People haven't stopped arguing about AI for years. To be fair, it's a broad topic. It encompasses labour concerns, the power of corporations, intellectual property rights. The frustratingly brilliant marketing choice to invoke the sci-fi concept of artificial intelligence (to describe a machine capable of tricking human brains into recognising a fellow sapient entity whilst actually having the same level of interiority as the racks of vacuum tubes and magnetic tape which constitute its earliest ancestors) has even brought all the ethical concerns and apocalyptic fears associated with that label into the collective discourse.
I have my opinions about AI, even a few positive ones which I'll get to, but what I personally find most compelling about these debates is that they act as a proxy for discussions of far more interesting and fundamental questions: What is it that we value about art? Is ownership a meaningful concept or just a legal abstraction? What if computer but too much? (the perennial classic) It's cool to see the general public wrestle with these ideas, previously limited to radicals and philosophers. What I want to discuss here is maybe not as broad or consequential as these, but it is something I think about regularly during my life and work outdoors.
A few years ago I finally started to make a conscious effort to learn the names of the wildlife around me, starting with trees. I already knew a little more than the average person, largely thanks to the efforts of my mother who would certainly be the first person to spring to mind if anyone asked me where my love of nature had come from. But I wanted to know everything better, to fill in the sizable gaps in my knowledge, and so on my walks I would bring a guidebook and a phone, and either look things up as I went, or take pictures of distinguishing features to check later.
For a long time I've considered this to be a firm second best for ways to learn species identification. It's a much more practical alternative on most days than the absolute best which is to find someone who already knows everything and just follow them around (which is how I solidified a lot of my butterfly knowledge). However, recently I've become more and more aware of another contender, one which is hugely popular among professionals and the general public alike: AI identification apps.
I find myself turning to apps like Merlin, Plantnet and Obsidentify fairly often now, when a plant I don't know catches my eye, or I hear a bird which is not one of the three or four I can confidently recognise by song. I find these apps to be excellent at compensating for gaps in my knowledge, and their level of accuracy is impressive. This is my one major positive feeling about AI; that it is excellent when used for very specific purposes, like recognising small differences in the leaves of plants.
On the other hand though, I've often found myself scanning the same plant twice on different occasions. Once the computer has analysed whatever photo or audio recording I feed it, it spits out the answer and my curiosity is satisfied. But the knowledge simply does not stick in my brain in the way it does when I look it up in a book or even in an app like Shroomify, an app for mushroom identification which is something like an electronic version of a classic field guide. I'm not a psychologist but I can think of a lot of things that may contribute to the slipperiness of the information I get from AI powered apps. Perhaps it's the hyperstreamlined process of identification: just press one button to feed an image into the algorithmic black box, then a few moments later you have your answer. No need to look through a whole book and compare against various reference images and check all the details and distinguishing features, increasing the chances of something sticking in your brain. Perhaps it's the format. When you scan a plant with Plantnet you are given a name and a few photos. For more information you need to click through to another page which then provides links to Wikipedia and other databases, whereas with a good guidebook you are presented with quite a lot of information as soon as you find what you're looking for. Or perhaps it's just the knowledge that you always have your phone which means that your brain doesn't bother wasting energy on storing information you could easily retrieve later.
From a purely practical point of view, maybe this isn't a big concern. How often, even as a conservationist or researcher, are you likely to come across a plant you want to identify without having your phone handy? Even if you don't have it, does it matter that much? Probably not. You could imagine a scenario where it would be vital to have this knowledge, but the overlap in the Venn diagram of people who own smartphones and people who need to weigh up the risks of starving to death versus eating some unknown berry is probably pretty miniscule. I won't dispute that memorising this kind of information isn't practically useful for most purposes, but I don't believe that utility should be the only thing that makes something worthwhile.
Writing that, though, puts me in a bind. It's impossible to try and sell a reader on the inherent value of getting to know their local wildlife without pointing to some concrete benefit that it gives you. I could talk about the mental health benefits of feeling more awareness of the world and its seasons, or the way that recognising different trees allows you to mentally sort through what would otherwise be a mass of green and brown and grey, and find small species or details which you might otherwise have missed, and the joy that can bring (as well as the advantages to, say, photographers). If I were more spiritually inclined, I could talk about the value of connection to nature. But, while true, all of these feel inadequate. I may in fact be bumping up against one of the deep questions I mentioned earlier: What is the value of art? Because like with most emotional experiences, any words I can write inherently fail to capture the truth of the feeling knowing these things gives you. Imagine trying to sell someone on the utility of romance, what would you even say? It makes you feel good? Studies suggest it improves your metabolism? You have an excuse to go out to dinner more often? It feels ridiculous. And so poets bend and twist language in ways that bypass its literal meaning and access something less conscious, and painters manipulate textures and forms and symbols to evoke emotions which are inexpressible in any other way. Art is communication between feeling entities and that is why AI art often fails, the message is being received but there is no sender.