Future Internet: An opportunity not to miss

I am attending the 2nd EU-Japan Symposium on the Future Internet, and gave a talk yesterday. It was met with stunned silence - not that I expected much else, we were already late for the party. But the point I tried to make is rather too important to miss:
Both the EU (in the 7th Framework) and Japan must involve users in the projects. And they must ask the right questions.
It is not sufficient to set up panels to ask people which resolution they prefer, although it is even worse trusting experts. Ask them that, and you have already biased the users. You must show them the stuff - and then try to figure out what they think. And as they will not know what you are talking about, you must build a prototype. Not to waste money and time, you must do it early in the project, before you have locked down your plans.
What happens if you do not do that? You may well end up with a nice middleware, which is impossible to sell (been there, done that - twice) because it does not fulfill a real need; plus a pile of research papers that do not get referenced, and people who managed to keep their jobs in a crisis - which is good enough if you are one of them, of course. But if the EU (and NICTA) wants their investments to have an impact on new jobs and more employment (rather than just researchers keeping their jobs), they had better make a condition for new projects to involve end-users. And make the results easier to find than searching through research papers.
When you are implementing new things, you can not expect scientific rigidity in the same way, either. There are established disciplines which are gaining success (virtual ethnography, for instance), but the truth is that the market is binary: Either it works, or it does not. Testing will show you how, but otherwise you may never find out. Then you tweak it, and reinvent it the next day. Eventually it works, but it may not be what you thought it would be, especially not when you started.
When you say this, people will complain "but the pace of innovation in the internet industry is so fast! We work with physics, we have much longer horizons". I still do not understand why. In the newspaper business, we had to reinvent the product every day (yesterdays newspaper is old news). On the Internet, the user is used to get a new product every time they open a web page - even if they have been there before. You just tweak the product incrementally - continous beta is expected, not condemned.
Google does this for the platform for the information products. The economics are easy: They make money from the large numbers, making sure there are many people using their products (this is why they try to close the web). They want them to come back and check out what happened (after all, they also sell advertising). There is no reason you could not use the same process for radio, or fiberoptics, other than mental block. And that mental block says "we are doing hardware, not software".
Everything is software nowadays, even hardware. Getting a new press tool to bend sheet metal takes time, for sure. But this does not apply to programming an FPGA. That is hardware in software; when you can store things in the regular memory, it becomes even easier. You can release it and tweak it and release it and tweak it until it takes off.
If you do that with a closed user group, the process is faster than if you do it on the open Internet (because you know who to ask for feedback). You could do both, of course. But you can not do it without involving end-users, and without being extremely critical of yourself. Kill your darlings is not just a slogan, it is a way of life - you remove that which does not work, constantly. Funny that the Japanese have not caught onto this, "nemawashi" actually means working this way. Of course, if you apply it to an old boys network, it will hurt.
Being critical of yourself, you can start with the user. Does anyone really need 3D, if the only programming is chess matches? Does it matter what resolution you give them? Maybe you should start with the programming, and then the technology will follow? And you could tweak it as it was accepted by users? Putting it on a testbed is an excellent start - but showing that it works in a network you can do in a simulator. You need users to be surprised, and still the "living labs" leave a lot to desire in that regard.
It would be nice to see the European Commission and NICTA start thinking this way. I will not hold my breath. But you never know. The new Japanese government may be smart enough to open up research to the world; and the European Commission is being replaced as their terms expired. Perhaps there will be some breathing room for initiatives which give result in the future.