My Unprompted Thoughts on AI

This should start with a disclaimer. I am not an AI researcher. I am not a security researcher. I am not a computer scientist. I am a nerd with a website. If you want an actual evaluation of the state of AI, please find someone with more credentials.

Remember when NFTs were going to revolutionize digital content?
Remember when blockchain was going to revolutionize financial transactions?
Remember when VR was going to revolutionize the way we work and live?
Remember when Self-Checkout was going to revolutionize the way we shop?
Remember when e-books were going to revolutionize literature?
Remember when <technology> was going to revolutionize <subject field>?

Technological development will always march on. That is a good thing. But true technological revolutions are few and far between. Most importantly for my point, the public is slow to adopt technology. Technology starts as a fad, and gets tons of press, and looks like it is going to change everything overnight. But it is always awful under the hood. And then people move on with their lives. And then, as we are not paying attention, it slowly gets a little bit better, and we start seeing it more and more.

And now you’re thinking, “This is the point where she says, ‘this technology is eventually great and everyone uses it!’”.

But no, that is not what I am going to say.

e-books are locked behind expensive, pay-to-read publishers. The advantage that they offer is usually outweighed by the simplicity of the experience for those people who really like reading books. Self-Checkout has proliferated, but companies still have all their registers open on busy days. Routinely, one of the systems will be down, or need intervention. VR is still a glitchy mess, accessible only to people who want to pay for a novelty that quickly wears off. Blockchain is back to the obscurity it came from, without the anonymity and technical weirdness that made it useful. NFTs are a laughingstock.

AI is… Well, it is certainly here. It is a neat toy, and far more accessible than most of the technologies that came before. But that is entirely due to mass subsidization from the companies running these AI platforms. Sure, it will let way more people access art, and literature, and other creative tools. But it is not changing the core behavior of humans. People have always cheated on tests. We used to just pay strangers on the internet to do it. Now we pay a machine to do a worse job. People have always stolen art. Now we just pay a machine to make a worse version that needs correction every time. People have always written bad documents. Now there is just no proofreader before a lawyer gets himself laughed off the international stage.

AI will do for SOME fields, what 3D printing did for manufacturing. It will let you get an outline, a rough draft, a prototype quickly. But that does not mean it’s replacing human creativity. It means all those writers will be editors and proofreaders. It means all those artists will be touching up work they would rather do themselves. It means all those high school students will get caught even faster when their essays don’t make sense.

AI is a tool. But it is not truth. It has no concept of truth. To call its mistakes “hallucinations” grossly misunderstands how it works in the first place. There will never be a replacement for human creativity. I am not saying that from a moral perspective, but a technical one. There will never be a replacement for human attention to detail. There will never be a replacement for people who want to pay their favorite artist, not because of the art but because of the person who made it.

AI is not going to change the world, because fundamentally, the time of technical advancement in communication is effectively over. Sure, we can find technology that makes us telepathic, but I can already reach everyone I want to immediately. I can already find any image that I want, or at least close enough that it is better than AI. AI is not going to change the world, because that will require a change to humanity. We like our books, and how they smell. We like how cashiers bag our groceries for us. We like how cheap and accessible normal computers are. We like how our financial transactions already work and let us hold people accountable. We like how images and content online are free and easy to share. We like our favorite authors, our favorite screenwriters, our favorite artists, and photographers, and every other creative person out there. Because what we like is not necessarily what they are making, but because they made it. Because we know what to expect from them. Because they are familiar. And AI will never be familiar. Not like real humans are. AI is fighting to be good at the wrong things. It’s climbing the mountain of technical challenges, when what it needs to succeed is to climb emotional and human challenges. And once it does climb those mountains, we will be uncomfortable, and fight it, and it will have to fight hard to find a place in our everyday lives.

One last note; the reason AI will not succeed is the same reason that remote work is failing. Institutions like having someone to blame. Someone they can walk up to and hold accountable. Someone they can “motivate” by making an individual reliant on them. Ai is not reliant on institutions. It cannot be brought into a manager’s office and told it needs to do better or it will be on the streets.

So, to all the creatives; All the writers, and artists, and screenwriters, and designers, and programmers, and everyone else: AI is not coming for your jobs. It may change those jobs, and make them more tedious, but they will still be there. And AI is not coming for your jobs because while capitalism talks a big game about replacing human labor, it never has so far. And it likes way more to have humans to praise and crush than machines.