@TL
I have been reading your posts wherein you have mentioned about "product" innovation in IT companies. But then what is a "product" for an IT company. if you feel google, MS, Yahoo are product companies and Infosys, TCS, Wipro is not, then i must refute.
When you compare Infosys and MS don't compare MS Office and Finnacle, compare Visual Studio, .net. When you talk about Google, talk about Android OS, the versatile little OS that is running more and more hardware these days. When we say innovation of companies like Sun (Now Oracle), we talk about the Java Virtual Machine which can give you a Java Runtime environment on any piece of hardware.
You also mentioned china has billion of $$ product companies, and i take it in IT as the thread is about IT, i am sorry you are wrong. in IT field, by any standard, china is way behind india. Perhaps you are using the word product for industrial product. in that case china is way ahead of us, but then this is the difference in our growth models. Industrial base vs service industry.
The Chinese are way ahead of us, in computer science. They are working closely with American companies and now actually producing all their hardware. Initially it was just Taiwan and now its spread all over China.
Coming to the topic i feel the innovation should be in riding up the value chain. and in the last 5-6 years there have been signs that it has also started to happen. i do agree we took a lot of time in the transformation, but now it has started, it will not take long to happen because there has been increase in the competition at the lower end service level from many countries, and in order to maintain the competency and lead, the indian IT companies will have to move up the value chain. its the most logical process.
The idea in facilitating innovation is to push research in universities. This is something Pakistan is doing poorly. IT is taught in every university, but there is no research being done on it. Students need good research equipment, funds, labs and mentors. Can't say much about India, but in Pakistan research students are sort of duking it out on their own with minimal university assistance. Raw talent brings them up most of time, luckily. However thats not a good plan.
Also good computer science serves other sciences. You want to use computers to research genetics, neuroscience, quantum mechanics, cryogenics, protoscience - you know the stuff that's really out there.
Again I'll comment only about Pakistan, but it seems like we have a limited number of careers:
1) IT 2) Business 3) Doctor 4) Engineer (electrical, aerospace(few), architects)
We need to drag in more sciences. To reiterate, we're serving somebody else's needs in a way that everybody knows it can be done or worse it has been done before but we're just delivering a product thats more value for money.