I used to manually download all my movies, shows, music. Stored it all on big terabyte hardrives. Used to have to buy hard drives all the time. Now? probably weeks since the last time my computer had to actually wake from sleep and spin that drive up. These days, Rdio is a thing. Netflix, appleTV, hulu these are all things now. The whole idea of using windows/mac file manager to fiddle with media at the file-level, across all my computers, laptops, tablets, phones that’s crazy batshit talk. Who would do that anymore? There’s dozen great platforms that virtualize all that for me. Thanks to airplay, even virtualizing which device is playing the media and which is displaying it.
Enough to make you wonder, how the hell is Tivo still in business? How are DVRs still a thing? Millions of people paying millions of dollars for local hardware to make millions of recordings of media that could perfectly well be served by one copy stored on the cloud? WTF is up with that?
Someday there will be a real netflix even for PC and console games. Built from the ground up to be virtualized.
Just to play games people buy expensive graphics cards, power-hungy PCs, console boxes that only get upgraded every 5 years. five! Millions of these get distributed, but the duty cycle is really low. Most of the day they just collect dust, take up space, consume idle power. What a waste. Why can’t I rent or subscribe to some awesome gpu and compute cluster at edge of the cloud to pump-out awesome, immersive experiences to any screen, regardless of it’s local silicon limitations? Huge opportunity for disruption here. Of course, we’d all need to have fibre or much better internet speeds that we get today.
AMD gets this. NVDIA gets this, they’ve recently introduced products designed to be run in racks for remote rendering of games. Valve/Steam, understands the vision as they continue to experiment with ideas of streaming gaming across remote or local networks. However we’ve yet to see a real compelling platform player put all these pieces together. MSFT and Sony missed this boat with their latest generation, they still try to support physical media of all things. Why didn’t they just but a cassette deck in there.
Webmail, salesforce, facebook, the web in general have been proving for decades that there are huge classes of apps that work a lot better when they are architected to be entirely native to the web.
But it’s all piecemeal. I can run almost all my daily apps on the cloud, but i can’t run my whole desktop from the cloud. I have to install apps for all those apps on all of my screens. I can’t seamlessly move from screen-to-screen and keep all my state and harware resources available. Although, in cases when I can, it does feel magical: like searching for a maps location on my pc and seeing the driving instructions pop us as an alert in android, like pausing netflix on my bigscreen and pressing play on my tablet. Why can’t my whole computing experience be seamless from screen to screen by default?
We need to redefine what we mean of a ‘computer’ or an ‘operating system’ from a machine-centric perspective to a cloud-centric perspective. It’s almost like “the network is the computer”.
today amazon announce this:
“Amazon WorkSpaces allows customers to easily provision cloud-based desktops that allow end-users to access the documents, applications and resources they need with the device of their choice, including laptops, iPad, Kindle Fire, or Android tablets. With a few clicks in the AWS Management Console, customers can provision a high-quality desktop experience for any number of users at a cost that is highly competitive with traditional desktops”
Maybe the enterprise will go first. Amazon is talking about supporting legacy desktop apps and IT management. However, as innovation in business tools is increasingly driven by cloudware as opposed to desktop apps, it makes more and more sense to just offload the whole stack to the cloud. Particularly with the benefit of portability across pc, mobile and tablet devices.
However. This great virtualized cloud future comes with several major barriers. These barriers to further innovation have names:
- The telcos: Bandwidth and latency are the key to enabling smooth remote and visualized computing experiences. Until fibre is much more ubiquitous, until consumer (and business) network speeds and bandwidth caps continuously increase by orders of magnitude, nothing will happen. Why does it seem like desktop PC innovation has entirely stagnated over the last 10 years? In part it’s because (unlike with mobile devices) network bandwidth available has barely improved since broadband became ubiquitous. (in fact, some consumers have even seen their usage caps come down in recent years).
- Open spectrum: Freeing up more licenseable spectrum for 3G/4G is good too, but unlicensed spectrum is even better and even more sorely needed. no spectrum gets better utilization than the tiny slivers (at 2.4 and 5.6GHz) that are reserved around the world for open usage like wifi (and a thousand other things). The aggregate economic benefit could be vastly greater if governments would set aside just a bit more, instead of always auctioning all new spectrum for brief one-time revenue gain.
- Microsoft: This company still has waning, but enormous control over our basic computing environments, especially in the enterprise. After all these years, they’ve really only come up with two really killer apps (Windows and Office) that just happen to owe their entire revenue model to per-machine licensing. As a result, Microsoft has always had a very conflicted strategy with regards to the cloud.
- The cablecos: Protectionist and anti-competitive strategies by cablecos and telcos to protect revenues on last century’s media distribution models is probably the single biggest barrier to ongoing innovation in computing. See also #1
- The powers that be: Trust in governments is not at an all-time high. And recent revelations haven’t been helping. Until we can somehow mutually resolve issues of due-process and security from domestic and foreign governments, a lot of people are going to be reluctant to allow all their base onto the internet. Maybe we just have to come to grips with the fact that spies are always going to try to do spy shit and move on.
Very interesting post. With Google wiring whole cities and taking on traditional on Microsoft with Sheets etc, perhaps enterprise will just make the leap beyond MS.