AI is dangerous and

we need to be better educated about this.

I do not use AI in my work. It tries to wriggle in sometimes and I have to push it firmly out again. The reason for this is that if AI gets it wrong then the rest of us on the team could put someone's life in danger, someone could die, a building could crumble, a bridge could fall down, a road might not be where it is supposed to be, a fight make break out, someone could be insulted and start anything from a minor altercation to a war or they might commit suicide. No, I am not exaggerating. Communication needs to be accurate - or as accurate as we can make it. There is no room for the sort of errors that AI can make.

This last week someone had tried to rely on AI for information and ended up screaming desperately for help from one of my colleagues. The mess was eventually sorted out and calm has descended again but much was said about AI while all this was going on. People have been warned again. Do not rely on AI to give you the correct answer if your life and the lives of people around you are at risk. Please just do not do it.

Now there is a story in this morning's paper about a lawyer who has apparently relied on AI to support a case s/he was preparing. The court has, rightly, complained it has been misled. The cases "found" by AI do not exist. They do not support the arguments put to the court. Apparently there has been another such case in another state with a similar result. 

This is serious, more serious than the general public realise. Accuracy is very, very important in law. Getting a comma out of place can change the entire meaning of something. There can be no "but this is what they actually meant" in law. It is what is there on the page which matters and nothing else. Relying on AI to provide something accurate in an area which is so important is dangerous. It could mean someone being incarcerated for a crime they have not committed or a murderer going free. Yes, extreme examples but still possible.

The judge in this instance has said that Gen-AI cannot be trusted to provide accurate information. If that is all the judge had to say on the topic I would be surprised. 

We are being told that AI is the way of the future and that we will need it more and more. I am wondering how it will be used and by whom. Perhaps I am wrong but I am beginning to think we may need more and more people who are well educated and can think for themselves because the idea of AI taking over is not going to work well if we really need accuracy - and accuracy may save many situations and lives.