The Doctor Jeckyll and Mr Hyde of AI and what investors should do

There is a good, a very good, a bad and a downright evil side to technology. What can investors do?

Article updated: 8 October 2019 10:00am Author: Michael Baxter

When humanity discovered how to split the atom we thought we had learned how to create almost unlimited energy, but we also made the atom bomb possible. And so it always is with technology — it’s good, bad and evil.

Take AI. I read recently that AI is now as effective as doctors at medicinal diagnosis. There are two points here. First point, if AI is as good as humans at medical diagnosis today, imagine how effective it will be as it advances. The second point is that the study was limited — and actually, most experts agree that we are nowhere near the stage where we can rely on AI without human supervision for diagnosis.

But that’s not the issue, healthcare is one example of how AI could have a truly transformative effect upon us and very much for the good.

AI is being used in energy too — as renewables take up an ever more important part of the energy mix, AI can help solve the problem of intermittent energy supply, by channelling energy generated when conditions are favourable for renewable energy into non-time sensitive applications such as storage heaters.

AI is being used to help automation in the services economy. Technologies such as robotics process automation, which lean on AI, and will increasingly lean on AI, provide the potential to revolutionise the mundane, super boring tasks, that are currently carried out by humans in banks, insurance companies and many other process driven operations.

Then of course, there is the application of AI with autonomous cars.

AI is set to transform the economy. PwC has projected that AI could contribute $15.7 trillion a year to the global economy by 2030. Okay, that’s eleven years off, but the growth trajectory between now and then will be phenomenal.

The bad and evil

Yet AI has a nasty side.

For one thing it can be bigoted. We know that if the AI, or let’s call it machine learning, is trained on data that is not fully representative, it can throw up algorithms that can be racist or sexist, for example if AI is used to sort CVs for a job application.

For another thing, there is cybercrime. Gone are the days when cyber criminals worked from their parents basement, eating McDonald’s takeaways all day long. Today, cyber criminals are ‘professionals’, they work in teams with other ‘professionals’ and they are experts at using AI.

Then there are deep fakes. This is where AI is used to doctor a video or sound recording. The classic example is of a video of Barack Obama, apparently, saying things we just know he wouldn’t say in real life.

The fear of course, is that someone could deep fake the US president threatening another country.

Recently, it was reported that the CEO of a UK energy company transferred around £200,000 after he thought he had received a call from the boss of the parent company. In fact, an AI algorithm was used to help impersonate the boss’s voice.

Then there is surveillance, and the use of our personal data in ways that can manipulate us. Thanks to AI, we have never been closer to living in the kind of world envisioned by George Orwell. By the way, the EU’s General Data Protection Regulation — or GDPR — was introduced to try and counteract this danger.

Then there are the military uses of AI. Not many AI experts take the risks of a Terminator type scenario that seriously. But the risk of AI controlled weapon systems designed by humans, taking actions that common sense would avoid, are real. As are the risks of AI drones — amoral, no sense of right or wrong. These risks can be reduced by the use of a so called off-switch. But the idea of a war in which AI plays a key role is terrifying, not that war without AI isn’t frightening enough.

As for investors. Well, don’t ignore it. Fortunes will be made and lost from AI, but I suspect that very few companies will flourish over the next ten years unless they make use of AI.

Investors could put their money into AI companies, but they need a wider approach. Look for companies that genuinely seem to be adopting AI. If you can find a company you already like, and then you discover it has extensive plans to apply AI, then that might be a company worth looking at. I note, for example, that Unilever often talks about its use of AI. That does not mean it’s necessarily a good investment — there are a lot of considerations, but it is an interesting point.

But I would introduce one caveat. There is a lot of hype and nonsense spoken. Many companies say they use AI, when in fact they don’t — it’s just marketing hype. Do your homework.

These views are those of the author alone and do not necessarily reflect the view of The Share Centre, its officers and employees

Michael Baxter portrait photo
Michael Baxter

Economics Commentator

Michael is an economics, investment and technology writer, known for his entertaining style. He has previously been a full-time investor, founder of a technology company which was floated on the NASDAQ, and a director of a PR company specialising in IT.

See what else we have to say