I agree with Dr. Stein, although 10 years seems a bit long. I have watched as automotive M-CAD and E-CAD development have been exorcised from the car industry over the last 15 years. The car companies turned against their own children (human and software) with fury and hate. Of course, they are perfectly happy to continue using the software long after the developers have been 'offed.' The basic process involves using public derison against both the developers and the tools they made. COTs or other great commercial hopes are held out as the answer. When these prove inadequate or late, the old software is used in the extended interim.
I have managed to survive by staying ahead of the curve. Unfortunatly, the curve always dies in frustration and hate. Bioinformatics will be reduced to just another service, probably done offshore by transnationals who will charge a s mall 'on demand' fee for the requests made by service consumers. Except for the top people, the rest of the Bioinformatics community might as well get ready to train their replacement sooner rather than later.
Yes, I do plan to learn Bioinformatics at the graduate level starting this winter, but I have no illusions about a sustainable future for that field in this country (USA.) We will simply have the bottom cut out from under us. Do it for love and plan to kiss it goodbye really soon.