Bill Gates, Steve Jobs, and Steve Wozniak were part of the Digital Revolution where they wanted to decentralize data and put computers in the hands of the people.
Now it looks like we need a backlash.
No, the solution isn't centralization of our data systems. You can already see where that is leading with the high profile exposures today (Sony, Target, et al). It is a fallacy to assume corporations have all the answers, or will act in the general public's best interests. Short term profit is the only thing that has any meaning in that system.
At the same token we can't continue going along like we are - as that is already proven to fail.
The very thing that makes the internet useful for communications and commerce for large populations spread all over the globe, is the same thing that is at the core of it's weakness: public key encryption. To be more specific, computers are designed not to be random, and the systems we've devised to get around this problem have limits that may be exploited. When paired with encryption these limits open up potential exposure, and advancements in computing technology allow those exploits to be more readily used. For certain short term transactions, this level of exposure may be an acceptable risk - for data that is transient in nature, and not useful to someone at some future point in time. However, much of the data we trust to encryption could be useful to a 3rd party in the future.
We could ensure our systems (personal or corporate - doesn't matter) are completely secure from a remote attacker - by placing them inside a Faraday cage, and disconnecting them from the internet. While the data would be secure, it wouldn't be very useful in the broader context of communication and commerce - but for some types of information it might be an appropriate approach, and I imagine is what some sensitive government networks opt for their classified systems. For all other systems it would be as useful as throwing them into the deepest part of the Pacific Ocean - secure, but useless.
In order to communicate on the wider stage then, we must accept a certain amount of risk. I think we are all in agreement that the current risks are unacceptable the way they are today. I also think there is no single magic bullet. I think you will see the teams focus on the following areas, assuming corporate interests are not overly impacted by the potential solutions:
Tools - tools need to be devised that don't allow neophyte application programmers to shoot themselves in the foot.
Training - training has to be developed based upon new approaches, and made available widely.
Willpower - everyone - corporations down to individual developers - must have the willpower to do some things that might be hard at first (e.g. code reviews of all code - including libraries, refactoring/rewriting same in light of security issues etc) - and these things need to become habit.
Whatever the outcome, there will be no silver bullet.