This thread has become very interesting. It seems there are two main reasons why some divers here do not like AI. In my view, some of these reasons are tangible like the reliability of the gear, but most are philosophical.
1. There are really only two reliability concerns: the signal can be dropped between transmitter and receiver, or battery failure of the transmitter (actual physical transmitter failures seem to be extremely low, almost on a par with SPG failures) Signal drops are, almost universally, very temporary with the computer picking up signal again in just a few seconds. Also, the anecdotal evidence seems to be that transmitters from the last 5 years or so really do not suffer much signal drop if they are mounted properly on the first stage. These have been addressed dozens of times and the loss of signal condern is best refuted by the hundreds of thousands (ormore) AI computers in use, and the lack of distinctt or reliable evidence they cause people to go OOA.
I If the transmitter is not mounted properly, or the battery dies, that is not an equipment problem, it is a diver problem.
2. Which brings me to the philosophical issues. First, is that it makes a diver "lazy" and so that diver won't keep proper track of their gas. You are saying that having your gas on your wrist at a glance makes you less aware? Or that you are more aware of gas when you do the grab-and-check of an spg? Sorry, this one does not compute. As a former old school and now AI diver, I can tell you that I am more aware of my gas status now. Especially when carrying a big camera or using a dpv, when the "grab and check" process is really, really inconvenient.
The second philosophical concern is a variety of things all lumped into the general idea that "divers don't know how to properly manage gas." What this has to do with a computer is beyond me. I presume that all divers know how to use their gear and understand what it is telling them. If they do not, that is not a "gear problem." In fact, if the outcry over lack of knowledge is true, then a computer's ATR function should actually increase, not decrease, diver safety.
So, some say "don't trust ATR". Well, if you don't, then just use the computer as a pressure gauge readout. That is its basic function after all. How you manage your gas is entirely up to you.
Some say "ATR doesn't account for two divers having to surface." And an SPG does? In fact (at least on my computer), turn pressure alerts can be manually set for whatever pressure you want based on you own calculations if you are a "rock bottom" diver (which I am not).
For NDL diving (not deco diving) I find that the ATR function would suffice for a quick ascent sharing air. A practical example, on my last dive, the ATR (for a surface reserve of 500) sent me up with 750. Means it calculated I needed 250 for a standard ascent and 3 minute safety stop. That also means I have 500 extra to share with a buddy if needed--which, if he needs 250, still leaves a 250 cushion. Even if we would be breathing more on a sharing ascent due to stress, under those circumstances we are doing a faster ascent and skipping the stop. So, I am satisfied that I can rely on the ATR to get us up, for no-deco diving.
One poster said AI leads divers to "push the limits." That is not an AI problem, it is a diver problem. See my example above, AI can be made as conservative as you want it to be (ie I could set my surface reserve for 750).
Again, it's knowing the gear and how to use it. For everyone who wants to do it "old school" that is great, but it is NOT an objection to AI technology.
I LIKE my AI computer (and still carry an spg)
I LIKE my back-up camera (and still look).