WOOT! I did it!!!

[Zoombie]

Neat. Now KIA is ever more...killy.

[FizWizz]

KIA != KAI

[Neddie]

You cheat because you have a hundred million years of evolution going for you, compared to the AI which has a couple of days' worth of code. I think it is perfectly reasonable to even things out a bit. The better AI's will cheat more subtly, though.

Evolution that has absolutely no bearing on your ability to perform in a constructed game, sure. But thank you for the buzzkill.

[Arco]

I think KAI has got its cheat thing perfect. all it does is know the rough location of units. it dosnt see cloaked or get a resource buff. it plays well, but it dosnt learn which means you can find tactics that work and use them. KAI is the only AI I see as a good match for any player at EE (with the appropriate handicap for some.) and the best around for a fun game.

Hmm, it's been a while since I last used KAI thanks to development in AAI, but one thing I do remember is it seeing my cloaked units. Which was very bizarre and killed a lot of my available strategy. I forget if that was indeed later fixed or not.

Also, you speak of KAI not learning and therefore being more of a fair challenge despite the cheating because you can exploit static flaws in its gameplay. This is never the right approach. Exploiting bugs is not gameplay. Do we speak of matches where one player of inferior skill utilizes exploits to compete as "good matches"? No. We go, "man, that total loser had to resort to cheating to do anything, what a failure." I want to fight an AI as fairly as possibly--this means both it and I see the same things and can do the same things. Maybe they won't stand a chance against a good player. Well, then those players can fight multiple AIs at once, or use handicaps, or impose limitations on themselves. I'm not a fantastic player and on some maps even AAI can mess me up pretty well. I know that if I'm getting hit hard, it's because I didn't expand enough or I split up my forces poorly or I didn't have enough radar coverage or I didn't make a good unit mix or whatever--not because the AI can see things that I can't or get more resources than I can.

The problem with cheating AIs is that they change the rules of gameplay. For instance, one good tactic for fighting a player is to kill their scouting units when possible. The less intelligence they have, the less of an advantage they have when attacking. But try as I might, it's impossible for me to blind KAI because it sees everything regardless. That means a whole element of gameplay has been removed. That does not enhance my experience. That means that instead of being creative I just have to try to throw more units at it. It reduces the number of strategies I can use. How do people ever see this as the "right" way to make a competitive AI?

[Forboding Angel]

Kai Is definately not a fair AI, but it is one hell of a challenge, which is more than I can say for any other ai available.

[mehere101]

I have to agree with FA. If we didn't have AIs that could cheat people would be complaining that the AIs sucks. Lets face the facts, springs Global AIs are well ahead of most games AIs. A lot of them use not only map hacks, but resource hacks.

[krogothe]

Yes, KAI hacking pisses a lot of people off, its really not fair, but its a design decision. I have planned to make the cheating less obvious, but its far back in the list of priorities. The reason for that is if you want a fair game, you can use any other AI or play with humans.
KAI is there to fill the niche for people that DO want a cheating AI that can hit your weak spots, keeping you on your toes so you can improve your game. Ok well, it sucks balls atm and doesnt really improve your game at all, but thats the plan for the future! Good things come for those who wait!
Ill take a look at this unit list later and hopefully decypher what bad choices KAI is making (its been so long since i tested it i dont even know anymore).