2025-05-08
A few thoughts on AI.
The above image was generated by ChatGPT's new image generator. You can kind of tell, there's something not quite right about it. But it is clearly a lot better than it was. This version understood the nature of 35mm film, what a 50mm lens looked like, and what its depth of field might roughly be.
I haven't played with AI image generation for a long time, for all the reasons so many of us have been avoiding it; I'm not a fan of the tech bros, I don't like their whevsy view of copyright etc etc.
At the same time, though, I am interested in new technology and what it can do. And AI is unarguably here to stay and it's going to have a huge impact on all of our lives. So I don't think that pretending it doesn't exist is a reasonable stance. I believe there are fights to be had on the topic, steadfast positions to be taken on the rights of creators to own their own work and for that work to not be exploited or scraped without permission etc. But I don't think we can win those fights if we're not fully aware of where the technology is at any given point and its direction of travel.
There are people who will object to me even using this image in the header. I am drawing attention to the tech, I am destroying the livelihood of real photographers (I was never going to employ a photographer to take a picture for this morning's post). Just playing with the technology makes you the bad guy in some people's eyes.
So why, this morning, did I choose to open ChatGPT? Well, it was because of this piece by Tobias Van Schneider, creator of the excellent MyMind app:
Tobias van Schneider
The arguments made in that piece are pretty cogent, and align with a lot of my thinking; AI raises the game of mediocre operators. It does not replace true creativity. Used properly, AI can raise the game of truly original people. There is a cost to it, just as there were costs to the industrial revolution. Serious costs. But we don't avoid them by ignoring them, and protesting AI generally seems pretty futile - I have had medical scans that were enhanced by AI, to the benefit of my health. I'm pretty much in favour of that.
I think, as a general rule, I'm OK with creative people using AI as a tool; an AI image mask might save an hour's work in Photoshop, a well-formed Perplexity search might throw up interesting pointers to new background research. I'm less OK with the image above; a whole-cloth generated image that I had very little to do with. But if the tech exists, someone is going to use it. We can't put the genie back into the bottle, so we have to figure out how to move forward with it.
There might be a path through this that protects livelihoods, but it's going to be complicated. The technology is not going away, it's getting better and better (and it would need to, because a lot of it is really crappy at the moment).
But if there is a path to be found that minimises job losses and doesn't decimate human industries, it's going to be found by people who understand the terrain, not people who refuse to acknowledge it. Before your job is taken by AI, it will be taken by someone who can use AI; just as people who could use word-processors supplanted those using traditional typewriters. (My grandfather worked his whole life repairing manual typewriters. For him, electric typewriters were the beginning of the end.)
Personally, I like working with creative people who are good at what they do and I have no interest in replacing them with technology. But if those people use this tech to make their great work better or easier, is that wrong?