With regard to PvM, its fun to read the speculation. The reality is that its not very hard to create a creature that can outsmart any player in the game, regardless of their stats and resources. The problem that Jagex faces isn't that of insufficient artificial intelligence... it is a problem of creating an enjoyable challenge with good repeat playability. If a creature has a super-high kill rate on players that challenge it, one of two things will happen (at least typically within the realm of Runescape). In the absence of a worthwhile payoff, most players will give up on it after a while, because of the futility of the effort. If there is a reward worthy of such an extreme challenge, then it may skew other areas of game play by giving the owners excessive advantages of one sort or another (the one possible exception being if it grants access to an exclusive area for additional gameplay or some such other unique feature of the game). Any monster foe needs to be hard enough to be challenging, but not so hard that people are consistently discouraged from facing it. Ultimately, Jagex wants to create playable content that will appeal to at least a reasonably-sized audience, otherwise those development resources are better used elsewhere. They have been responsive to the increasing skill levels of players by progressively introducing more difficult creatures for us to confront. Make no mistake, though...this is not a reflection of the upper limits of their programmers abilities (or the Java platform) in most cases - it is simply a reflection of the practical realities involved in developing content for a diverse audience. So once it is all said and done, I think the important question is "Is it worthwhile to develop a monster that regularly outsmarts most of the players".