AI ethics boards are often cheap cover and theater.

AI ethics boards are often cheap cover and theater.

There’s a move to put AI ethics boards in companies, particularly in AI companies, to review the ethics of artificial intelligence.

That sounds like a great idea, but the problem is when you really look at the makeup of most of the AI ethics boards, they’re made up of people who have an inherent need for the AI to be successful.

Everything AI has ever done that was unethical was done because humans told it to do it.

Even AI agents that are set up to do specific things repeatedly are turned on by a human being.

If, for example, I tell AI to do a deep face of a celebrity in a compromising position, it’s not AI’s fault that I asked AI to do something unethical.

Creating ethics boards is a nice gesture, but there’s no way they’re going to be able to account for every possible contingency that that AI could find itself dealing with.

Government watchdogs are also not going to solve the problem. 

First of all, government officials are notoriously out of touch with how things work. When you hear a government official ask a question like, “Will this go up on THE Facebook?”, you know they have really no idea of what’s going on.

Yeah, the industry itself could build ethical  guidelines, but that’s still not going to stop the rogue agents from doing things.

Moreover, AI’s ability to write code means that now almost anybody can write an AI agent or more AI code in order to build whatever they want.

AI ethics boards, unless they have teeth and are led by people who truly understand how AI works at its core, are not going to be able to do much.

They’re more or less a show to let the public and politicians know that somebody’s paying attention to the complaints, but there’s almost nothing they can do to stop this train.

How can we guide AI to make better decisions? How can we keep AI ethical?

#AI #AIissues #AIethics #business #AISensum #AIagents #technology #responsibletech