The Rise of the Machines

  • Posted On: 12 June 2023

Artificial Intelligence (AI) has been very much in the news lately, with views from that akin to Dad’s Army’s Private Fraser (`we’re all doomed’) to how beneficial it will be to mankind, with increased leisure time and improved health for all.

It reminds me of an essay written in the 1930’s by John Maynard Keynes saying that in the developed world the grandchildren of that generation – my kids as it happens – should enjoy a 15 hour working week, thanks to mechanization and computerization, with more time to spend on the Arts and self-improvement. It wasn’t a forecast so much as an aspiration, but of course it hasn’t quite turned out like that – whilst we have lost many of repetitive and hard manual jobs (often simply moved to other parts of the world) for many the world of work remains an unrewarding and harsh place with long working weeks and poor pay.  Even the gig economy has not delivered on the individual freedoms it initially promised.

Now it seems another revolution is underway, led by AI’s, and this time it is white collar work that is likely to go – customer service centres replaced by chat bots, expert advice by Chat GPT – all for our benefit.  In practice what we have seen so far is a simple shift in the workload from the provider of services to the user of services - from scanning our own shopping, booking our own holidays, to doing our banking on line – without any benefit of advice or help, taking up our limited leisure time – and this somehow empowers us. When was the last time that you, as a consumer, managed to speak to a major service provider?

I read very recently that the results from another survey (never a reliable guide by the way) about banking that 20% of people aged 20-34 would prefer to deal with a local branch than being on line. This is astonishing – this is the generation reared on the internet – and yet 1 in 5 would rather deal with people? It was of course hailed by the bank as justification for the increased closure of branch’s, presumably because 4 out of 5 in that age group aren’t bothered.

It’s coming to the world of insurance as well, and in many areas, it’s already here. In Trade credit insurance API links between client and insurer have been around for years, and one insurer very recently launched a wholly integrated on line offering. No need for help or advice here, and this is undoubtedly the way the market will go, at least if the insurers have their way (that it makes changing insurer more difficult is, well, just coincidence). For the actual underwriting much has been done by algorithm for some time, a trend that AI’s (not sentient by the way, but learning machines) can only exacerbate. After all, computers work 24-7, they don’t have bad days, they don’t show up late when the home team wins, they don’t have bereavements, or child care issues. They’re easy to manage, and unbiased.

Well, not quite, they carry the bias of the programmers of course, and also will extrapolate trends from slim evidence as anyone who buys credit information should know. Corporate failure is notoriously hard to predict, and not just because insurers are forecasting based on historic accounts, but also because businesses often fail because of the human factor – from simple bad decisions to investment or banking `bubbles’ that burst overnight.

We provide an advice and broking service. We’re here to make sure clients understand and get the best out of the insurance cover they buy for their business. I for one am incredibly proud (as modern argot has it) that we provide this support for our clients. Yes, we have to deal with on line systems, but primarily we deal with people. Long may that continue.

«February 2024»