Original URL: https://www.theregister.com/2011/02/25/cloud_computing_could_lead_to_monopoly/

All hail our cloud computing overlords

Pray that the IT priests beat the beancounters

By Tim Worstall

Posted in SaaS, 25th February 2011 12:49 GMT

Comment We all know what's going to be so frabjous about cloud computing: it's going to make the damn stuff work. Yes, yes, we have computing systems that work now, but they need that priestly caste made up of Reg readers sitting in front of them, in charge.

Think of it this way: a coach and horses and a car both manage to get you somewhere. But in the one you need a skilled coachman to guide you, while in the other you can give any old idiot a key (and judging by the roads today, we do give any old idiot...) and they can manage it themselves.

What cloud computing should mean is that the user will, just as with the modern car, not need to know anything at all about what is under the bonnet. Pedals and steering wheel, keyboard and mouse, that's all the user will need to know about.

Do note that this doesn't mean then end of the priestly caste: there are far more mechanics and autotuners around now than there ever were coachmen – and they're better paid too. It is just that the invocations and spells will be woven in private, not out in front where they frighten the horses.

But is it really all going to be that lovely and wonderful? That rather depends on a few little arcane bits that come with the economists' secret decoder ring. For there are two possible concerns: monopoly and whether, in the absence of such, competition actually does lead to better outcomes.

It can at times seem almost blasphemous to ask the latter question. But of course competition leads to better outcomes! Well, yes, except in those times and places which it doesn't. Competing armies are what gave us the Wars of the Roses and that wasn't generally thought to be all that enjoyable or improving.

More recently there's been some intriguing research into competition in healthcare. The last set of changes in the NHS gave us a natural experiment we could explore. And what was found was that competition in healthcare on the basis of price did not lead to better outcomes. It led to cheaper prices, yes, but to a deterioration in the standard of the healthcare itself.

However, when the competition, the market, was rigged, so that it was a fixed price market and the competition had to be based on the quality of the healthcare, then the quality of the healthcare did indeed rise. But we've not quite come to a final conclusion on exactly why this should be true for some items and not for others; competition on price seems to work just fine in many areas of the economy but obviously not in all of them.

My vaguely informed guess is that this problem arises when the user of the good or service is not the one who is making the decision on the quality/price trade-off. If we were trading off the 1 per cent or 0.5 per cent risk of death during a hip transplant (not far off likely numbers) against the £5,000 or £7,000 bill (similarly) then we, looking into our own wallets and summing up our own reactions to risk, could make our choice. But when it is some bureaucrat deciding for you, then that trade-off could well be miscalculated.

Who will win the battle?

cloudcomputingFight

There can be only one...

Which brings us back to cloud computing. Almost by its definition it will be the large organisations of this world that really take to it: which means that it ain't gonna be the individual user who decides which, if any, of the systems get used. It is going to be something of a battle between the IT bods side, arguing for quality, and the accounting side, arguing about price.

And who will win that battle? Who will be able to steer the decision about the various competing suppliers, and which way will it go, between quality and price? Which of these will determine whether competition in cloud computing leads to a price-driven race to the bottom or a quality-driven race to the top?

I think we all know the answer to that. For the sake of all our economic futures, I hope it is you, the IT priestly caste, that wins, and not the beancounters.

As to a monopoly arising, this seems like, at first blush, a rather odd worry. The cloud, when properly implemented, shouldn't care what equipment is hung on the end of it as input or output devices. Operating systems (to the minimal extent that there are any) will have rigid interfaces with the cloud.

Other than that there's no lock-in at all: so we should have less prevalence of monopoly, fewer ways in which there can be a monopoly OS, or desktop environment, and even less possibility of a monopoly program (such as Word, Excel etc). If all inputs and outputs must conform to the (likely, browser) interface and what happens either side of it is a free for all, then how can there be a monopoly situation?

There could be an effective monopoly, sure, in the sense that someone, somewhere, has provided the best bit of kit to do a certain job – and so everyone uses it. However, there's no lock-in there, which is a necessity for a company to be able to set up a monopoly. And would it really be so bad if we all ended up using the best tool available for a particular job? But there just doesn't seem to be room in such an architecture for someone to be able to play, say, the Microsoft game of owning the desktop.

Except for one thing: data. The whole point of it all is that both the processing and the data sit off in the cloud – you don't know where and you don't care where. But, umm, what happens when you want to change your supplier for the storage or processing of your data? You need to know, absolutely, that you can retrieve everything that is yours and switch it elsewhere. You also want to make damn certain that no one else can get to your data, even if you do have a full copy.

Yes, I know, you've all already thought this through and have the necessary solutions, and I'm just worrying like an economist might. But if you thought that having to buy your kit or OS from someone who had you over a proprietary barrel was bad, just wait and see the pain that will come from having to cough up to someone who has all of your data tucked away somewhere. ®