The first time I came across the use of artificial intelligence in the fly fishing  world, it benefitted me. At the very least, it had the potential to benefit me. A spike in website traffic to Casting Across was the result of a link to an article I had written. The link was on another fly fishing website. After a few minutes of reading, something felt off. The writing was not good. Now, people can be poor writers. But this wasn’t just C- writing. It was soulless. It was undoubtedly AI.

I’m hardly an angling luddite. Casting Across is, after all, an online platform. I have hundreds of podcasts and dozens of YouTube videos taking up digital space on servers somewhere. Technology is not a problem. AI isn’t even necessarily a problem.

Recently, I received a marketing email from a company trying to sell me on how AI could help me connect with my audience. Even though I was confident that the answer would be a strong no I read their pitch. Basically, they used natural language processing to comb through the available data to create a list of topics you’re interested in.

What did I do with this data?

I recorded a podcast episode. And, according to the feedback solicited from you, the robots didn’t do half bad in pointing out the kinds of fly fishing information you were looking for.

There are uses for AI in fly fishing. The above seems like one of them. It is pretty benign. It takes what might require an hour of time and compresses it into a few minutes. I’ve toyed around with some of the embedded AI features from Google and Microsoft. They give reasonably accurate gear recommendations and local fishing opportunities. Similarly, automatically generated online fishing reports are just fine. The kind of general information, big-picture variables that a lot of online fishing reports consist of are usually a copy/paste job with a few pertinent and timely details added in. Also, it was fun to watch AI use my prompts to generate the image at the head of this post.

Once more: not necessarily a problem. As is the case with most things, I suppose moderation and discernment are key. Tools that make aspects of life easier are fine. Tools that fundamentally change life ought to be treated with caution.  Difficulty and challenge aren’t the enemies of life or of fly fishing. Within fly fishing, what AI tools might encroach on could be a bit detrimental. Online fishing data and equipment recommendations should only augment what real people provide. It shouldn’t circumvent human interaction. It is a problem if it makes a stop into a fly shop unnecessary. It is a problem if it takes a brief streamside chat and replaces it with a tap on an app. The internet is already an on-ramp to this issue. AI widens the lanes and increases the speed limit.

At the heart of the matter is the same aspect of soullessness that you see in AI generated content. Fly fishing is poetic and introspective. Fly fishing, even when pursued for a sense of solitude, has legitimate communal elements. A robot can’t feel. A robot can’t relate. And I am 100% confident that even if a robot one day catches a trout using a five weight and an Adams, it won’t get it.