Vellron2005, on 10 September 2019 - 11:34 PM, said:
To all the people reminding us of escort and academy A.I...
The point of the post is that PGI is supposed to be much more proficient at coding A.I now that MW5 is near release?
I always hope they take what they learned and apply it elsewhere.. Be it mech modeling skill, A.I coding, or all the ideas they adapted from feature suggestions to MW5..
The idea is that once MW5 is out, they can work on Clan invasion DLC and keep funding development by extending MWO shelflife, until they either make MW5 PVP Solaris DLC, or MWO2..
Bots would surely extend the shelf-life of MWO and could open up new avenues of making existing modes, like Solaris, viable, but also adding PVE content..
MWO need not be PVP exclusively..
That's assuming a lot, first that PGI has managed to make the AI in MW5 so good that it could plausibly mix in with humans in a pvp setting. I really doubt that, there isn't really such a thing as good AI in games yet when it comes to completely unscripted behavior. Most of the time the sequences in games that are perceived to have convincing AI are where the AI is carefully scripted to the environment, for example in Alien isolation etc.
I would be extremely surprised if MW5 had AI worth a damn, even by single player standards. Now I do hope I'm wrong here, but there are almost no games with convincing AI as I said, it is the biggest missing thing in the game industry IMO as everything else has reached a high level but AI is still crap. Now maybe we'll see neural networks becoming commonplace in consumer products in a couple years and maybe then they will take over in games too and we might have a shift towards actually convincing AI. I hope so.
But let's assume MW5 AI is good enough to fill out teams against human players, and let's also assume the transfer and implementation of MW5 AI into MWO is trivial and cheap enough, and that MWO has a long enough predicted lifespan for PGI to bother doing it.
Let's assume all that for the sake of argument and the question still remains when would you actually pull that AI into matches? At what point of waiting time is it actually preferable to add a bot over waiting for another player to fill the match? 5 minutes? 10 minutes? Matches are still firing way faster than that so I'm not sure how there is even a need for AI. And if we get to a point where it is actually difficult to put a match together that would also mean the population is so low that it wouldn't make much sense to keep the game going anyways.
I suppose you could let AI take over disconnected mechs, that's honestly the only realistic place for AI in MWO that I can see.
You're not going to see any serious PvE content for MWO like missions or even training scenarios vs bots, even if that would be nice it's a major investment beyond what can reasonably be expected. You'll see stuff that can be done with XML edits and that's it, possibly new mechs, that has already been true for a year and development is winding down not ramping up.
7
