Yes, its dying in its current form.
When you think about it, the 'personal computer' is an incredibly inefficient device. The fundamental flaws of unreliability, (due to local OS, local software and local storage) and expense, (due to the need for hardware sufficiently powered to drive local resources) are being addressed now, and will drive change in how we manage devices and information management/storage.
Cloud computing and high speed broadband will drive a shift to thin clients. Thin clients will drop in price to the point where they are a disposable item - much like a low end bubble-jet printer, or a monitor is now - when it dies it will be cheaper to throw out/replace than to repair. this will start a demise of 'box-level' repair work. (this is already beginning to happen with laptops and the increasing customer mantra of "I could buy a new one for that". Which is truer now than it was 12 months ago and will be truer again in 12 months time. This trend will not diminish over time.
Once widespread thin-client computing is viable, virus removals will be a thing of the past, (no local OS). Data recovery on a domestic level also will diminish as everything will connect to the cloud therefore no need for 'internal' and 'external' storage repairs. There goes the bulk of our work as we currently know it ('box' repairs, virus removal, data recovery).
I agree with the post above. We need to get diversified as this thing of people bringing us their sick boxes is impendingly time-limited. 'professional IT' will increasingly be a corporate and infrastructural/organizational arena. Work for individuals will dissipate over time due to a transference of the work of computers away from local device level, and the consequential shift in device-affordability / disposability.
This debate has been had on here before (can anyone throw the OP a link? - I cant remember the thread titles).