In 2012-2015 ish, I was building FPV quads, planes for universities.

(I say quad because at the time we were staunch drones kill people….)

I really don’t think you need ai for any of this.

The stuff today is so easy plug and play. There’s so many groups you can connect with.

It’s also much more durable.

I don’t see a lot of discussion by the author about transmitter (TX) or receiver (RX) choices. Which are important.

Fatshark goggles are a great choice for most people.

If someone is having fun though, I don’t want to dissuade them. But focus on the fun. Practice flying. Practice flying line of sight. Then focus more on fpv.

But focusing on how the things work (and less worry about the batteries… quality charger, quality battery - high c rate for most small quads…)

Brushless motors help a lot for longevity.

All that said, in the hobby I’m back to los nitro planes, as that’s what I find most relaxing and enjoyable.

I’ll occasionally fly 200-450mm size quads, sometimes even fpv, but it’s just not as much fun to me.

In fact, I think the most fun fpv flying is with something more akin to a sky surfer, up a bit higher, longer flight time, etc.

But again— everyone’s fun is different - however I would still encourage not using ai, and trying to understand how different components are working together.

Truly the stuff today is so much easier , no bind plugs usually, no crystals, no lengthy manual pid tuning.

But I will stand by understanding the transmitter receiver tech, both for controlling and for video, and not so much focusing on what’s needed battery storage, but instead on how the power system as a whole is working, along with things like prop pitch, your motor type, etc will lead to a better grasp of what’s going on.

Simulator time as well, while good for some initial practice, tends to be very detached from The real world experience at the RC level.

I remember building my first quadcopter with kk5.5, it was a fun time indeed, and the technology has gone really far

So, today was my second day piloting drones ever. I am absolutely not sure what I would need any bit of AI for... it is quite enjoyable to zip around the trees here.

I did the same process as the "author", namely Joshua Bardwell's videos and tutorials Liftoff using a remote, and now I have a tinywhoop and analog goggles.

But it was mostly just normal practice and learning in the sim (I have maybe 25-35h of time in that) which has translated pretty well to the actual drone. I was able to fly around my yard pretty well. I don't think I'd have had as much luck if I'd not crashed many hundreds of times in a simulator.

For me, the amount of thought outsourced to LLMs has been a bit strange lately. Like, I just puzzled this stuff out and followed a couple of videos other folks had made on Youtube. It's just not that complex these days and there are plenty of (human intelligence) resources already.