One mistake, whether by the AI agent or you, could be fixed quite easily. But what if there is a series of them that cascade into each other?

And who at Discovery are you going to phone to sort it all out if just about everything is now done by an AI agent?

BusinessLIVE also reports that some of the other, what you could call “more traditional” banks are going to be moving into this area more slowly than Discovery.

I think that’s wise. And Mr Ludd would be pleased to see that his spirit is still alive and well, at the start of another revolution. DM