Marc Andreessen wrote an article at the Wall Street Journal the other day, and I was impressed by it. I mean, not exactly by the article, I was impressed by the reactions to it. Lots of people got impressed by the article, and I simply can’t understand that. What exactly is good and new and insightful about what he wrote?
The growth of the importance of software in our lives, in the economy and society, shouldn’t be a secret to anybody nowadays. I would say at least ever since Cybernetics risen in the mid 20th century, no intellectual should be caught “discovering” that now it’s not all just about material goods, but “software” can also be of equal or larger importance.
The revolutions brought by modern electronics are found in telecommunications, control, data storage and all things computational. That happened in the 20th century already. Software got big 50 years ago, and still is. What has Marc “discovered” after all?
His article made me think if Eben Moglen. He has some quite interesting speeches about software and society, and one thing he never tires of saying is that software is to the 21st century what steel was to the 20th. I like it when he talks that, specially beause after that he talks some other interesting things, brings some interesting insights and draws some interesting conclusions. Here is one speech of his at FOSDEM 2011 that does start with that and end up somewhere else, forward.
I think if you want to hear a great speech about how software is “where it is” now, that is one good recommendation. But Marc’s article… What was that? My theory is that it was just a kind of a lobby for investors to keep betting on IT companies. There is probably some buzz going around in the investors world that a new dot-com bubble is about to burst and that investing in “software” is dangerous right now, and Marc wrote that to try to counter that, and restore the morale of software investors. It was published in the Wall Street Journal, by the way, that certainly supports this theory a bit.
One thing that I was very angry with the article is about how the lumps so many different problems under the word “software”. I won’t say that is not a very frequent and annoying habit that also plagues other writers, maybe even Eben himself… But in that specific article this was very prominent to me.
I think at least three things should always be considered separately, even if you want to talk about how the “software” revolution has had a big impact on these, they are very separate issues. They are the issues of communication — common people to people communications, then what we call sometimes “digital content” — mostly video, music and text stored in digital format and easily accessible, and sometimes easily copied, and then the third separate thing, that is the actual “software”, computer programs that are created to perform certain tasks, and are used as tools for work, etc.
The lines between these can get blurry sometimes, it’s true. It’s hard to say for example, if a computer game is “software” or “digital content”. A blog can start as a small communication platform between few individuals, to become something as relevant as a big newspaper (if you can pay the web hosting bill). And yes, the “software revolution”, or more precisely, modern electronics is at the basis of all that. But what do you really win by looking at all this from this angle, that it’s all “software”?
All you win is a nickname to tell people “hey, if you are not investing in ‘software’ you are not cool”… If you are really interested in society and economy, and all the transformations we are going through, and what we might like to do to drive society to some other place, and what we might like to do to avoid some other places, if you are concerned about that, that article doesn’t tell much, really.
Now how central really is “software” to contemporary society?… As much as I think it’s cool to say software is the 21st century steel, I still see a lot a steel out there. I’m looking right now through a window to tons of steel. Really, tons. Lots of cars moving around. Consuming gasoline. Some of them are trucks carrying coffee beans imported from Brazil to feed spoiled computer programmers in the Silicon Valley.
Energy, tool manufacturing, real state, transportation… Matter still matters. I was very glad to see Chris Anderson say on twitter yesterday that “Real stuff is 10x more important than software”. I even asked him about Marc’s article afterwards.
A couple of days after that article was published, company Solyndra declared bankrupcy. The company was going to manufacture photo-voltaic cells that carried lots of design innovations (i.e. “software”). But it seems competition was too hard, and the price of their software didn’t make it worth compared to other manufacturer’s cheaper “hardware”. I even heard that the whole problem was a fluctuation on the price of silicon… Yes the price came up, then it looked like their economical design would be a killer. Then it got back down (i.e. yes, then sand got cheap again) and they just couldn’t hit the right spot in the cost-benefit curve again. So billions were lost, and hundreds of workers were laid off. Many of them skilled workers, who can create and operate complicated software.
I think this kind of story is nice to show how things are complicated. It’s great to read about macroeconomics, but you shouldn’t miss the trees for the forest.
The whole software vs. hardware thing does not fall far from the old services vs. industry and agriculture phenomenon that economists have been studying for a long time. We must bring that all together. Avoid treating technology as magic.
We must discuss things like licensing of “digital” music. Free software. The question of owning software, and software as a service… This is where we should be. That’s where I am at least. And Marc pretty much ignored all these issues in his article. I guess Wall Street is not ready for all that yet, and is simply treating software as the new steel, with no recognition of how they are actually two very different beasts, and not just labels on the jars where you are choosing to put your money in.
Anyway, “soft” stuff like design, algorithms and control systems have been around for a long time, and so has “hard” stuff like power, energy, steel, rare earths, plastic etc. They coexist, and are all parts of big puzzles. There is no simple answer. We won’t simply stop selling steel “cold turkey” the day that someone come up with some excellent new functional language that make creating more and better software easier. These are different dimensions. Our lives sprawl through multiple dimensions. And the curse of dimensionality is upon us. From time to time someone will point into a direction and tell you to project the whole universe there. Sometimes that helps for some reason, for some time. But the real big and complex world is still out there with its fractal inextricability.
Bottom line: software is great, but people are still dying of hunger out there, and there is no software that will _solve_ that, because it’s not just about finding better logistics, etc, it’s about making things happen. What do you say of THAT??…
Only robots might be able to do something… Steel robots!