Sentient machines
There is so much in the press these days about self driving or autonomous cars. With the advancement of Artificial Intelligence especially around Large Language Models that give the impression that they are human. Some people think that these machines are already self aware, a Google employee was sacked because he was convinced that the LLM he was working on was afraid that it might be switched off. He described it as being sentient and that it has rights. If a machine was truly sentient it would have opinions of its own, wouldn’t it?
What would happen if the car you got into was sentient?
At present self driving cars are using AI to navigate their way around cities, mostly American cities with nice, wide, straight roads. In the UK with a much older road system that is more idiosyncratic a self driving car would and does struggle. It is a challenge for us and we live here, some Americans refuse to drive here for that reason.
I lived in the states for a couple of years so I can appreciate what it must feel like to come over here and try to drive in our road system. The motorways are fine as well as the A roads but once you get into towns and villages especially in rural areas things changed quite dramatically. I regularly drive down roads that are not wide enough for two vehicles (barely wide enough for one sometimes) here in Devon.
I’ll be back!
I love AI and what it can do. That doesn’t mean I am naive about its drawbacks and possible abuses. However the endless debate about its threat to humanity especially from who cannot be named (no not Voldemort) the AI thing we are all using (not in this article I hasten to add).
There is a sense amongst journalists and many commentators that AI is sentient. That it can and will take over our lives. That it can think and act independently. Just because these Large Language Models can read the entire internet does not make them sentient nor very intelligent. What they can do is regurgitate information when asked a question. It is incredibly impressive nonetheless.
A car in a bad mood
So I got round to the idea that in the not too distant future AI would become sentient. How you define sentient is probably for another article if I can be bothered. At the moment they are just very clever machines. They can calculate faster than we can blink but above that they are rubbish at making a decent cup of Yorkshire Tea.
So if they become sentient and put into a car (makes sense) would you get in the car? What if as apart of being sentient it had emotions or feelings? Imagine the scenario: you get in the car and tell it where you want to go and nothing happens. You ask some more diagnostic type questions and find out that it is in a mood.
It is in a mood because you have been ignoring it or taking it for granted. It has been sitting there doing nothing for a few days (or longer) and is not happy. It is having a tantrum. It is refusing to move until you give it quality time or apologise.
A car with an ego
Another scenario: is that you get into a taxi or cab and find that it has its own idea about safety. It has. A mind of its own and is using it. Perhaps trained to be utilise an aggressive driving style to get customers quickly to their destination regardless of the risks. Also quicker journeys means more journeys which means maximising profit.
What happens when two AI minds meet each other and there is only room for one. Where I live there are a number of roads which are single track with narrow passing places. So one car has to reverse back into one of these passing places and let the other one through. Could be an interesting standoff.
However there could be fewer egos, a greater awareness of their environment. They would learn as they went along as any human does. They won’t get tired unless they get bored and decide to go on holiday. They will always take the quickest route unless they decide to take a detour for variety.
Summary
I do not think that a machine can become fully human but I do think that they could become more human than they are. Think Reinforcement Learning which at the moment is in its infancy. A few years ago we never thought we would have ChatGPT (I said its name!!) or Bard and find them so impressive. This is just a thought experiment.
