The Future of Fly Tying using AI

[Further North]

https://spawnflyfish.com/blogs/spawn-blog/the-future-of-fly-tying-using-ai?fbclid=IwAR22_vnk7JgymvqzwPBZnIqsLILr-DmfHEV0ZTMmgVMb9n9SZCZ3i_9SksM

 

The Future of Fly Tying using AI

 

 

Fly tying is an art form that has been around for centuries. From the early days of using only feathers and fur to modern-day synthetics and cutting-edge materials, fly tying continues to evolve. However, one thing has remained constant, the human element of creativity and experience. We all have been drawn to this aspect of our sport and that will not go away.

We believe AI is here to stay and is going to impact every industry including ours. With the advent of artificial intelligence (AI), fly tying is set to enter a new era. AI technology has the potential to revolutionize fly tying by assisting fly tyers in creating new patterns, providing innovative material ideas, and incorporating unique design elements into their creations all with the click of a button.

 

 

One of the most interesting aspects of creativity AI is pattern generation. Using multiple AI tools we have extensive knowledge on pattern generation. Every image on this blog is created with AI and it is frankly terrifying. However, at this moment it is not as good as we are when it comes to creativity. Through machine learning, AI can analyze existing fly patterns through photo uploads, prior knowledge or descriptions and create new ones based on similar design elements. This means that AI can help fly tyers create new patterns as well as introduce new variations of classic patterns. Here is an example of a simple more classic style shrimp pattern I created in AI then asked it to provide variations.

 

 

 

 

In addition to pattern generation, creativity AI can also provide innovative material ideas. With access to a vast database, AI can suggest new materials to use in fly tying based on the intended target species, fishing conditions, and desired presentation. This means that fly tyers can experiment with new materials that they may not have considered before. All of this without having to read the past literature on these fisheries, blogs like this that we write, or even without watching the YouTube videos we all love. AI will do it for you.

 

 

Finally, AI can assist with incorporating unique design elements into fly patterns. By analyzing existing patterns as well as real insects identifying successful design elements, AI can suggest new design variations that may be more effective at attracting fish. This means that fly tyers can create patterns that are not only aesthetically pleasing but also more likely to result in a successful catch. We are a bit skeptical on this but time will tell!

 

However, despite the potential of creativity AI in fly tying, there is no replacement for the human element of experience. The joy of catching a fish on a pattern that you created, tied, and cast is a feeling that cannot be replicated by AI. The passion and creativity that fly tyers bring to their craft cannot be replaced by technology. At Spawn we believe AI will only bolster what we do, it will draw people to more real experiences like tying a fly and catching a fish without the assistance of AI. We know our hands on experience creates not only a product we love but a truly human relationship with what we created.

 

AI isn't going away but either is the creative fly tier.

 

 

Apr 06, 2023

[TnRiver46]

Well, I see this as both positive and negative. Hand tied will always be a better bug, but hopefully this eliminates some of the sweat shop style production 

[Further North]

How he did it:

 

 

[txchaser]

brilliant!

[bulldog1935]

I'll give you a like for glossy effort (great photos) and enthusiasm, but also want to remove it for missing the boat.  

Still the best 1987 photo I have, partly because our will-be guide buddy behind the camera will become Phil Shook's inshore guide, and get write ups with action photos in Parks & Wildlife, and FR&R.  Also, my buddy Stevo (in the middle) and I spent a few socked-in lubricated weekends working out the perfect shrimp and crab imitations and picking up on hill country flies from Billy Trimble to David Train.  

 

The Austin Angler kept a wash tub of iced free beer. Joe Robinson was always perched to involve an unsuspecting visitor for a couple of hours at the vise.  My very first vise time was with Billy Trimble, and a decade later, was able to join him and a few other trout wizards on week-long runs along the CO/NM border - they tied tiny midges by Coleman lantern light - I just held out my midge box and poured the Scotch.

The clerk at The Angler was Alvin Deaudeaux, who may be the best guide in the country today - he certainly sets the best lunch.
The Angler was on Congress Avenue, a street-level window box entry, and a stairway window box climb to the 2nd floor.
Downstairs was Alfa of Austin. In a way, they hawked each others' wares, since a trip to the Alfa dealer might involve killing a couple of hours, and when you took a rod to the alley to test-cast, both sides of the alley were lined with classic coachwork Alfas.

 

Besides its blazing inaugural run from visiting relatives in TN, down Naches Trace in a blur, and on to the '84 NOLA World's Fair, Stevo and I made a run from camping in Palo Duro Canyon to fishing Cheeseman Canyon in 4 hrs 10 min.  

 

But the point, what good is fly tying without a washtub of iced beer, good friends, and the memory after they each become "world famous in TX hill country" -

- no AI can do that.  

[Further North]

3 hours ago, bulldog1935 said:

I'll give you a like for glossy effort (great photos) and enthusiasm, but also want to remove it for missing the boat.  

...no boats missed.

I post things I think people will find interesting, and to try to get a conversation going not necessarily because I think it's the path to follow.

Your post is a great example of a fun conversation!

3 hours ago, bulldog1935 said:

But the point, what good is fly tying without a washtub of iced beer, good friends, and the memory after they each become "world famous in TX hill country" -

- no AI can do that.  

I'll still do my own "creative design" outside of computers, there's too much that hasn't been programmed in yet, and I'm not sure it ever will be. 

All of what you mention, plus watching and evaluating how a fly actually performs in the real world  - how does it cast, how does it work in the water, does it catch fish? - are things that can't presently be reduced to code and fed to machine.

[flyfisher]

I see this as a parallel to baseball and the strike zone.  Many are against the computer controlled strike zone which will make every strike or ball the same and consistent across all games and levels.  Others, argue that there is a human element to the game that should not be removed.  

I think the AI fly tying is cool as **** and I would be all for it.  Will a hand made fly be better quality?  I have been on a lot of fly tying pages and I can tell you just because it is human doesn't make it better, myself included.  I also know that a fly similar to that first one would cost a fortune if hand tied but I will say that if you could get me an AI in a reasonable price I'd buy one and fish it to see if it is any better.

 

I am all for innovation and remember working in a fly shop when the gummy minnow came out and all the purists were up in arms about it not being a fly etc...I used the material and caught some good fish and tunes changed.  I can't imagine what those people think about articulated flies that are so prevalent.

 

Guess I am not a purist but I know if I have a choice I am throwing a fly over a "lure" and sometimes even a lure that I use as a fly.....

[Further North]

I don't know if AI will design better flies or not...that's for the future at this point.  As I mentioned above, I feel like there's too big a gap between design and evaluation yet.

In the video it's obvious that the choices in design are still driven by human interaction and that the AI is really just rendering, not "creating".

Automation of actual tying would be another step (some of which I think is already happening) for consistency, but natural materials add variation no matter what (almost 40 years making corrugated teaches you that).  As we're prone to say in our fly tying club, "Nature doesn't make any perfect bugs/fish/critters."

 

15 minutes ago, flyfisher said:

Guess I am not a purist but I know if I have a choice I am throwing a fly over a "lure" and sometimes even a lure that I use as a fly.....

I'm not any kind of purist...I think as a fly angler focused on pike, musky and bass, that's impossible...and I always enjoy the discussions about whether a this fly or that is "real fly fishing". 

I do often question why some anglers try so hard to duplicate gear lures and baits though...when I want to use those...I just fish gear.

[bulldog1935]

When 3D printing arrived, everyone was ga-ga over e-mailing each other new fly reels.  

The cost of the powder-metal printer may not be what you're thinking.  

The one I saw in a gas turbine shop filled a room - the only worthwhile parts it could make were a handful of small burner parts, and it cost $18M (useful for a regional GT support shop).  

 

Most fly tiers make the mistake of duplicating a mirror static fly.  

This is a shrimp in the bay as a fish sees it evading, also perfectly imitated by stripping line.  

(insects on the water surface tension look very different to fish than we see in the air)

[flyfisher]

31 minutes ago, bulldog1935 said:

When 3D printing arrived, everyone was ga-ga over e-mailing each other new fly reels.  

The cost of the powder-metal printer may not be what you're thinking.  

The one I saw in a gas turbine shop filled a room - the only worthwhile parts it could make were a handful of small burner parts, and it cost $18M (useful for a regional GT support shop).  

 

Most fly tiers make the mistake of duplicating a mirror static fly.  

This is a shrimp in the bay as a fish sees it evading, also perfectly imitated by stripping line.  

(insects on the water surface tension look very different to fish than we see in the air)

No fly is a perfect imitation and this is no different.  It may imitate it enough to fool something with a brain the sized of a peanut though.  I do get your point though as realism doesn't matter except when it does....

[Further North]

3 hours ago, flyfisher said:

I do get your point though as realism doesn't matter except when it does....

I think of this when I see perfectly rendered crayfish flies...the craftsmanship and attention to detail is simply amazing...even though bass have a marked preference for crayfish without claws...