In a previous most I showed how I used AI to provide me with the top fishing spots on a specific lake based upon seasonality, weather, time of day, etc. I didn't get to test those results, but this past weekend I did on another lake. My premise is that if AI is accurate, along with map study it can help me break down new lakes quicker. This is a big help since I fish out of a kayak and often fish tournaments on lakes that are new to me.

 

So this weekend I headed to Lake Keowee. I actually fished this once or twice before as a back seater in and MLF tournament, but I have no idea where we fished. Keowee has been taken over by spotted bass, so I aske AI to target that species for Sunday based upon seasonality, weather, etc. and provide me the top 5 locations on the lake. AI also provided me with a file of the GPS coordinates that I uploaded to my Garmin.

 

I was supposed to fish Saturday, Sunday, and Monday but that didn't work out. Saturday I had to replace a bearing on my trailer and I chose to spend time with my son instead of fish on Monday.

 

Anyway, AI provided me with the top 5 spots and of those I chose the Cane Creek area since it was close to my girlfriend's lake house. I got a late start and didn't get on the water until 7 am, but my first spot was close the ramp. I immediately saw what looked to be a few bass cruising on my livescope. I took a few casts with a frog with no takers. Then I saw a few spotted bas swim by in the crystal clear Keowee water, tossed my frog in their direction, and they ignored it.

 

I switched over to a crankbait and immediately caught two. Nothing big - just a 12 inch and a 15 inch. I pushed further into that creek channel and caught nothing, then caught two more on the way back out. 

 

The next spot was further up Cane Creek, but along the way I got bored and fished a few spots and caught a small pumpkinseed on a crankbait of all things. It was only 8:30 am when I reached the next spot. According to AI it was a spot I should hit in the afternoon, and maybe it was correct as I saw nothing in this deeper water on livescope, not even bait. Still, I tooled around and fished the area but didn't get a bite.

 

The next stop on my AI list was further up to the inflow of Cane Creek, but at that point the jet ski's and pontoon boats were starting to hit the water so I started back to the ramp. I fished a few spots of my own choosing along the way and caught a few on a ribbon tail. Then I hit that AI spot missed a few bites, caught one more, then packed it in before the lake started looking like an amphibious freeway with the holiday boat traffic.

 

That's too bad, because one afternoon spot was deeper water, 40-50 feet deep. I almost never fish that deep and I wanted to try it out and get some practice.

 

Conclusions? One test and one outing isn't a definitive answer. I wish I had the time to try the suggested spots on the other parts of the lake, especially at the times of day it suggested. But it did lead me to some fish.

 

I may try this experiment on another lake in my area next weekend.

 

That’s real neat, I tried something similar a couple of months ago. Having AI design a plan on where to paddle to on the kayak and what to throw at what time to see how effective it was. Problem was I did it in a spot I was familiar with, so after getting some success, I deviated away from what the AI was telling me to do. I did have a very successful day using it.

 

Im curious to hear how it goes if you try again next weekend. Surely with all of the info out there that AI can use to compute a strategy on a given lake on a given day, it can be a really useful tool

2 hours ago, Kayak Koz said:

 

 

I may try this experiment on another lake in my area next weekend.

 

would love to see your prompt, and choice of models...

8 hours ago, txchaser said:

would love to see your prompt, and choice of models...

My prompt was along the line of: What are the top 5 spotted bass fishing locations on Lake Keowee for Sunday, August 31st 2025. Take into account the time of year, water clarity, wind speed, wind direction, and water temperature. Provide GPS positioning for each location.

 

What it did was not only provide me the locations, but also described what cover and structure was in the area, the water depths, preferred fishing methods, bait choice, bait colors, and the time of day to fish each area.

 

What I also plan to do is test different prompts for the same day and lake. For example, will I get different results if I as it to take into consideration the weather for the previous 3 days? What if I factor in rising or falling barometric pressure?

This is what I got from my Microsoft Co-Pilot this Morning for fishing tomorrow bank fishing:

 

"If you want, I can also give you a precise hour-by-hour “best bite” forecast for largemouth bass at Flint Creek tomorrow so you can time your moves perfectly. That way, you’ll know exactly when to leave the dam and hit those other spots.

Do you want me to pull that bite window forecast for you"?

 

Good Fishing

 

1 hour ago, GreenTrout said:

This is what I got from my Microsoft Co-Pilot this Morning for fishing tomorrow bank fishing:

 

"If you want, I can also give you a precise hour-by-hour “best bite” forecast for largemouth bass at Flint Creek tomorrow so you can time your moves perfectly. That way, you’ll know exactly when to leave the dam and hit those other spots.

Do you want me to pull that bite window forecast for you"?

 

Good Fishing

 

That's very interesting.

 

I think I will try multiple AI services and see how their information varies.

Inputting on Microsoft Copilot using your prompts, changing dates and location of course, I’ve come up with very similar recommendations as my Bassforecast app. Although the AI gives more in depth presentation methods in stead of just recommended lures.

I like it! I think this can be an aid in breaking down and fishing new lakes. Well done. 
Thanks for teaching an old dog new tricks!

Using AI and GPS and Livescope ain't enough. You gotta go all the way, man, and figure out how to incorporate a VR headset too. Use it in conjunction with Livescope and have AI transfer the 2D scope image to 3D models of your lure and any fish in sight (plus mapping the surrounding structure) so you can see up close digitized models in 3D space of how the fish are reacting in real time to your presentation. Should probably deploy drones too, both airborne and underwater. 

10 minutes ago, Motoboss said:

II think this can be an aid in breaking down and fishing new lakes. 

 

That's what interests me. I often head to tournaments on lakes that I have never fished before. Even if I manage to get a few days to practice, it's hard to cover a lot of water on a kayak.

 

Other anglers roll in with teams of 5+ people and they split up and pre-fish. That's puts single anglers like me at a disadvantage.

Just now, SpellCaster said:

Using AI and GPS and Livescope ain't enough. You gotta go all the way, man, and figure out how to incorporate a VR headset too. Use it in conjunction with Livescope and have AI transfer the 2D scope image to 3D models of your lure and any fish in sight (plus mapping the surrounding structure) so you can see up close digitized models in 3D space of how the fish are reacting in real time to your presentation. Should probably deploy drones too, both airborne and underwater. 

 

Actually, I do know how to incorporate AR with Livescope and the Active Captain app. I thought about trying it, but I don't want to put on a full headset. There are AR sunglasses, but the resolution isn't great. Give it another 2 or 3 years...

 

And if you think the pros aren't experimenting with this stuff, you're crazy. These guys are fishing for big money and looking for every advantage.

3 minutes ago, Kayak Koz said:

And if you think the pros aren't experimenting with this stuff, you're crazy. These guys are fishing for big money and looking for every advantage.

 

For real. Chris Zaldain mentioned AR goggles in conjunction with Livescope a few years ago. Didn't say he was using it but knew someone who was. 

I ran this prompt for my local lake just to see what it said. It came back with the names of legitimate locations that I have fished. The problem is that 3 of the 5 coordinates were not even in the water...not even close. One coordinate was in a different state. Be sure to check your coordinates before you get on the water.

50 minutes ago, Fishlegs said:

I ran this prompt for my local lake just to see what it said. It came back with the names of legitimate locations that I have fished. The problem is that 3 of the 5 coordinates were not even in the water...not even close. One coordinate was in a different state. Be sure to check your coordinates before you get on the water.

A different state? That's funny!

 

I had ChatGPT export a file for the coordinates, then transferred them to a micro SD car and uploaded them to my Garmin unit.

AI Searches available data fish reports including fishing sites like BR that a lot of us have contributed to freely then relies to the topic.

Thus technology is in it’s early stages and will improve building on it’s own data.

Tom

AI seems pretty effective at summarizing and even synthesizing data.  When it creates its own, I wonder who will conduct the peer review?

16 hours ago, WRB-2.0 said:

AI Searches available data fish reports including fishing sites like BR that a lot of us have contributed to freely then relies to the topic.

Thus technology is in it’s early stages and will improve building on it’s own data.

Tom

My experience shows it does that and more depending upon which AI you utilize. AI will tag where it derives specific information.

 

I'm doing some experimentation and analysis of different outputs. My latest comparison has yielded varying waypoints and varying strategies as I get AI to drill down based upon my additional prompts.

 

One of the keys in my analysis will be to see if AI changes output based upon different weather conditions within seasonality.

 

My next step is to create a GPX file and upload it to my Garmin, adding relative notes and separating the different AI waypoints with different symbols.

 

But it will take a lot of data and a lot of fishing to draw any conclusions.