On the way home, Brenden and I talked a little bit about machine learning or artificial intelligence. I believe the subject came up after Brenden spoke of the impossibility of answering certain questions. I spoke of the probability of certain answers. For example, the last hour of fishing this morning was slow for me. I had lost what I believed was my last bright purple Yum Dinger and had put on a white and green one. I did catch two bass on it, but we'll never know for certain if the slow-down was due to the long shoreline we fished being inherently less productive today, or whether it was that Yum Dinger. I know it's thinkable that the purple color was fishing better, because my friend Fred Matero and I have experimented with the colors of Senkos, an exercise which seemed to show us results. I also know that at least in the past, that last shoreline of the day fished well. It makes you wonder about the color. It's at least a little probable that it made a negative difference.
Machine learning itself leaves so much to imagination, because it (artificial intelligence) is off on its own, and no one knows exactly how it learns. Nor does anyone know exactly what conclusions it will draw. It's thought that it can do enormous good in fields like medicine, but we don't know what it will think of the relationship between it and us. And...what it will do.
Fishing is Like Artificial Intelligence
A fishing trip is like that. It's something of itself. We might like to think we master the situation, which very much was the attitude during the 1970's when I began fishing seriously. The whole idea that "Knowledge leads to fishing success," to quote the master himself, Buck Perry. A man who had taught high school physics. The 1970's were a simpler world in which knowledge had an absolute quality. (I was fixated on the fact that Einstein's relativity hadn't made it into the high school physics program at Lawrence where I grew up.)
Today, anglers have much greater access to knowledge and information. But "the more I know, the more I realize what I don't know," as Albert Einstein put it decades before the '70's. A man ahead of his time. It's not that a fishing trip masters us, but an outing is an interaction, which our knowledge participates in more than it determines results. We don't exactly know how we do it--just as artificial intelligence is less determined by us as the creator of it, than being a field of interactions in the sense of AI being a self-initiated development of understanding. Just the same, when we fish a particular shoreline, let's say, edged by weedbeds and dropping off into eight or nine feet of water, we don't know how much that environment involves itself in meeting us. We take it for granted that we're after the bass. We never stop and think--as maybe we begin to now--that the bass as representatives of an organic environment are directed towards getting caught. But that they do get caught at all implies that they're at least receptive to that. I don't mean at all that they are available in any way that would be obvious and easy, but it's not a master and slave, or predator and victim, relationship. Even the rabbit grabbed by the coyote fulfills its possibility as nourishment for the advantaged. The "advantaged" is that only in light of the rabbit. And I mean "light" in the sense of making something possible, of making something visible as it were. Just as the rabbit is a positive agent in that respect, so is the bass to us.
Now artificial intelligence has implications for the ecological scene. When both we and AI are ready for the changes. Even as the two of us are separate. the two of us are striving to understand one another.
We used to think we have the upper hand on how an outing will work out, and then we discover we've run out of purple Dingers. Later on, as Brenden and I emptied the canoe, I discovered I had one more purple Dinger in a box I hadn't checked. I did have the knowledge of that Dinger. On some level, I knew it was in that box. But knowledge can only function within a system. If my organization of the items I presume to bring along and keep at my disposal isn't lucid, then doesn't a glitch exist in the system?
Or you could say, being denied that purple Dinger opened a whole other can of worms.
(It's made for an interesting article.)
Artificial intelligence might learn from a similar glitch in its system. In any case, why is it that of a very successful day--I caught 15 largemouths in less than six hours, much of them spent fishing a Mepp's Giant Killer for muskies--what I might remember best is that white and green Dinger?
It poses a question.
Fishing is Messy
And Brenden wasn't altogether at the Zenith of his Form, either, which of course we would expect of ourselves during that Ideal Decade of the '70's. He told me yesterday he might bring one rod, so I emphasized that I was bringing two, sensing that he was making a mistake if he did that. He did bring two, but they were both casting rods intended for musky. He never fished a Senko or Yum Dinger. He ended up catching two or three largemouths on jerkbaits. I can't remember if it was two or three, but I think it was two, so all the more power to the uncertainty and messiness of what we do. And think.
He regretted not having brought a third rod, although he did lure three muskies into following his spinner and jerkbait to the canoe. Water temperature ranged from 79 to 83. He told me it was in the mid-70's late in May when he caught three muskies on one outing, the largest at 42 inches, and a single 47-incher another time.. I think with 83-degree water, even though that's at the surface, as Brenden pointed out, the likelihood of getting one to hit diminishes. There are no muskies in Florida but plenty of largemouths.
No comments:
Post a Comment
Comments Encouraged and Answered