Steve Wozniak: "I worry about everything going to the cloud"
france24.com
france24.com
"I really worry about everything going to the cloud," he said. "I think it's going to be horrendous. I think there are going to be a lot of horrible problems in the next five years."
He added: "With the cloud, you don't own anything. You already signed it away" through the legalistic terms of service with a cloud provider that computer users must agree to.
"I want to feel that I own things," Wozniak said. "A lot of people feel, 'Oh, everything is really on my computer,' but I say the more we transfer everything onto the web, onto the cloud, the less we're going to have control over it."
I wouldn't be surprised if many people around HN do not agree with Woz. After all the cloud is how many here earn their living.
That said, Woz is absolutely right. I wish more people of his standing would speak up.
There's a weird split. People running a service do not want to have to depend on a 3rd party service; but they do want to encourage other people to depend on the service that they are providing.
If I'm the provider of said service, Amazon are a potential competitor.
If I were to use a cloud to store files, I would treat it like my hard drive, ie, I would want to back it up. Nothing vital to me would ever sit in or on a cloud. I've seen too much go wrong to ever trust them.
Overall, I just don't understand why people who have decent, cheap local storage and back up ability already, while retaining complete control, would suddenly hand all their vital data over to somewhere where you lose all control. Even if I want to get my data from elsewhere on the net, I have a cheapo Iomega drive that allows that easily. This cloud lark simply makes no sense to me at all. It just looks like another corporate data grab, which added snooping.
Cloud = Kings New Clothes. I feel like the little boy.
> I see no practical use and only danger. Outages,
> ownership, authority snooping, seizures, hacking,
> etc, etc.
> Nothing vital to me would ever sit in or on a cloud.
> I've seen too much go wrong to ever trust them.
> Overall, I just don't understand why people who have decent, cheap local
> storage and back up ability already,
> while retaining complete control, would suddenly hand all their vital data
> over to somewhere where you lose all control.
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.
Woz is right on the money - anything we put in the cloud make us be on the mercy of provider's business plan, reliability and plain quality of QA processes.
I couldn't care any less if my Facebook posts or Twitter updates were wiped off the face of the earth.
I would care if my email archives, photographs, music collection, etc., were destroyed, but with varying degrees of concern. For example, I'll often buy individual music tracks from iTunes, but if I know I want a whole album, I buy a physical CD. I might only use the CD once to rip it to my computer, but I have an automatic backup that, so far, seems pretty robust, as I still have working CDs that I bought in the mid-1990's... which is more than I can say for my email archives!
But there is no reason we couldn't do that with a basic home or small office network, adding basic VPN functionality to allow remote access if necessary. It's not particularly difficult to set these things up today for anyone who can set up a broadband router and home wireless network. If they were the must-have technologies being promoted instead of the cloud then it would get even easier, as the people making the equipment built more user-friendly interfaces that didn't need much in the way of technical skills at all, and as ready-made server software to co-ordinate your mail or calendar or music collection or whatever became as simple as a one-click download.
My contacts? Not vital at all. My presentations? Not vital at all. My music? Not vital at all.
Somehow I suspect you'd disagree if all your contacts started getting phishing e-mails that looked like they came from you, or if you blew a million-dollar business deal because your presentation wasn't available when you visited a prospective client, or if you had thousands of dollars' worth of music that you'd spent real money on and it disappeared overnight because the service provider closed down the system or you didn't realise you couldn't transfer something between accounts or you sync'd the wrong thing the wrong way. Things like these can and do happen, but I think there's a perception of "It'll never happen to me".
Because when I figure in my time needed to configure and maintain this decent, cheap and local storage it is not that cheap anymore.
I guess you're looking at this from a personal point of view. I'm not sure it's really true even then, but I acknowledge that setting up and maintaining IT systems today does require technical skills that most home users don't have, and if your new tablet or laptop or smartphone ships with the vendor's pet cloud services already integrated then of course that's an easier user experience since it just works (as long as you're happy to use those services and don't want any others).
I don't really buy that argument at all for businesses, though. Usually there are all kinds of gotchas if you're doing serious work in the cloud. Just look at the string of pathetic excuses and weasel words every time Amazon's services go down about how you should have used redundant something-or-other and different availability-wotsits, and then look at how many businesses either hadn't understood this or hadn't wanted to spend the extra money or simply hadn't done it yet and wound up offline as a result. I defy anyone to claim that the time and money to configure cloud services to be reliable per all these recommendations is harder than setting up equivalent equipment on-site or with dedicated hosting and a decent off-site backup system, assuming that you have someone who knows what they are doing in each case to do the work.
Powerful vested interests working to amplify this myth because in the end you get a indentured customer base that you can fleece for years.
Uhh. You must have a really tech-savvy family...? I get asked about setting the microphone volume.
I believe the main reason we don't have that already is that the cloud computing PR machine stole the developer mindshare first. The kind of people who might have started companies to cater to the market I described above have instead started companies based on cloud technologies.
But there is no technical reason I shouldn't be able to run a simple installer on any home computer that sets up, say, a simple mail store and an IMAP server so all my devices can connect to it. There is no reason we couldn't have GUIs no more complicated than today's e-mail or calendar or IM software to configure fetching your mail or planning events and sharing invitations with friends/colleagues. There's no reason, aside from obnoxious DRM schemes perhaps, that we couldn't have a one-click-installed home media server that can act as a hub for your music/video/streaming pay-per-view/whatever content.
The technology to do all of these things is basically already there, it's just too often the case that it only runs on one OS, or requires messing around with a command line and text configuration files to set it up, or requires figuring out an absurd amount of technical details from 57,249 half-complete README files and three-year-old HOWTOs.
I'm actually a little surprised that Microsoft hasn't fought back against the on-line/cloud push, where it has been utterly destroyed by the likes of Google and shows no signs of resurrecting itself any time soon, with a comprehensive bring-your-own-cloud campaign that emphasized things like security, flexibility and reliability where the cloud hasn't exactly lived up to its own hype. They already had the required dominance on the desktop and in the business software market, and I can't help thinking that if they had tried to out-Blackberry a severely weakened RIM by offering a solid business product range instead of launching one underwhelming consumer phone after another, the mobile device market would look very different today. And once MS had the brand and credibility at business level, it would have the experience to attack consumer space from a position of strength, and the warchest needed to acquire a few key on-line/mobile players to help that move. But what do I know? I'm sure Windows 8 and the Surface (whatever that brand means this week) will be much more successful. ;-)
If you find running a personal server too difficult try tonido or any of those plug computers. You don't need to take my word. Just try and see. You will be pleasantly surprised.
> The fact is if you can change the light bulbs you can
> run a personal server.
I'm personally more interested in personal cloud type services. Little plug computers that you hook to your wall and run services such as FTP off of. You pay for brain dead easy DNS service at whatever rate and all of your data belongs to you. If your DNS provider gives you the finger you switch and give it right back.
Right now these devices are expensive. In time though they'll be cheap. So I'm just waiting on it until I can build my own or buy one for <$50.
(1) Like code windows open next to this one, right now.
I'm not exactly a patient man. If I suspected I'd be waiting long I'd start my own project.
http://en.wikipedia.org/wiki/Homomorphic_encryption http://blogs.teamb.com/craigstuntz/2010/03/18/38566/
Would I worry about all my data living in one enormous datacenter in the USA right next to everybody elses, all owned by one gigantic corporation? Sure.
Would I worry less if it was hosted locally by a company in my home town with a datacenter that I can actually go and visit and some agreement in place so that I can have take a copy of the raw data myself and have their copy deleted? Certainly, not to mention that the latency would be better.
I think fast upload is the answer to this and I believe that google fiber plans on providing this. In that case you can simply have a home server which is basically a NAS+Router on steroids with a static internet IP (thanks to IPV6) , this server could provide a pretty web interface for "normal" users and a unix shell for l33t h4x0rs. Since upstream is fast enough the service is going to be indistinguishable from something like dropbox in terms of speed and I will have full control.
The issues here are data and power reliability , what happens if the disk in my home server dies? There is only so much redundancy you can cram into a small low cost box. I guess the solution would be third party backup services that take advantage of fast upstream to sync the contents of your local box and perhaps can also perform as a "failover" box if your local node fails. The difference here is that you would get a choice of third party backup suppliers, though I imagine most ISPs would provide this as part of the broadband service.
One thing that does concern me somewhat on HN and other development oriented websites is that people do a lot of talking about stuff like APIs and Web services rather than protocols and standards. Web APIs provide the illusion of data portability since you have some way of interacting with it, but the reality is that you need a way to dump all of your data and you also need 100% compatible software to move it to.
I guess it's difficult to make a business case for designing Email 2.0 unless you know that you can monetise users through some kind of lock-in. Otherwise you could design a fantastic open service only to have someone copy it , undercut you on price and steal all of your users. For example if you could set up your own third party "facebook server" that allows users to interact with FB users without agreeing to facebook TOS or seeing facebook ads.
http://gizmodo.com/5926598/the-amazing-contents-of-steve-woz...
Is he not a bit past his sell by date?
After seeing the TWiT.TV show about Roll your Personal Cloud (http://twit.tv/show/know-how/1) I have started using Tonido (http://www.tonido.com) for my remote access and sync needs. It can be run completely local and authentication is between my browser and the Tonido device. I am happy sofar and completely own my data.