AI is a great asset for learning when paired with a more authoritative source. I use it to learn coding and math and for basic/foundational knowledge, its great at explaining things in ways that are more understandable than alot of teachers and professors I've had in the past. That's usually because foundational knowledge usually has a pretty clear and well established ground truth to it.In order to decide what you can use for AI, just ask yourself "how obscure is this topic and how much contention is there regarding its contents?" If you want to know how basic radar might work, sure use chat gpt. Its even better if you read an online book and have chatgpt explain things you dont understand. If you want to ask chatgpt about how sensitive is a rafale's radar, good luck - you'd be an idiot to believe anything it says. Sometimes even "sparring" with chatgpt will cause it to deliberately begin to agree with you.
I'd like to add to the list of sites to avoid the South China Morning Post. It is notoriously bad just for how much it distorts PLA news, which itself is already opaque enough. Its even worse than the national interest. Almost every single time I've forced myself to read some garbage by them, its some science paper cited as gospel and then imaginated into some super weapon by the author, who by the end of the article most certainly needs a new change of undergarments.
On the subject of youtube channels, I find a number of ex-mil pilots to be terribly unreliable and annoyingly so. The ones im referring to might be great sources for cold war and GWOT era stuff, but dont expect them to be good sources on current topics. I won't name names (dont think I need to cuz im sure everyone who likes military stuff and watches youtube knows who im talking about), but a number of them are biased against stealth (namely F-35s), which to me is batshit loony town considering the current operational challenges in the aerial domain. Most of their commentary on modern stuff is also boilerplate and uninformative at best and at worst, overtly reductionist and lacking nuance. They rely too much on their past experiences some of which, let's be honest here, is approaching obsolesence when considering recent developments and the current peer threat.
