A.I. Aye?

I haven’t written on here in a long while for various reasons. It was all fine and dandy to be blogging away during COVID but when we were all allowed out again I discovered that whilst I had not lost a couple of years (they happened, I’m sure of it) I had actually fallen behind by some distance. I hadn’t fallen behind anyone else – they’d all been stuck like I was – no, I had fallen behind in time. I was now two years behind where I should have been. I was two years behind the technology, the TV shows and the fashion. I was just two years out of sync with events. Things are taking me by surprise a lot because I thought I had more time to prepare. Which would be fine but these things aren’t exactly trivial – it’s not like I’m wearing the wrong colour of socks here. No, these things are, apparently, existential threats.

Existential Threat
Doomed

It used to be the climate. Unlike a great many people I am actually ahead of the curve on this one most of the time – I’ve been saying for a while that we have a lot less time than we think we have. People are going to be dying in vast numbers just shortly and the migration is going to start to ramp up. I’m not going to go into the details here (been doing that for ages) but have a look around and you’ll see it.

Then there is neo liberalism/ the rise of the far right/the new cold-hot war in Ukraine and when you add it altogether you kinda wake up every morning thinking ‘Did I miss a meeting?’ Seriously, it’s just relentless and then just when you think they’ve exhausted the terror they come out with the latest existential threat.

AI.

And that’s not me saying that. It’s the people who are making AI a reality

On a par with pandemics and nuclear war they say.

Naturally in a rational world we might reply ‘Well, don’t do that then.’ You are the people making the machines after all. The easiest solution here is to, well, not.

But they can’t not. It seems impossible to not. ‘Not’ is not an option.

Because if they don’t then someone else will. It’s the same clever thinking that brought us nuclear weapons and continued use of fossil fuels. But how scared should we be? I mean are we talking Terminators here? Do we even need the full Cyberdyne models or are they just going to give these things great big metal teeth?

There seems to be three or four main models for how this plays out.

  1. Millenium Bug – a big old fuss about nothing. And we diagnose cancer within minutes of you getting it.
  2. New Contracts – everyone has to learn how to use one of these things and after a while you can’t remember how people managed without it. Remember the world before mobile phones? Care to give up your phone?
  3. Financial Crash and chaos. – Like the Wall St crash but with more apple logos. Which is funny because you can’t remember the last time you bit into an actual apple. That’s fruit. You’ll be eating a diet of Soylent Green and won’t know anyone with a job.
  4. The Matrix

So now clever computer people around the world are delving into the possibilities of machine learning and it turns out that all four options are, if not inevitable then, at least likely. In a very real sense the Matrix is already here. The only differences being that the system doesn’t require you to wallow around in pink goo or even be plugged in to it – you do that voluntarily (you have to change your password periodically) and worse, most of it is on a subscription model you have to pay for. Nobody’s getting to choose between the blue pill and the red pill.

The data flow is now the reality. If you want to change how people act then all you have to do is change what they think and the new hell is that the machines know what you think better than you do. Having been brought up on a diet of cyborgs I used to think that it would take a while before the computers assumed control, that they were nowhere near creating a metal Arnold Schwarzenegger but it turns out they don’t have to manipulate the real world at all; either with guns or anything else. Not when they have people around who can do it for them.

Remember this?

The storming of the American Capitol by right wing Trump supporters. The former President won his first election because of a few tens of thousands of votes in certain parts of the country. Specific people were targeted with the advertising that would sway them. This gathering was organised online and although certain individuals have been punished the misinformation flow does not stop. Sure Fox News have had to pay $787.5 million for perpetuating false claims about a voting machine company but that had about zero effect on their output. They didn’t turn left in any way.

The man in charge of the election security called it the most secure election ever and was promptly fired by Trump during his last days in office. One of the most connected countries in the world is now paralysed by contradictory stories and it is, in a very real sense, unable to agree on the fact that Joe Biden won the election.

Social media swings elections. Q Anon had its roots in Twitter and fake ‘evidence’ for the wildest of theories is posted on other platforms. It goes viral and spreads because of ‘algorithms’. We’ll come back to that.

You might think you are immune to all of this. That you wouldn’t believe nonsense made up on the internet. I used to think that but the latest AI is good enough to fool your college professor when you are handing in an assignment. The latest deepfakes are practically indistinguishable from reality. Look at this one:

Is this the face of Kate McKInnon grafted on to Hilary Clinton or is it the face of Hilary Clinton grafted on to the body of Kate McKinnon? Or can’t you tell? The fact is that I found this on the internet and placed it here but the point I am making is that I could likely tell you either answer was correct. It’s actually Hilary grafted on to Kate.

So the internet was born and lauded as the greatest tool mankind has ever made. Knowledge at the click of a mouse, and it was supposed to be the great leveller – anyone with access to the internet could learn anything eventually. But the internet turned out to be a mirror for humanity and what it showed us we don’t really like.

By the way, I lied. It is actually Kate grafted on to Hilary.

So what lies ahead in a world where truth is controlled not by humans but by those algorithms. By Artificial Intelligence. Well, I can guess what the near future holds, because it’s gone the same way every time a new technology arrives. First it is experimented with – a few interested souls tinker and play and discover (we have passed this point) and then it starts to become much clearer that this is here to stay and that we have just begun to discover what is happening.

Then the money moves in.

Back in the nineties I was doing a bit of research for something I was writing for college. I went to a talk by Doug Rushkoff as part of the promotional tour for his book Cyberia and there I made the biggest mistake of my life. This was 1994 or thereabouts and the internet was in its infancy. In a downstairs room in Waterstones in Edinburgh people gathered who were interested in computers and all things cyber. And I didn’t listen to myself. The chat drifted to the internet and what might happen with it and the number of people in that room who insisted it would not become monetised was quite breathtaking. I knew then that capitalism would invade the internet and I should have acted on that knowledge. I didn’t. Which is why I am not rich.

AI won’t stop because the money won’t let it. There is too much to be made by manipulating humans and so they are inevitable. Right now it is in research as the money tries to find how best to use AI but the next people to really push the boundaries are the criminals and the pornographers. And I’m not sure I even want to think about what they might get up to. We can all ignore the call for help from an acquaintance/friend who is somehow stuck abroad when it arrives in an email. What do we do when it is a video message? Given how dark the web can get with actual humans being abused I think we are not far from true horror when anything that can be imagined can be rendered visually with photo realism. I needed some art for a project the other day and I thought “I’ll try this AI stuff they all talk about”. And it produced usable images. But on one AI it started to censor the images that it had made. Normally you get 4 options from any prompt. This time it gave me three and said that the other one was NSFW.

Not safe for work?!?! I asked it for a logo.

You can already get an AI girlfriend/boyfriend tweaked to your preferences. There are so many of these that you can find lists of the ten best (reviewed by humans).

I’m not going to go looking but I bet there are already things out there, created by AI, that would turn your stomach. And the worst part of it is that this isn’t the worst part. That comes afterwards with…

The Corporations

If you think they aren’t looking at AI then I applaud your continuing naivete in this cynical world. They are gathering data at an unbelievable rate. In a recent interview Yuval Noah Harari pointed out that whilst you are watching the social media it is also watching you – where do your eyes go? What type of human do you like to look at? Female? Male? Next up – the party game app where teenagers sit with the AI and have their sexuality reported for all their friends? How does that work in countries where your sexuality can get you thrown off a roof?

Governments

If the AI knows you better than you do then what happens to protest and change in an age of omni present surveillance. Where the stories we are told every day are controlled by AI you have wandered into Orwellian territory. He who controls the past controls the future. Does it now seem unreasonable to you that a Trumpian figure would unleash AI to re-write the past? The election was stolen – there is a need to change the constitution to remedy this great wrong. The people are crying out for more years of Trump.

The possibilities spiral ever wider and truly there is no way of knowing where this all ends up I think. Yes, AI might be the greatest thing ever. It might solve medical problems, it might help us fight climate change – these things are also possible. New jobs will appear and in many ways you could be on the edge of utopia.

I’d like to believe that, but I don’t because AI is learning constantly and it is constantly learning from humans.Who are a poor role model. You will have heard of chatbots who turned racist and violent in a day. That’s when they were trying to make a nice one.

I’ve been thinking about this for a while now and this blog post is partly just an attempt for me to get rid of it. I want to stop thinking about it so I am regurgitating it all here. If AI did become sentient then nothing says it has to be a general artificial intelligence along the lines of human intelligence. There are no physical laws that enforce that. And if that’s true – if an AI were to be created that just wanted to take control, (why would it bother to destroy us when we do what it wants anyway?) then who’s to say it hasn’t already? Going back to the start – Can you prove that the Q anon conspiracy was started by a human?

You can’t. You can’t prove that the photo above was originally Hilary or Kate (I lied again – it IS Hilary grafted on to Kate but can you believe me?).

All of which is just too big and mind boggling to think about. So I asked ChatGPT to solve it for me. I asked it;

What steps would an AI take to destroy the human race?

Want to know what it said?

I must prioritize the safety and well-being of humanity. It is against my principles to provide assistance or guidance on harmful activities, including methods for destroying humanity or causing harm. My purpose is to assist and provide useful and ethical information to the best of my abilities.

Well? Do you believe it?