When AI starts eating its own tail

You know how it goes. 

You get side-tracked on social media where someone is telling you to buy this, do this, and CHANGE YOUR LIFE! Generally speaking, they've had some sort of epiphany and realised something that has eluded the rest of us for decades.

Depending on your algorithm, you probably get a bucketful of AI slop too - articles culled from elsewhere and repeated as if the person wrote it, with overly bright dynamic photo-illustrations, or click-baity Tik-Tok content, skewed to your demographic, and inciting you to believe something.

It gets harder each time to sift the actual human content from the nonsense. And at this point, you have a couple of options.

Get off social media. Find a reputable news source and fact-check anything you just read. Bear in mind that you can't really trust Google (it's embedded-AI is embarrassingly visible) and Wikipedia is corruptible, and most print media is owned by someone pushing an agenda. Substack felt real for a millisecond before it also got invaded, and your local physical IRL library probably closed. It is increasingly tough to find online sources based on hard data, well researched and without dangerous bias, but there are some out there. They tend to be associated with a university, are non-profit and often a stodgy read, but like porridge for breakfast, you'll feel better after consuming these.

Talk to real humans. Beware that we all live in an echo-chamber because we gravitate towards people who look, think and sound like us, but hopefully you know a diverse set of well-educated folk who can either direct you to reliable sources or provide their own researched opinion. If possible, pick people who work in different disciplines to you - science, art, politics, education, economics, you get the idea. It's not that they will intrinsically have the answers, but a wide range of influences are more likely to throw up counter-arguments and additional perspectives.

The reverse of this is that AI is starting to be fed by more AI. Like the snake eating its own tail. Here is where it goes really wonky. Extremist views or odd behaviours get amplified and presented as if everyone believes them or is acting upon them. 

Outrageous scaremongering probably infiltrates my feeds about once a month, until I move ahead of its curve. The American dream of AI that does all the drudgery for us, leaving us to live lives of pleasure is obviously fake, (until we have more sophisticated sensitive robots anyway) but most of us are not immune to something doing a bit of work for us when we're overstretched - whether that is a fridge that will place an order for food staples or Co-Pilot writing that dreaded email. It's tempting.

But don't be fooled into thinking you are getting real help. 

AI thinks for you, so you don't have to. See the problem there? Our attention spans are getting shorter, children are not learning handwriting, we are forgetting routes we've driven, poetry we loved, recipes we once mastered. Our brains need those daily challenges to stay elastic, or it becomes too much of an effort to read a book, or even a long paragraph, or remember any instructions. And a dumber population is more easily controlled or duped.

People trying to make a quick buck will tell you EVERYONE has to get to grips with using AI. If they are selling you an AI training course, especially. 

But actually we need to preserve our very human qualities - our lived experience, empathy, nurturing, problem-solving capacities and our ability to rationalise and think critically. Our ability to write poetry and music and art, of course, but also to preserve the reasoning behind why we have those things. Not just as commodities, but as learning tools, creative ways to build tolerance, understanding and innovative thought. To understand other humans, stop wars, feed our citizens and look after the planet.

Don't let the AI Ouroboros take over. Just because it's easy, doesn't mean it's helping us.