Friday, December 14, 2007


After the debate about what Prof. Rao said in his article a few days ago, here's a guy putting things in the right remember that software guy from Seattle? I think he succinctly conveys a balanced perspective, recognising the importance of both IT skills and the fundamental importance of a good basic science background. One more message; read books.
"A solid working knowledge of productivity software and other IT tools has become a basic foundation for success in virtually any career.

Beyond that, however, I don't think you can overemphasise the importance of having a good background in maths and science.

If you look at the most interesting things that have emerged in the last decade - whether it is cool things like portable music devices and video games or more practical things like smart phones and medical technology - they all come from the realm of science and engineering."
That's the point. When you climb high on the branches, you may forget the roots. But there would be no branches without them.



Blogger Saket said...

In the semiconductor industry we were able to forget our roots for sometime. The trouble is that now as the end of silicon integration approaches, basic sciences become more important.

Carbon nanotubes, nanowires and quantum computing are all the rage on university campuses, at least in America. There are always some tiny improvements needed though. The latest hafnium laced gates are being used in Intel's chips at 45nm. There was active research in high-k and low-k dielectrics for different purposes on chips.

On the SW side, I cannot see the same. You can forget your roots and get by. Knowing stuff from algorithms, number theory (cryptography) and graph theory can help in some new CS work. Again I am not an expert there, so I may not know.

What I am saying is that you can afford to forget how an invention came about at least for sometime. New problems crop up, that need research. Bottom lines of companies are affected plenty enough that they will invest in it themselves. Intel does...

3:16 PM  
Blogger Ashutosh said...

You make some great points Saket. Given that there is a natural limit to how much you can squeeze on a chip and how fast you can harness its power, technologies like nanotechnology and carbon nanotubes have inevitably become important. So are things DNA computing and quantum computation that have arisen from pure science.
Your observation about the SW side is interesting, but don't companies like Microsoft employ a decent number of theory and maths people? I do agree that there seems to be more focus on basic research in hardware. But as you said and I have the same thoughts, people can afford to forget about roots for some time. Then the technology starts finding its limitations and research in basics needs to be galvanized again. My concern is precisely this illusion that is being created, in the period when people can relatively afford to forget about the roots. Of course companies will invest in whatever they have to to achieve the bottom line. The question is, as you know, whether they will realise it soon enough to have time to build up capital from basic research. There's no point in suddenly realising that your technology is pressing against a wall and that you need to suddenly invest in basic research; as you know it is inherently a long-term activity that does not provide returns in a very short time. Until now, companies like IBM have shown enormous foresight in investing in basic research ahead of time. But one wonders if they will continue to do that, otherwise they are going to be in trouble.

3:26 PM  
Blogger Nikhil K said...

I too have a difficulty in believing that the innovation processes in place at companies as of today are beneficial in the long term. Take the case of Bell Labs, which were once the hotbed of discovery and produced the cosmic background radiation, fractional quantum hall effect etc. The unhindered freedom to explore which it gave to its employees was what allowed these achievements. But, after Lucent took more hands-on control, the greenfield discoveries have been whittled down. And we can expect more engineering than science to come out of Bell labs in the next decade. In the struggle between quarterly profits and a ten year plan, the former will always have an upper hand. Perhaps, this is good. A division of labor where the Universities look after the dreams and the corporations their profits.

11:45 AM  

Post a Comment

<< Home