2900 possible for GM's?

I think 2900 is quite possible if:

1) ratings inflation keeps increasing at the rates it has been increasing since it started suddenly in 1986 (source: http://members.shaw.ca/redwards1/)
2) there are more 2800+ players (which there will be given #1) to feed the top player: if Carlsen or somebody else were able to dominate like he has, and if there were lots more 2800+ players for him to dominate, then that would push his rating up towards 2900.

Kasparov could probably have gone higher too if he had had more higher-rated competition. I think the key variable for the rating ceiling will be the difference between the #1 and the average of the peers that he plays against. The higher the average of his peers, the higher his rating will be as a result of dominating them, and as the difference increases, the effect of the inevitable occasional loss will be that much greater.

That may be valid if they were playing close to perfect chess or something, but computers show that chess players are far from perfect. There is still a ton of room for improvement in chess.

Moreover, the length of the track doesn't change. With rating inflation, the chess players are metaphorically running on a shrinking track. 

Bator Sambuev was rated 2725 CFC in 2010 and 2753 this year, so if his rating continues to rise at the same rate, he should crack 2900 five or six years from now.

Other players to watch: Claude Bloodgood and Zaw Win Lay.

In my humble opinion, the Elo rating system in its current form will never allow a player to reach 2900. It is simply too hard, due to statistical reasons and the K-factor.

To answer the question: it would be much more likely if FIDE games were rated with the "Glicko system", which is what they use here, on Chess.com

Disclaimer: I don't hold a degree in mathematics or statistics, but I know a little bit about these kinds of things.

Zaw Win Lay for those who are unfamiliar was the Myanmar player whose FIDE rating skyrocketed upward as he played mainly his own countrymen. 

Claude Bloodgood, Zaw Win Lay and Bator Sambuev all share the common characteristic of being a big fish in a small pond.

But for someone that is actualy playing at that strength, Carlson maybe, It will probably happen, and this is Stricktly gms here so if they dont complete thier norms it wont happen for those cheaters

I'm not at all convinced that Bator Sambuev or Zaw Win Lay could be described as cheaters.  They seem to be stronger than their opponents, and the fact that they were not facing international opposition was probably not of their own choosing.  Both of them play in FIDE-rated events when they get the chance.  Zaw Win Lay's rating has dropped quite a bit since he started playing internationally.  The small pond effect is a product of Elo's system, but it can be fixed if you bring the pond into contact with the main pool of players.  Offshoot groups that play frequently though will lead to inflation of the ratings of their top players if there is a substantial difference in strength within the group.

In any case, ratings are always relative to the group of players whom the person is playing most frequently, and are probably less useful when estimating the relative strength of two players who share no common opponents.

No, it's pretty much people thinking they know more about the rating system then they really do, and they don't want to acknowledge that the players, with computer aid, ability to play chess better is improving, or the achievement of these GMs. 

There is very strong evidence of regional inflation, shall we call it.  None of the Myanmar players were titled at the time of the bubble, but Zaw Win Lay's peek rating was in the 2630 range.  There were other Myanmar players rated in the 2500 and 2400 range as well.  Localized bubbles probably happen all the time.

I should break 2900 next week.

Kasparov's rating as 20 year old was 2710 and his peak 2851. Carlsen's rating as 20 year old was 2826...

In the early 90's there was a time when only Karpov and Kasparov (>2800) were rated above 2700. In the latest Fide rating list there are 47 players above 2700 and 4 above 2800...

In order to be have one 2900 there must be several 2820+ or an infalible player

Agreed with Retguvvie98. You'd have to go back to Steinitz to find somebody who was as far above his contemporaries as Fischer was in the months leading up to winning the world championship. There are of course many different measures of "who was the best?", and on most of them, Fischer doesn't come out ahead, but in this measure, there is simply no dispute.

For some different ways of measuring "who was the best?", see Jeff Sonas's excellent four part series at chessbase, The Greatest Chess Player of All Time: part 1, part 2, part 3, part4. Sonas also finds that Fischer dominated his peers to an extent that Kasparov, Karpov and others never did (see part 2).

in 20 years we will be unable to find our fridge without having the app and gps available since our brains will get more and more into regression thus retarded.

Hikaru now has 3200+ rating.