Read Write Web ran an article (Google's Eric Schmidt on What the Web Will Look Like in 5 Years) which include the following summary of prognostications. My comments are indented from main topics.

Which may be true, but it will only matter to people who speak Chinese. Discovering relevant relationships in content that surface actionable insights will be key. If I were CEO of Google I would be worried a bit about this because making sense of the world’s information will require cross-links between languages, and very very good translation. The stuff the really matters really needs to be right, so having an accurate view of text across languages, as well as an understanding of cultural context and nuance, will be critical if this matters beyond the speakers of Chinese. Choose your dialect.

That is one view. It is not the only view. A major cyber-terror attack could actually reduce the amount of cloud influence and drive people back to a more myopic and insular view of information. Other information distrust events may cause the same behavior.

Moore’s law is about raw power not application. If we don’t start thinking about distributed and parallel computing models all of that raw CPU-power will go to waste. The key word in Schmidt’s sentence is “capable” which doesn’t mean “realized.”

Yes, but there will be new business models on the web as media attempts to figure out what it means. When all the pipes are big and all the media high resolution, we will need to invent new ways to differentiate.

Perhaps. But will it be corporate and personal, or will it be generative and interactive. Will the video be more game-like, more dream-like, less watch and more play.

Great, but if information is a commodity, making sense won’t be. The winner will be the company that goes from finding things to helping individuals make sense of what they find, and what it related to it that they didn’t know to ask for.

Duh. In ten years, I doubt that Twitter or Facebook will matter. They will either really matter, as in, they will be a platform around which our worlds revolve, or they will be a Wikipedia entry we all edit fondly with our recollections of posts past.

See real-time comment above. It isn’t about ranking,its about sense-making, which is a completely different business model. What, me worry?

There is another version where trust erodes and we return to vetted information. There may be so much user-generated content that we become overwhelmed with the uselessness of it all and even the entertainment value wanes. People desire to re-establish trust, and new models of trust will emerge, and with them, new sources of trusted content.

Forecasts should always be couched in the context in which they exist, with clear caveats. Not having seen the entire presentation, I can’t say if Schmidt did this or not – but all good forecasters should. If your going to talk about the future, prove that you thought about it robustly and understand the variables and the uncertainties in play. As I say at the top of this blog, you can’t be wrong today, but over the course of time, you run the risk of being wrong much more often than right, so thinking deeply, understanding context and the interplay of uncertainty is crucial to creating strategic plans that are meaningfully informed.