Artificial intelligence has been with us to a certain degree for a few decades now. It’s been a practical tool for large corporations for 10 – 15 years and it really entered the public consciousness in late 2022 with the opening to the public of ChatGPT (for the written word) and Dall-E (for image generation). Like most new things, two camps quickly developed: a) “It’s too often wrong to be considered a serious tool”, and b) “it’s not quite where it needs to be, but it is the future.” Of course, there’s also the third camp which is: c) “I will use this to deceive people to my own gain.” Myself, I fall solidly in both a and b camps.
As a journalist and consultant, I fear it a bit because it is taking work from humans and, for now, is contributing to the dumbing down of our collective information base. Given that most of the publicly available AI tools use the Internet as their primary language model to learn from and given that so much of what is on the Internet is intentionally or unintentionally wrong, how could we expect it to be any different?
On the other hand, I also see it as the future. There are applications that will certainly benefit from it and someday, I really hope that it will be able to do a better job of differentiating true from false. On the third hand, though, humans really struggle with true vs. false and right vs. wrong, so, again, how can we expect the machine to be any better. This is nothing new. It’s just an amplified and much faster echo chamber.
The featured image here is a Dall-E creation of Albert Einstein, Richard Feynman and Tom Lehrer singing together. I imagine that Dall-E arranged them in order, left to right as I put in my prompt, but it really looks more like each face in the image is a mash-up of all three.
The journalist in me holds onto the hope that the dumbing down of so much content will actually create more opportunity for quality human writing. However, only time will tell. At some point, the job of writer may become that of an AI driver. When that day comes, I may or may not embrace it and I may or may not have the skill set. In the meantime, I write and I think, and I do my best to produce better material than “the machine.”
Like every other new technology, the hype initially far exceeds the reality. And, like most successful new technologies, as it gets better, the hype will die down and the tools will become genuinely useful. As the tools get better, we will think less and less about it, and it will just become a part of everything around us. The key to successfully navigating a future where AI is ubiquitous is in understanding where AI makes sense and where human intelligence still fits. Again, that is no different than with any tool we have ever invented.