Seeing as how this has been a talking point recently, i thought i'd give my two cents, and instead of just bashing the hell out of the idea, provide some thoughts on how it could not necessarily be terrible as long as certain conditions are met, as well as a few advantages. I am talking about something similar to the PS TV+PS Now, not the XB1's implementation of cloud, as that one loses some of the advantages.

I'll start with the biggest pro; hardware failure. Something like a PS TV or any kind of streaming box is simpler, won't produce heat, and is devoid of components that'll risk making the device unusable under certain bad conditions/circumstances. This decreases costumer frustration with hardware problems, dealing with customer service, waiting for weeks for replacement units, etc. And in the event of the box dying after the warranty is over, it'll always be cheaper buying a new one than it is buying another console. Isn't that "supposed" to be related to one of the advantages of consoles over PC? Truth is i've had more problems with consoles than PC ever since the previous gen..

The other pro is the one that MS is trying to promote; increase in processing power without dependance on local hardware. Theoretically achieving the kind of graphics you'd expect from a 400-500 dollar console would no longer cost 400-500 dollars, but much less

For the cons, I won't bother going much into them because people are already quite aware of most of them; internet requirement, input lag, compression, etc

So instead of i'll say what i think about when this model of console gaming could be implemented; what conditions could make cons be less relevant and bring the pros into the picture

1: Game pricing. Unlike retail and digital purchases, this kind of service is 100% dependant on the service provider, meaning you always depend on it to play the game. This doesn't happen even on Digital, where, at most, you lose the ability to download the game (if it's "pulled"), but not the ability to play it. Because of this, there's an approximation to the renting model. Pricing should reflect this; more expensive than renting because there's no specific time limit, but cheaper than buying because the costumer knows he won't be able to play that game "forever". Retailers would become obselete and the service providers would not have to compete with them, only with competing streaming services, driving the game prices down to acceptable levels

2: Tech advancements. The most important one IMO here is buffering or temporary storage. Basically, the box would have some kind of flash storage memory that would temporarily allocate game files to prevent DCs if either your internet fail or the streaming server fails/goes under maintenance. It wouldn't do it with the whole game, but at least enough to let you keep playing for a while; somewhat similar to PS4's PlayGo system. Of course this applies to SP games only. Other things would be the reduction of visual and audio compression, as well as minimal input lag... Also mandatory dedicated servers for games with online components (meaning that a third party could not make a game with P2P MP and sell it on the service)

Seeing as how the above would certainly NOT be doable by the time next gen comes, i could only see it happening on the one after that. Or the companies could try to bring it next gen already but without these last "conditions" being met, in which case I'd just reject it entirely.

What do you think? Still wouldn't accept cloud gaming? Would accept it even without these conditions? Or are they enough for you?

game developers said they don't want to render graphics using the cloud. Its bad for graphics, but good for AI.

Once the law passes that allows ISP's to charge more for bandwidth heavy services, cloud gaming will probably be put on hold for the near future. This would mean you would have to pay for the cloud gaming service and a premium for the excessive bandwith use.

It will make costs rise to a level that it can't compete with the tradditional gaming systems.

I doubt cloud gaming will ever be used at large like sony is doing with stuff like PSnow. That model of it would be just too expensive to be on the cutting edge of gaming tech. Easier taking refurbished/reclaimed PS3s and turnig them into server racks to power PS3 games. What I feel will be done though is levergaing that power like how MS is proposing. Breaking the game up so that part of its processing would be in the cloud. So those PS3 servers can ultimately be used to handle physics, crowd physics, skyboxes...etc. Those are things that won't affect how smooth a game should feel.

Downside of that though, is the whole always online thing. And that is what I completely do not want, the day when being online no longer is a feature but an absolute necessity. I can get being online to play online multiplayer. But the idea of being online to just play ny game I have is very very unsettling.

I think that's a bad model, too much of a middle ground that soaks up disadvantages with minimal advantage. Commitment should be made towards either one side or the other. Otherwise you're wasting money on hardware that is only partially required and is STILL prone to faults

Eh, not interested in the cloud future anyway lol.

I agree. Though technically this model is exactly like what you pro of having a console that would cost les but be able to achieve more. So you can have a PS4 that cost $300 but be able to rival physics and group AI that you can only see in much much more powerful PCs. Which in turn frees up resources within the console to be used for other things.

But I don't fel going all in in one direction or the other would ever work. There are just some areas that game streaming simply doesn't work well with. 

You're second point defeats your biggest pro. A box that can download partial game data and play the game locally when needed is not a simple streaming box.

Your second biggest pro is debatable. Netflix doesn't provide blu-ray quality movies, why would game companies provide cutting edge graphics with streaming. It will be just 'good enough' to save on costs, which will happen to satisfy your point of game pricing.

Cloud gaming will work because of convenience, but it will never meet the quality or responsiveness of a dedicated console, never mind what you can get on a gaming pc.

I dont want to give some remote server complete control over my gaming life.

Firstly, just to throw it out there, cloud computing is in no way whatsoever new technology, just FYI for those who don't know. It has been used extensively for 7-10 years by large software and services companies, and the concept was invented decades before that.

Secondly, here is where I stand on it in terms of gaming. The benefits it could achieve are staggering such as - No new hardware. No complicated setup. No game discs. No digital downloads. No game installations. No game patches, latency improvement, unified device gaming capability instead of "exclusives" forcing purchasing decisions and on and on.

My honest opinion, it won't work. Overutilization would crush its stability if it was depended on globally. Clustering, maintenance, monitoring, virtualization, storage, patching, and everything else that goes with maintaining a dedicated gaming cloud infrastructure will be EXPENSIVE. There is no way those costs won't be forced down to end users who mostly already don't agree with the premise of full cloud reliability. These demands would greatly increase and demand powerful local hardware when 4K is the new standard which, IMO is not that far away. Take a look at OnLive. It is a good idea of what we would be able to expect from dedicated cloud gaming - lower resolution and reliability issues rampant.