Here’s a patent spoiler writ large …
Yesterday I learned something that I already knew but had not taken in properly. There are those out there concerned about patents on ‘life’.
Now there are rules about such thing which vary from jurisdiction to jurisdiction, but the basic principle is that one cannot patent a principle of nature.
That sounds clear enough but any linguist or any lawyer will tell you that this concept leaves a very wide door open for arguments as to what and what is not a principle of nature. And thus it is so in our courts.
We scientists have been working very hard to demystify life and in the process we are developing the tools to create and modify life. It’s as though living beings are just complex robots; which is exactly what they are to some, and not to others.
Are we just very complex robots? There are two types of people (and no animals or plants) that disagree – those that believe that we humans contain a soul and those that think we (and all living things in some cases) are ‘special’ even if we are just organic robots.
I am guessing that religious types often feel very uncomfortable with the idea of patents over say human genes, but possibly less uncomfortable with plant patents because plants don’t have souls.
Others may think that granting patents over genes will just encourage research corporations to fiddle with life and we shouldn’t do this because of the ethical and actual disasters that may follow. I have to chuckle at the concept of an ethical disaster – it would be all over the front page of the Sydney Morning Herald and not even mentioned in the Daily Telegraph.
Social justice is another motivator. Some believe that corporates automatically use and abuse knowledge and power against those individuals without means. This is a ‘fairness’ argument and people in this camp would prefer that certain genetic engineering technologies did not exist rather than these unfairly benefiting a few.
Financial dependency is another issue. Monsanto hasn’t done their position any favours by ruthlessly exploiting their technologies to the detriment of (arguably) the environment and a sense of personal and financial freedom of farmers.
In order to get a US patent four criteria must be satisfied: the invention must be useful, novel, non-obvious, and must meet what is known as the enablement criterion, meaning that it should have a detailed enough description that anyone working in the appropriate field should be able to make use of it.
In the area of genetics there are apparently four broad categories of patents filed; diagnostics, drugs that affect the functioning of genes, methods by which genes are extracted or manipulated, and finally, ‘composition of matter’ patents which are filed to protect inventions that generally stem from combining different genetic material and are typically filed for drugs and vaccines such as insulin and human growth hormone.
Putting on an inventor’s hat, in all four areas substantial R&D is required to create useful outcomes and through this lens I believe that it is reasonable to seek patent protection in order to get investment to do the R&D in the first place. Obviously there are others that think that this issue is not the most important.
In the US courts there has been all sorts of silly rulings about whether genetic material isolated in a lab has significantly different chemical make-up than when it is found inside the body and therefore if genes, once removed from the body, are a composition of nature or not. This is all Sophism in my books and if the debate is fought here the debate itself suffers from artifice and will be likely won by the party with the most legal fees or greater number of brown paper bags for political donations.
I am not completely sure where I fall on these issues. On one hand I believe in the requirement for patents to lower the bar for investment into technology and I do not believe that life is anything other than organic machinery. In any case patent rights are only for 20 years – a blink of an eye – so I can’t see the long term issue.
On the other hand, I question whether it makes sense to use a substantial fraction of our planet’s resources to artificially benefit a small minority through, for example, increased life expectancy via the use of intensely complex medical technologies.
On balance, I do not think there is a good argument to prevent patents in the field. However I wouldn’t be that unhappy if corporations couldn’t get broad patents in the area and as a result the pace of technology development was slowed substantially. This would allow the ethical and moral debates to catch up at least.
By the way, I think it’s very ironic that many who hate patents say that patents do not promote ‘innovation’ and yet on the other hand they are working very hard to stop patents over genes in order to slow down the pace of innovation in this area!
And finally, to the patent spoiler…
There are approximately 20,500 genes and 3.3 billion base-pairs in the human genome. There are databases for all known medical diseases that are continually being updated.
It would be relatively easy to write a few lines of code to concatenate the words that describe every single base-base and every combination of base-pairs with the words that describe every disease known to mankind. Indeed the same code could parse the combinations of base-pairs or genes and diseases into English-sounding sentences.
This could be done with the inclusion of phrases such as ‘diagnostic method for…’, ‘a drug for…’, ‘method for extracting…’, and ‘a combination of genetic material for…’.
Possibly billions of sentences could be automatically generated in a blink of an eye with just a few lines of code put together by a team of linguists, genetic researchers, software engineers and patent attorneys. Once published on a website they would represent prior art for every future patent in the area.
Oddly this may be an implementation of the Church Turing theory which states that a function is algorithmically computable if it is computable by a Turing machine. Which if it turns out to be true implies that we are in fact organic robots which would in turn remove some of the ethical arguments against patenting genes.
I would note that neither the computers nor the people would know which of the billions of automatically generated gene invention statements is true until they are experimentally tested. Which is an act of discovery not invention. Also any such experimentation could be automated so this doesn’t require human intervention or divine insight.
I would note that we already have combinatorial chemistry which is a stylised machine-driven approach to discovering new and useful molecules. Should the outcomes of such an approach be patentable?
More importantly if a machine can anticipate an invention then arguably it isn’t an invention. Which is a wonderful restatement of the Turing theory don’t you think? Possibly it is even better. How few believe that invention by a machine is possible?
Actually I do believe that machines can invent but generally they can only do so because of human input into their design. But I am not sure this will always be the case – and yet in an evolutionary sense computers only exist because humans once started designing and building them in the first place.
Maybe rather than limiting patents over principles of nature it would be more useful to (artificially) limit patents to inventions that could not be invented solely by machines – both experimental and computational machines. It might have the same outcome (because nature tends to follow ‘rules’ within constrained boundaries) but this new approach may be less prone to silly arguments.
Another deviation of thought; since all that we know of ‘nature’, in the biological sense on the planet earth, is just variations within a ‘bound’ complex system it could be argued that only discovery is possible within this environment, That is, invention requires the imagining of one idea from an infinite set of possibilities, and not from within a finite set of possibilities.
Back to patents – there is some requirement for prior art to be ‘enabled’ (that is, shown to have worked in practice) and yet I know for sure that 90% of, say, Google’s software patents have never been enabled. They are just ideas, diagrams and flowcharts. I would argue that in the case of the use of gene technology the approach to enabling any idea is so well-known and standardised, requiring trial and error within certain bounds, that any idea by itself is all that is needed for prior art. Of course this would be argued to death in the Courts.
Here is one definition of the US Courts’ opinion of the subject “Such possession is effected if one of ordinary skill in the art could have combined the [prior art] publication’s description of the invention with his [or her] own knowledge to make the claimed invention”. I think the automated gene invention machine could satisfy this requirement, especially if the knowledge itself to enable the invention is referenced by the invention statements.
At the very least I would imagine that this style of high-level prior art forces patentees into claims of specific implementations of technologies which then allows other parties the opportunities of ‘work-arounds’. The right result in my mind.
I wonder if the ideas of this blog entry could have been patented? A method of automatically generating inventions from within a finite set of input parameters by (a) computing every known combination of input parameters as potential inventions, and (b) automatically testing every such potential invention for practical efficacy”.
It makes you think doesn’t it?
