FWIW, as a confirmed cloud sceptic, I think it's about a lot more than just warrantless searches and accidental remote wipes. While these are legitimate concerns, I think we'd probably all agree that they are unlikely to adversely affect most cloud-hosted customers.
The more serious concerns, to me, are about cost/benefit, availability, inconsistency, and plain old quality.
On cost/benefit, a lot of the cloud hype seems to assume that outsourcing functionality and data storage will lead to cost savings, yet I have seen little evidence to support this case in reality. Certainly if you look at the raw numbers, using things like AWS or GAE are far more expensive per unit of processing power/storage/bandwidth than any number of traditional local or dedicated hosting options, unless you really do have computing demands that go up and down like a yo-yo over very short time intervals, which hardly anyone actually does. So then we're down to efficiency savings in operations/sysadmin work and redundancy in the event of hardware failures, and I struggle to see how multiplying the basic costs several times over can possibly be offset by those overheads, even if you assume (obviously incorrectly) that there is no overhead at all to configuring and maintaining cloud-based systems. (Edit: In case my point here wasn't clear before, this doesn't just affect those providing cloud services, it also implies a certain minimum level of mark-up for all these SaaS businesses that are built on these cloud platforms who want consumers and businesses to use their web-based software rather than running local applications.)
Availability is easy: If your Internet connection drops, you are a sailboat without wind. And while some parts of the world are enjoying near 100% reliable 100Mbps broadband, most parts of the world aren't enjoying anything like that reliability or those speeds.
Inconsistency is the Achilles heel of many consumer cloud apps, IMHO: I don't think the general population are going to be happy with endless minor (or occasionally not so minor) UI revisions that they can't control. Take a look at what happens every time Facebook do a major update, or the feedback about browsers that keep moving everything around as they auto-update, or try using Google Docs at work for more than a few days and watch your colleagues start pulling their hair out. The easy distribution and updating of web apps is great in some ways, but it does have a downside, and I'm noticing a marked increase in user frustration as more and more things are forever moving around.
Finally, again looking at cloud-hosted web apps from a user's point of view, the simple fact is that for a lot of software, what you get with a cloud app just isn't up to the same standards as what you get with traditional native apps. Google Docs (Drive, whatever) vs. MS Office? They're not even competing for the same space, because the Google stuff lacks huge amounts of elementary functionality, never mind all the automation/integration/customisation stuff.
In short, it's not just that some people want all computers to be iPads and all data to reside in the cloud, it's that they want all software to be written for five-year-olds and some of us have real work to do.
The more serious concerns, to me, are about cost/benefit, availability, inconsistency, and plain old quality.
On cost/benefit, a lot of the cloud hype seems to assume that outsourcing functionality and data storage will lead to cost savings, yet I have seen little evidence to support this case in reality. Certainly if you look at the raw numbers, using things like AWS or GAE are far more expensive per unit of processing power/storage/bandwidth than any number of traditional local or dedicated hosting options, unless you really do have computing demands that go up and down like a yo-yo over very short time intervals, which hardly anyone actually does. So then we're down to efficiency savings in operations/sysadmin work and redundancy in the event of hardware failures, and I struggle to see how multiplying the basic costs several times over can possibly be offset by those overheads, even if you assume (obviously incorrectly) that there is no overhead at all to configuring and maintaining cloud-based systems. (Edit: In case my point here wasn't clear before, this doesn't just affect those providing cloud services, it also implies a certain minimum level of mark-up for all these SaaS businesses that are built on these cloud platforms who want consumers and businesses to use their web-based software rather than running local applications.)
Availability is easy: If your Internet connection drops, you are a sailboat without wind. And while some parts of the world are enjoying near 100% reliable 100Mbps broadband, most parts of the world aren't enjoying anything like that reliability or those speeds.
Inconsistency is the Achilles heel of many consumer cloud apps, IMHO: I don't think the general population are going to be happy with endless minor (or occasionally not so minor) UI revisions that they can't control. Take a look at what happens every time Facebook do a major update, or the feedback about browsers that keep moving everything around as they auto-update, or try using Google Docs at work for more than a few days and watch your colleagues start pulling their hair out. The easy distribution and updating of web apps is great in some ways, but it does have a downside, and I'm noticing a marked increase in user frustration as more and more things are forever moving around.
Finally, again looking at cloud-hosted web apps from a user's point of view, the simple fact is that for a lot of software, what you get with a cloud app just isn't up to the same standards as what you get with traditional native apps. Google Docs (Drive, whatever) vs. MS Office? They're not even competing for the same space, because the Google stuff lacks huge amounts of elementary functionality, never mind all the automation/integration/customisation stuff.
In short, it's not just that some people want all computers to be iPads and all data to reside in the cloud, it's that they want all software to be written for five-year-olds and some of us have real work to do.