One problem with remixing/editing and sharing is legality and liability. Since Facebook, Myspace and most of those social platforms are owned by big media corporations I think they'll heavily fight that. Instead, I envision F2F (friend2friend); P2P over IM (Jabber preferably). Yet if the platform used is centralized data can be easily censored or filtered. So to protect our privacy and work for net neutrality I think we'll see increased usage of decentralized platforms, and cryptography including over 3G. Its not just that. If I make a picture of myself on a train station and someone else is on the picture they legally need to give consent. But in a blink of an eye the photo is uploaded on Facebook. Is the world adolescent enough to handle this responsibility? I think not, I think we're gonna see more and more (self) moderation on the Internet.

One thing we'll see a lot innovation with is augmented reality. I think Google Streetview will raise in popularity as well. I envision all kind of user scenarios there which are useful to life. Maybe the person on photograph is recognized by Google and has a privacy policy. I've posted elsewhere examples of grocery shopping using a mobile device for 0) scanning in the products 1) getting cuts & deals 2) calculator of price (from total of groceries to even fuel used by car) 3) logistics (amount of food, perishing, active/passive reminders) 4) recipes (say what is missing, suggestions, reminders, etc). Heck, you could even link your inventory with your neighbors if you're on good terms with them, and help each other out for a change (or if you live with students). My question is not if this kind of application will see light. I'm convinced it will. My question is when.

Mix OCR/TTS/voice recognition with augmented reality and you get all kind of crazy things. Get the point? Its about I/O. You have input in one way and want output in another way. Then you can process further. Very important and difficult. But once its there. Wow. Examples. You tagged your data accordingly and are able to use voice commands to 'embed' them. For example you tell your mobile device to e-mail a certain picture to a certain person. Or to get that picure of your 2nd car and apply photo manipulation on it to see how it would look in different colours. Or you could browse Wikipedia this way. If there is an object you don't recognize you make a picture of it and augmented reality with search engine does the rest. IOW, besides personal data I imagine a whole library, an encyclopedia, of data available on the mobile device. I also imagine automated translation services eventually being applied on phones.

Now, I've seen application for photo manipulation, soft synths, and much more creative applications on iPod touch. Yes, its nice and fun, but its toys. Its nothing like a real VST, never mind a real hardware synthesizer. The quality is bad, the UI is bad (could be worse but the hardware synth is mimicked, knobs are difficult to control), you can't export it over MIDI or anything. Maybe a touchscreen can be used for this but then the touchscreen has to be bigger. Otherwise I don't see laptops being replaced in the music scene any time soon.

A normal screen can be used as output (maybe even foldable LCD?) with the mobile device being a remote. Yes, that makes some sense IMO, if the remote control UI application is optimized for running on the specific hardware. IOW, one has to think out of the box of the existent UI and think of the limitations and potential of the device.

Speaking of remotes, I guess my mobile device would be like a remote control for all kind of objects which currently have their own buttons. Instead, those devices only have a WiFi connection to AP. You could also see statistics of these devices on your mobile device. So if your fridge dies you receive a SMS about that. If your espresso machine needs maintenance this is added to your todo list. We want integration between devices. RFIDGuardian (Nokia Cellphone User Interface, API), Blinkenlights (tradtionally controlled by SMS IIRC), any many more examples.

Eventually we will see commercial photo manipulation software on (small) touchscreen devices but their UI will be very much optimized, and they will have their limits. The result is not the final work, or its purely hobby wise.

For video editing I can imagine something similar. Effects like Cheese does. Or mixing yourself in a Google Streetview picture because you've been at that place. But serious video editing requires a lot of raw power, and a good UI and application. Even if there are tools like this for Linux, they're for the Linux desktop.

For UI we'll see gestures to move to different parts of the UI. These have to work well and make sense. Carman, for example, is an example I like.

For gaming I think of RTS, even in collaboration, using touchscreen. Racing as well using accelerometer. And the device being used as remote for gaming consoles. But eventually the fun wears off, and the hardware is as-is. There is only so little you can modify on it. For example, I don't see a N900 easily replacing a Korg's KAOSSILATOR Portable Touch-Screen Synth.

Eh, I think the porn industry will also eventually catch up with augmented reality and mobile devices...

Hmm, its also important to use automation such as Electric Sheep and Debris. Would be nice to make your own Nokia background with some kind of application like that (although I love the default; maybe it could resemble the cellphone wave used when called lol). Something like fractal art, which is not that heavy on GPU/CPU and is also a matter of formulaes I can see applied on a mobile device. This can also be used just to make proof of concepts, and later exported for fine graining. I think that is some kind of key here: that the task performed on the mobile device is not the final end result, that it almost always needs final pinches of salt and pepper to get an acceptable end result. For example, I've used software synths to get some basic idea and quickly play around on laptop because my hardware synths are too heavy and can't have them all with me. That is the advantage of a computer: it can mimic instruments without the required space, weight, and price. But it cannot fully replace them. You're still bound to MIDI (or Melodyne Direct Node Access?).

In an ideal world collaboration all happens over IPv4/IPv6 and WWAN and hopefully those protocols and formats are open standards. Could be real-time, but I think is better to upload your 'audio diffs' of source of music to collaboration platform and allow your (specific) friends or the whole world to hack further on it like version control system or wiki. Same for photos and video. But with Apple strong foot in multimedia and artistic industries maybe they'll get the more serious stuff only ported to their platform, being incompatible with anything else.

PS: No, I'm not on psychedelics... merely pain relievers because of recent operation...

PPS: Oh, and Twister (tm)

Here 2 videos of touch screen MIDI controllers 1, 2

A table touchscreen where one or multiple musicians can compose using objects as instruments. The objects, in various forms and symbols, will be controllable by the way they are pointed towards the screen, connected towards each other, and in each other's influence field. The touchscreen provides additional control. ReacTable is the most astonishing multi touch audio application I've ever seen (ReacTable on Youtube, Official homepage), I'd love to try it out, have to see it in action some day... also impressive is AudioTouch and something like VJing.

In the end, size matters, and it should be clear for what 'size' this discussion is aimed at. Bigger touchscreens easily open up a can of worms like above whereas smartphone/internet_tablet size is more limited. Both will have their market, eventually.