Read my posts about technology, gadgets, data, privacy and the internet.


How worried should we be about artificial intelligence and robots?

I had completely forgotten that I wrote this back in 2017. It has though aged rather well and is still very relevant given how much AI has developed since its early days. So enjoy…

There’s a lot of talk at the moment about how artificial intelligence, powered by machine learning, may develop. Will we end up in a world where not only have robots taken our jobs but where their artificial intelligence has become dominant, unintelligible to human minds and powering robots to have control over us?

To answer that, let me take you a little into the very plausible future. It’s one where many households have domestic robots. Those robots don’t understand exactly how to behave or follow human orders from day one but require some training up.

The voice recognition takes time to learn orders just as it takes time for humans and voice-activated assistants on smartphones to take time to tune in fully to each other currently.

To speed things up humans have learnt how to provide other feedback too ‘rewarding’ robots that behave the right way with physical responses that indicate approval.

What do these robots do? Providing entertainment is the key appeal, with of course guilty busy parents also getting suckered into buying these robots with some additional blurb about how good they are at educating children in learning how to interact and look after others.

The more scary part, of course, is that we don’t really understand how the robots work. Sure, there are the broad theories that help us know how to train them well. Take them apart and prod any individual piece, and we know how the mechanics of it work. But add all those pieces back together again and the complexity of the overall system is beyond our proper understanding. We keep on getting surprised – sometimes pleasantly, sometimes not – at how the robots behave.

One result of that lack of control are the occasions when robots misbehave and cause problems. Sometimes they get themselves into the wrong place at the wrong time and are seriously damaged. Perhaps worse, sometimes they behave in ways that cause minor damage to humans (speedy brief contact with human skin easily leaves scratches and can even draw blood).

We don’t understand really how they work. We struggle to control them. Sometimes they even cause us harm.

The name for these robots?

They’re called cats.