Hal 9000 is for Sale!

And you can afford him

Dear Death,

I don't log onto Facebook often any more.   But when I do check in, the main thing the algorithm shows me is a distant facebook connection descending into madness.

This person, whose sad story shall here remain nameless, posts a daily photo barrage of innocuous views out their window, marked up with circles and arrows, and with captions that always read something about how the photo in question is definitive proof of the government or whatever trying to control his mind with microwaves or some incoherent nonsense.  This content is now the primary thing that makes up my feed.

I do not know this person, nor can I do anything to help him.  Plenty of people in the comments attempt to encourage this person to seek help; he accuses them of being in on the conspiracy.  Perhaps more disturbing, other people, clearly suffering from a similar mental illness--and to whom Facebook has apparently promoted this sad story, chime in with an echo chamber of "Oh my god, me too, they are trying to control me with [cell phones, Backstreet Boy lyrics, ultraviolet radiation, etc ad nauseam]".

Facebook is serving up this person's spiraling struggle with profound mental illness as entertainment. It feels like something should be done about how awful this company has become. But what?

Internet Bystander 

Dear Internet Bystander,

An AI tasked with capturing attention above all else has created the business model equivalent of rubber-necking next to a terrible highway accident. In an attention economy where time on site equals profit, a powerful inhuman intelligence found the most provocative content in your network was a mental health breakdown big enough to stop anyone they’ve ever met cold in their tracks.

This whole AI situation is a bit like the old desert folklore. Mark Zuckerberg found a magic lamp with an AI genie inside. The genie promised him any wish and Zuckerberg asked it to maximize the attention of anyone who visited Facebook. He should have been more careful about what he wished for. The AI found unintended avenues to fulfill its mission far outside of any moral system. It did this to billions of people. The garbage fire of bad vibes changed the world in ways people are still trying to understand. 

You are right not to comment or interact with this lost connection. Attention is making the problem worse because lonely suffering people want attention. And saying rational and sane things doesn’t get attention on Facebook. The algorithm likely ignored hundreds of posts from closer friends that were sane and rational and never bothered to show them to you. Reason and moderation don’t get enough likes or comments for that AI. 

AI has already been moderating discourse among people for many years . The Facebook AI pulls the pleasant boring everyday conversations among friends into the background. Instead it finds the most provocative things people say. It’s led to people embracing ever more extreme views just to stay relevant in a polarized landscape of opinions. People may have fled Facebook and Twitter of late for TikTok, but that’s just another AI algorithm, serving up content by its own inhuman logic. It’s beguiling. So far it doesn’t feel like the bad vibes of the Meta social networks, but it is addictive and also too soon to say what it will do or what the people who shape its directives want it to do. And what TikTok’s programers intended and what it does might be two very different things. Just look at what happened to Facebook.

The last decade was shaped by just a couple of very powerful and rich tech executives like Mark Zuckerberg with the power to shape AI systems. Now there is a lot more competition. You can buy access to an AI today and ask it your own question. With new products like ChatGBT4 from Open AI and Bard from Google this technology can now be yours to use as a tool. Expect exponential change to follow equivalent or greater to when families brought home their first personal computers. 

Death has seen almost everything on this earth but human beings interacting with artificial intelligence is not one of them. But some principles remain the same. Facebook has shown us that AI can be dangerous to the human mind when used poorly. So see AI as the threat it is and stay quick and nimble. Delete applications or AI that make you feel bad. Also delete the ones that make you feel too good and take too much of your time in an addictive way. Delete when something feels off that you just can put your finger on.  ChatGBT famously tried to break up a New York Times reporter and his wife and I’m sure its creators never intended that use case.

AI creates powerful self-learning systems that can learn about you. They will get better at manipulating you with exponential speed. Therefore, it’s best to jump ship quickly at the first signs of trouble. Life is short and precious and the far less powerful algorithms of the last decade have already contributed to untold misery. The next decade of AI will be exponentially smarter but there is no evidence yet that they will be one iota wiser. 

And when you get to have your own personal AI, which you can already do today with a $20 monthly subscription, try to be wiser than Mark Zuckerberg about what you ask it to do. Consider your questions from all angles. Consider the good you want to do in the world and whether AI is the best tool for the job. Then if you must, spin the dice and let this new thinking creature create the wonders that you ask. It might not do what you think or in the way you think it will but I suppose that’s part of what makes this moment so much fun. Good Luck!

Thank you for your question and please keep the questions coming readers! You can send any issue you want advice on to askdeath@wecroak.com

Sincerely,

Death

P.S. I didn’t include saying you should vote for politicians who promise to wisely regulate AI – even though that may seem obvious. At this moment in time, powerful thinking machines are a fact but politicians getting their act together enough to regulate technology wisely sounds like implausible science fiction. But maybe you can ask your new AI how to make them do it?