How Software is Harming Science and Engineering

A recent column by Netscape co-founder, software entrepreneur, and noted Silicon Valley venture capitalist Marc Andreesen caused a stir in the tech community. Andreesen postulated that the software industry was “eating the world” and “poised to take over broad swathes of the economy.” This is delusional.

It’s a clear case of someone with a hammer—Andreesen developed software, ran software companies, and now invests in software companies—seeing everything as a nail. The irony is, software is hardly a hotbed of innovation.

While technology races ahead in many other fields, software has advanced but meagerly in the past 20 years. In terms of solving grand challenges, software has largely failed to deliver. Take the case of voice recognition. It’s much better than it was in areas like airlines’ reservation phone trees. But despite billions of research dollars no company has produced commercially available, affordable voice recognition software that can understand and transcribe, from voice to text, conversations involving multiple voices. Likewise, voice recognition software requires training to work well—it’s not speaker independent. Yes, an IBM team did take on live Jeopardy! champions and beat them but the Herculean effort required to program a supercomputer to accomplish this just illustrates the enormous chasm that continues to exist between software and the solution of truly great challenges.

Compare this to advances in fields like DNA profiling and decoding. Over the course of a mere two decades, the ability to sequence or perform tests on DNA has become orders of magnitude cheaper—even to the point that sub-$100 DNA testing services will likely emerge within the next three years. Or how about the field of 3D printers, a mind-bending class of devices that fabricate 3-dimensional objects and even devices with moving parts. It can do this in a matter of minutes by layering precise patterns of materials painstakingly and accurately, with the help of software and smart computers (note: software plays a supporting role here). In the race to innovate and serve the developing world, companies like General Electric are developing medical imaging technologies that cost 1/10th or 1/20th the price of comparable devices sold in the U.S.

Yes, software has made some limited progress in key areas. Search engines have had a material impact on the world. Some types of enterprise software have made a huge difference in business efficiency. But I’m hard-pressed to think of any other software-based product that has enabled revolutionary changes in society due to the innovative nature of the product and not to the innovative way people use the product. And, of course, Andreesen does give a nod to the other enablers of the growth of software such as cheap Internet-ready devices, the global telecommunications grid, and the microprocessor.

Andreesen’s portfolio of companies includes many that are highly touted but thoroughly unoriginal. Twitter is, basically, another way to do SMS using the Internet. Facebook is Friendster 3.0 hacked up by some kids in a door room that has enjoyed good timing and deployed excellent UI. And then there’s Groupon, an enterprise that has achieved a single feat of innovation—creating dubious new accounting terminology to justify inflated IPO valuations.

In fact, I’ll make a bold statement: I believe that software is draining talent needed in other areas of science and engineering. Smart kids in college major in computer science rather than mechanical engineering because that’s where the money is. Yet some smart kid coding social games for Zynga serves very little societal purpose—particularly when that same kid could have instead decided to build innovative low-cost drip irrigation systems to serve famers in the developing world where irregular irrigation, dwindling water supplies, and poor infrastructure are a crushing trifecta.

The last thing we need is a world consumed by software, Marc. Please invest in more startups that seek to change the world in a meaningful way, and not just to make a mint in social media and useless software companies.

Vivek Wadhwa is a visiting scholar at University of California-Berkeley, senior research associate at Harvard Law School, and director of research at the Center for Entrepreneurship and Research Commercialization at Duke University. You can follow him on Twitter (@wadhwa) and find his research at wadhwa.com.