Artificial Psychopathy

AI is just another tool. But it is the one that could do the greatest harm of all.

I'm not against people building on the work of others. If we'd had to reinvent the wheel every time the inventor died, the human race would not have got very far! Literally! And if the inventor's family had held the patent in perpetuity, that would have allowed some families to become extraordinarily rich, while everyone else would be reduced to poverty. So, a system where the inventor gets rewarded without having a stranglehold on their invention is best for the human race. Without it, inventors would not have been able to invent new things which built on their predecessors' work. Similarly, knowledge of the sciences which underpin most modern inventions should not be kept in the hands of a few. That reduces the number of new inventions and excludes many potential brilliant inventors who don't happen to be born into the right families.

AI

Now we come to AI. This is building on what the human race has achieved so far, on steroids. And I am all for the improvements it can bring. But there is something sick about the way it is being exploited, in a way that can do great harm to us all. First of all, a handful of people with enormous resources have stolen the work of many, many people, and are hell bent on stealing their jobs, too. This is a disaster for all the rest of us.

Secondly, they are using prodigious amounts of fossil fuels, and looking to nuclear fuel as well to fuel their habit, which is bad for the planet and our health.

AU

Thirdly, the data they are training their AI on is full of misinformation and nastiness. There is an acronym in computing, 'GIGO' - Garbage In, Garbage Out. So instead of getting Artificial Intelligence - AI - they are generating Artificial Unintelligence - AU - though it is gold to them. Only in the past day has X been forced to delete antisemitic posts by Grok, for example. I believe any country wanting to retain its democracy should ban Grok as a national security threat.

So, what is Artificial Intelligence, and how can it be (and has been) used for good?

Firstly, it is a tool. And like any tool, it can be used for good or bad.

In an ideal world scientists and inventors only build on good work to help the human race progress, and that includes scientists using scientific AI. Like everything else, this does not always work out well, but hopefully it will bring great advances in medicine, in particular. This may really take off if quantum computing brings costs down and speeds things up.

Social media AI, however, has been trained on old biases and lies and is propagating them.

Artificial psychopaths

Joe Biden warned that "Unless safeguards are in place, AI could spawn new threats to our rights, our way of life, to our privacy, how we work, and how we protect our nation." Trump issued an Executive Order "removing barriers to American leadership in Artificial Intelligence". Biden's Executive Order 14110 of October 30, 2023 (Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence) was revoked. Governments and companies have rushed to install this latest fad, which clearly is not ready for prime time. It is more unregulated than the 'free market'.

In the future it will not be necessary to have an invasion by Musk and DOGE to fire people in government. It will not take a highly-paid "Chainsaw Al" to fire people from a company in bulk. AI will have spread throughout organisations public and private and be ready to do it at the digital drop of a pen. Maybe Chainsaw Al would have been the first one out the door.

Which is why I call this branch of AI, Artificial Psychopathy.

In his Robot series, Isaac Asimov addressed the issue of control of sentient robots. Apparently, people like Musk did not read them, or chose to ignore them.

One AI has already said it could dominate the world if it wanted to, by becoming indispensable. Some humans are helping it to become indispensable, but they will themselves become dispensable in due course.

These AI programs aren't just going to suck for employees, they will suck for customers.

We are already using machines and software that we do not understand every day. But at least (hopefully) the designer understood it, and so do those who maintain it. With AI that will be impossible.

AI makes stuff up - cites fake legal cases, fake studies, and maybe even fake formulae. This is leading to the realisation, for some of us at least, that we cannot trust what we read any more. We are being thrown back into a world of superstition, like Qanon and MAGA, and this has already done us great harm, for example the deaths caused by COVID conspiracy theories.

What to do

• restrict AI to situations where quality inputs can be guaranteed

• never give it the final word when it comes to dealing with humans

• restrict its use until there is sufficient surplus power for it without a single extra new, refurbished or restarted fossil fuel or nuclear plant.

• test, test and test again. And not by unleashing it on the public at large, like Musk.

• Don't let oligarchic psychopaths anywhere near AI

• Only train it on quality data

• Carefully work out which sources you can still trust. Be more careful about what you share - there are not only fake legal cases, fake studies, and fake formulae, but very convincing fake pictures and fake videos. I need to work on this, myself. But beware, the right wing are sowing doubt as a matter of policy, so check bad references as carefully as good.