Wednesday, February 3, 2010

IT, Biotech and Patents: When Worlds Collide

Some of you know I like to "throw the long bomb", in that I think about what could happen at the limit, or the end-game, when I prognosticate. This is a handicap on Wall Street, where being too early is exactly the same as being wrong. And it's either mind-expanding or migraine-inducing, or both, when thinking about how our Universe will wind up (cue Peggy Lee singing "Is That All There Is?").

But recently I've been spending time around non-IT folks, specifically, people involved in start-up life sciences and biotechnology ventures. Their perspectives on intellectual property issues are very different than my experience in IT. Thinking about the current state of these two different worlds has caused me to imagine what might happen to both if they proceed at their current course and speed.

DISCLAIMER: I am not a lawyer, nor do I play one on TV. These opinions are my own and if you get harmed by taking them as gospel, then shame on you.

For software, the patent system has become so unwieldy as to be useless. Many have opined on this. It is currently almost impossible to write a new line of code that doesn't infringe on some software patent somewhere. I don't think that's what Jefferson intended.

Patents were written into the Constitution to promote the development of new ideas by providing a government-granted monopoly for a fixed period, after which the ideas would pass into the public domain. We as a country benefited from the advancement of knowledge, and inventors benefited from the fruit of their labors. Life was good.

But life then was also slow. The pace of technology development was slow enough that a 17-year patent lifetime made sense. Moreover, the target environment of a technology improvement was also slow to change. Prior to the PTO's allowance of software patents, the target environment was mostly the physical world (we are excepting design patents from this discussion). Our physical world changes slowly enough that 17 (or now 20, to sync up with the rest of the world) patent lifetimes made sense. For example, a drug providing a certain therapeutic benefit in humans is likely to provide that same benefit 20 years hence - evolution just doesn't work that fast.

Things are really different in the world of software. There is no physical world target, just a physical embodiment of an abstract computing machine. Any software technology makes assumptions about the state of development of all of the layers of abstraction all the way down to the hardware abstraction layer. Each of these layers is not subject to any natural law governing its rate of change. Depending on market conditions, each layer is constantly evolving at different, yet increasingly rapid rates.

Ultimately, hardware needs to be built to instantiate the lowest layer of abstraction. This hardware is subject to physical laws. Back in the day, the pace of electronics development was leisurely enough that 17 years was a generation. We can argue about the actual pace of electronics development today, but I don't expect anyone to claim that the pace is slower than it was when ENIAC roamed the earth.

So, with electronics and software technology driving sub-two year product cycles and 3-5 year total technology refreshes, how are new ideas promoted by granting inventors 20 years of monopoly on novel, useful and non-obvious inventions with a 3-5 year utility? It's plain that patent owners benefit, but it can be argued that the rate of innovation might actually slow as the effort and resources required to navigate the patent minefield overwhelm the effort and resources required for invention.

Intellectual property lawyers have added to the problem. Since discovering this fertile yet unplowed ground, an army of lawyers has built careers around patent law. Smart litigators have made large sums in the courts, which has attracted more litigators and has also driven corporations to get better at playing defense. But the resulting friction loss for individual companies and industries is considerable, measured in both cash and time. I built a successful, innovative software company in the 80's without filing a single software patent. This was partially out of my belief that they were wrong, but, more importantly, my corporate counsel did not think that they would increase the chances of successfully executing our corporate strategy. Today, I would be advised that a multi-million dollar IP program might not be sufficient. Is that really progress?

Many others are asking questions like the ones posed above, and I suspect that we are on a slow trajectory toward changes in patent law that sharply restrict the number and kind of software patents granted. At least I hope so.

Regardless of the patent situation, in my IT world the game was always about getting to revenue and profitability in a 3-5 year timeframe and an expense of $5M to $10M, plus or minus. Today, web startups can get farther on less, for reasons that could be the subject of another discussion.

As I am learning, the world of life sciences and biotech is way different. Although revenue and profitability are still goals, the expense of research and development is generally much higher and there is the added fun of government approvals and clinical trials. It is not unusual to spend $500M and take 10 years to get a drug to market. The good news is that successful drugs can generate billions of dollars at fat margins. If you think of the market as a casino,IT startups are the $5 blackjack table and this stuff is baccarat.

In this world, patents matter a lot more than in IT. They represent an option on success, which can have substantial value before a company delivers dollar one of revenue. It is the rare IT startup that gets bought for big bucks while its product is still in development, but it is common for drug startups to be bought before Stage III trials are complete. The selling company bases a substantial part of its value on its IP portfolio, and the buyer is willing to pay for monopoly power in a market segment for a period long enough to generate substantial returns.

When patents in the life sciences area are for things like medical devices or molecules, they fit snugly into the patent framework established hundreds of years ago. Unlike IT, medical devices or drugs have a relatively stationary target, i.e. animals or humans (there are some interesting exceptions, like HIV/AIDS and fast-mutating bacteria, but hold that thought for a paragraph). So a long patent life allowing recovery of enormous development expense seems to make more sense in this world than in IT.

But when I hear people in this world talk about things like genomics, custom sequenced DNA, designer drugs and synthetic biology, I hear "software". The ultimate instantiation might be a molecule or cell, but, to manage complexity, layers of abstraction are being developed and ideas are being captured and manipulated at these higher levels. This is analogous to a Java program ultimately executing on an iPhone. Only the hardware is different.

If patent law changes to reflect 50 years of development in IT result in shorter-lived or substantially diminished monopolies for software, and life sciences technologies start to look more and more like software, then it's going to be hard to argue that they should be treated differently, unless patent law further changes to explicitly consider development costs, which could result in different lifetimes for patents in different fields.

Given the glacial nature of changes to the patent system, it is not unreasonable to consider this may be the "golden age" of IP law for life sciences and biotech. There could be substantial turbulence ahead.

1 comment:

  1. The wierdness for me is the idea of patenting a life form..or, say,a gene. I can understand reserving a method of making a gene, but not patenting the gene itself. Does the history of Quaker Oats provide any guide? Lou, thank you for blogging. Jack Vaughan