How Computer-Automated Inventing is Revolutionizing Law

This post is long over due, and quite frankly it is because I thought I had already posted it several months ago. As I am getting back into the swing of things and gearing up to write more about software, software patents, Parts III and IV of the History of Software Patents (see Part 1 and Part 2), I want to highly recommend The Genie in the Machine: How Computer-Automated Inventing is Revolutionizing Law and Business, written by Robert Plotkin (who not too long ago wrote Why Wishes Should be Patentable for this blog), a patent attorney located in Boston, MA, who is also an Adjunct Professor of Law at Boston University. If you are pro-software patent you need to read this book because it will likely give you some wonderful insights that you can use to help you convince non-believers, and maybe even persuade a patent examiner or two. If you are anti-software patent I would also recommend you read this book as well. Plotkin’s positions are somewhat radical in that not only does he think software should be patented, but he wishes should be patentable as well, and that is exactly what will happen as computer automated inventing becomes increasingly more realistic.

What follows is a series of Questions and Answers with Robert Plotkin.

1. Why did you write The Genie in the Machine?

First, I realized that we were in the midst of a revolution in how we invent new machines, but that few people knew this revolution was underway, and I wanted to spread the word. Second, I realized that the transition from old-fashioned manual inventing to computerized automated inventing had radical implications for patent law – the “law of invention” – and that we need to begin to reform patent law now to reflect this new way of inventing. Third, I saw that automated inventing is significantly reducing the time, cost, and risk involved in creating new products, and that both creators and consumers of such products could benefit from understanding how automated inventing works.

2. Who should read The Genie in the Machine?

Several groups of people will benefit from reading The Genie in the Machine:

  • Inventors, who will learn how to leverage artificial invention technology to boost their inventive abilities and become better inventors.
  • High-tech businesspeople, who will learn how to evolve their business strategies to take advantage of artificial invention technology’s ability to design new products more quickly, inexpensively, and reliably.
  • Consumers, who will learn how to leverage artificial invention technology to become inventors themselves.
  • Patent lawyers, judges, and legislators, who will learn how to update patent law and the administration of the patent system to reflect how artificial invention technology is changing the nature of inventing.

3. What is “artificial invention?”

Artificial invention is the use of computers to automatically design new products – to do the work that only human inventors could perform until now. Just as the assembly line automated manufacturing in the Industrial Age, artificial invention technology is automating inventing in the coming Artificial Invention Age.

4. What are some real-world examples of “artificial inventions?”

  • An antenna that is now on a NASA space mission.
  • An anticoagulant drug that is now in pre-clinical trials.
  • The “cross-bristle” feature of the original Oral-B CrossAction toothbrush.
  • An aerodynamic nosecone for the Japanese bullet train.

5. What is the “genie” in The Genie in the Machine?

The genie is a computer equipped with artificial invention software, which enables the computer to automatically transform an inventor’s abstract description of the problem he wants a new machine to perform – the “wish” – into a design for a machine that solves the problem – the “wish come true.” The Genie in the Machine documents a wide variety of technologies that behave like genies in this way.

6. What do scientists, engineers, computer programmers, and other inventors need to know about artificial invention?

Yesterday’s inventors focused their energies on becoming expert at designing the physical details of new products. Artificial invention technology is rapidly automating this task. Therefore, inventors who want to avoid being made obsolete need to shift their focus to learning how to identify the problem that a new product needs to solve, and how to describe that problem in a language that artificial invention technology can understand. They will need to become more like Henry Ford and less like an automotive engineer.

7. What do businesspeople need to know about artificial invention?

As artificial invention technology automates the process of designing new products, it will become increasingly difficult for businesses to obtain a competitive advantage merely by acquiring skill in the details of product design. Forward-thinking businesses will shift their focus to acquiring skill in:

  • using artificial invention technology to design new products;
  • identifying the needs of their customers accurately and efficiently; and
  • translating those needs into a language that artificial invention technology can understand.

8. What do patent lawyers need to know about artificial invention?

Patent law reflects the manual inventive processes that inventors have used throughout the ages. The law, therefore, is already starting to break down in the face of automated inventing. We see this as patent law continues to grapple with computer software – the earliest kind of automated invention (programmers write code which a computer translates automatically into working software). The Genie in the Machine explains how the current problems with software patents can be understood as symptoms of the law’s failure to accommodate artificial invention technology more generally, and how we can reform patent law to address all of these problems so that the law continues to promote innovation in the Artificial Invention Age.

9. What do consumers need to know about artificial invention?

Tinker with and improve their PCs, iPods, and Xboxes. The Internet now enables such “prosumers” to collaborate with each other and share their designs instantly and across the globe. Three-dimensional desktop printers that can automatically manufacture products at the touch of a button are quickly becoming as affordable as yesterday’s laser printers. Combine these developments with rapidly-improving artificial invention technology, and tomorrow’s amateur home consumer will have the R&D capability of today’s mega-corporations.

10. Are computers going to make human inventors obsolete?

Not if inventors learn how to leverage artificial invention technology to boost their inventive skills, just as yesterday’s inventors learned how to use slide-rules, calculators, and eventually CAD software to automatically perform increasingly high-level aspects of the inventive process for them. Inventors who fail to update their skills risk being replaced by software that can not only invent more quickly and inexpensively than human inventors, but which often blows past the blindspots and biases that inhibit human inventors from pursuing fruitful but unconventional design possibilities.

11. Is computer-automated inventing going to make patent law obsolete?

Not if we begin to reform patent law now. Although artificial invention technology is automating the physical design of products, we still need human inventors to create new artificial invention software and to instruct existing artificial invention software to solve today’s pressing technological problems. All of this requires significant technical skill, and as long as we need people with such skill to innovate, we will need patent law to encourage and protect the results of such innovation. The rules of patent law, however, need to be updated in light of invention automation, and The Genie in the Machine lays out a detailed roadmap for such legal reform.


Warning & Disclaimer: The pages, articles and comments on do not constitute legal advice, nor do they create any attorney-client relationship. The articles published express the personal opinion and views of the author as of the time of publication and should not be attributed to the author’s employer, clients or the sponsors of Read more.

Join the Discussion

13 comments so far.

  • [Avatar for New Here]
    New Here
    September 7, 2009 11:40 am

    Patrick Davis-
    It is work to think about that is for sure, it seems the future talked about years ago is fast moving into our lives. On the subject of computer Inventing using this or more advanced technology seems will become a greater challenge to inventors when thinking in terms of any innovation that computer Inventing technology lacks. I talk about this because in buying technology today like software, used in the creation of productive work always seems to fall short of my needs and the terms Iam thinking in. On this point, the thinking of the inventor using this computer inventing technology, seems will be limited if dependent upon the technology. An inventor with a dependence on computer Inventing technology is forced to do the work computer inventing technology cannot and could have saved time without computer Inventing technology in my opinion.

    A long way to a question, I would like to ask do you believe there is / will be a market sustainable by inventors that in my view having to change themselves to asking questions and work from there. Could I ask your opinion about would inventors want to use this computer inventing technology if such technology based on software patents is the case in view of this patent landscape today ?.

    Thanks, I enjoyed the read at the link you provided.

  • [Avatar for Patrick Davis]
    Patrick Davis
    September 7, 2009 12:31 am

    Good to know about new way of ques What do consumers need to know about artificial invention? but for more static information you can click on:
    Thanks & Regards
    Patrick Davis

  • [Avatar for Mike]
    September 2, 2009 10:38 pm

    OldTimer: The patent wouldn’t be awarded to the computer, it would be awarded to the person who first conceived (in this case saw the computer’s result after operating the computer) the invention. So, your concern doesn’t really advance the issue because, even if I stumble upon an invention out in nature, for example I find a wheel and apply for the patent on that (assuming it’s 10,000 BC), I’m allowed to be called the “inventor” even though it was just a discovery. This is analogous to the person operating the computer.

  • [Avatar for OldTimer]
    September 1, 2009 02:59 pm

    Query: If in fact a computer, and not an individual, conceived of an invention, then would the invention be ineligible for patent protection? The U.S. Patent Law awards patents only to inventors. I’m not sure the statute contemplates non-human inventors.

    In a case in which a human inputs a problem to computer and the computer spits out one or more solutions, it is not at all clear to me that the human conceived an invention. Recognizing the problem is not sufficient to confer patent rights onto the human.

  • [Avatar for broje]
    September 1, 2009 11:21 am

    Not having read the book, I get the impression that these computer programs are basically configurator tools that permit a user to specify constraints and then seeks one or more solutions that satisfy those constraints. Thus, the computer is acting as a highly skilled technician reducing the invention to practice, while the user, in specifying the set of of constraints, conceives of the invention. Having arrived at the set of constraints, the user has enabled the invention sufficiently that the computer can reduce it to practice.

    If I understand it correctly, should it be sufficient, then, to merely disclose the set of constraints in the patent application, and leave out the contribution of the computer as not needed for enablement? What about preferred embodiment?

  • [Avatar for breadcrumbs]
    September 1, 2009 05:29 am


    Is your human insight and leap in logic analogous to the discredited flash of genius? If so, you may want to revisit the 1952 change to 103(a).

  • [Avatar for New Here]
    New Here
    August 31, 2009 09:55 pm


    I was not talking about Plotkin’s book, only in that context, misinterpreting ?, I think not, because a computer in truth is not being asked a question nor giving an answer, is it ?, please, the “Genie” is in the design of the software that is it’s brain, takes formated input then parses it, makes “matches” based on that parsed input is all a matter of programming by someone, right ?. As I mentioned above but with less detail, this programming is done with knowledge of the scope of the questions, so designed within that scope knowing the questions to be asked to offer answers to them, how could it work otherwise ?…because a computer alone is not aware of the world around it and not able to learn of it alone …its only software run on hardware that software is programmed. The fit with patents, is these scopes do/will hold knowledge (ideas), most if not all patented I believe for this reason, that the people that develop “Genie” must offer their own ideas not those of others, if others ideas, that is an IP grab, and a big legal problem if those ideas that are patented ?; so patent protection for “Genie” “scopes” makes for better-safe development for everyone. I have a big problem now with owned knowledge that is taken from knowledge, too much of this is a patent danger with “Genie” or whatever they are called, because the use of a “Genie” only places new ideas in those terms, terms owned and knowledge lost because that question of owners, of that IP known as knowledge, has not been answered by credit(s) and without, how would my IP (knowledge) be credited to me without a patent is my point.

  • [Avatar for Mike]
    August 31, 2009 04:17 pm

    It would seem that the computer automated system is a method of solving problems. Having invented a method, that requires no human insight or leap in logic between any of the steps, then do all products that flow from that method become obvious?

  • [Avatar for Derry Stauffer]
    Derry Stauffer
    August 31, 2009 03:02 pm

    By the way, if you simply ask the Computer Automated Inventor Genie to “invent” a solution to a general problem (“Make me a better mousetrap”), I think you’ll get about the same result as Douglas Adam’s computer that was asked “What is the meaning of life?”… The answer was “42”, based on the computer’s boiling down the question to its most basic root…. to wit: “What is 4 times 6?” ! – Derry

  • [Avatar for Gene Quinn]
    Gene Quinn
    August 31, 2009 03:02 pm


    Sorry about losing your response. We never received it, so either the problem was on our server and the comment never got in, or it was in some server in between. I am afraid it is lost. This is exactly why I am not a fan of EFS. At the critical point of needing to rely on a computer for even simple tasks there so frequently seems to be a glitch.


  • [Avatar for Derry Stauffer]
    Derry Stauffer
    August 31, 2009 02:43 pm

    Anyone know how to retrieve text typed into the comment box but lost when submitted due to “server error”? That’s what happened to my “brilliant response” to New that I tried to enter 2 hours ago. I’ll summarize: I think the term computer automated invention is misleading. How about CAI = Computer AIDED Invention? This is really just the next evolutionary step in the progression from calculators to computer aided drafting/drawing (CAD) which then essentially became Computer Automated Drafting then Computer Aided Design (especially with 3D rendering, solid works, and 3D “printing” to make prototypes). In all of these cases the computer is really a design tool being used by the inventor (typically indirectly through a design engineer hired to help the inventor implement his/her inventive concept). In IP law, the inventor is the one who conceives the concept in enough detail to make it possible for someone of ordinary skill in the related arts to reduce it to practice (the hired designer). The designer is hired by the inventor because they know more about known design elements/methods than the inventor so they can do a better, more efficient job of implementing the inventive concept. If the designer needs to be creative beyond what the inventor suggested, then the designer may end up being a co-inventor, or an inventor of an improvement to the original invention, but the concept originator remains an inventor. CAI is not an inventor, just a design engineer/tool. As is well known: To solve a problem, one has to ask the right questions in order to define the problem in enough detail to make it solvable using otherwise known solutions/methods. The inventor has asked the right questions before anyone else thought to do that, and then “explains” the concept resulting from those questions to the CAI computer so that it can go about its task of applying what is known in order to implement the inventive concept. – Derry

  • [Avatar for Gene Quinn]
    Gene Quinn
    August 31, 2009 01:30 pm


    I think you may be misinterpreting the premise of the book. It is certainly radical to say that wishes should be patented, and there is a fear that this could mean ideas are patentable. I don’t think that is what Plotkin is saying though. With the increasing sophistication of computer automated systems the law will increasingly be confronted with the situation where an “inventor” came up with an idea and nothing more. Then computer automated tools were used to bring the invention into being. So who is the inventor, the person with the idea, the computer automated tool or the person who created the computer automated tool? I don’t think Plotkin is suggesting that disembodied ideas that are not enabled and do not satisfy the written description requirement be granted patents.


  • [Avatar for New Here]
    New Here
    August 31, 2009 10:42 am

    Computer-Automated Inventing ?, the “Genie” is in the collective ideas in a given scope of development. There is no Genie other then to those that would like such an idea, because it requires little or no knowledge from them, seeing the older way to do the same development. So the developer using a “Genie”, only has a wish, not having the knowledge to see it through on one’s own ideas, only to learn how the Genie works and point it the the direction of the wish. I have had ideas that I wished to patent, years ago, believed at the time I’d created something someone would want, however, when I think about the answer to the question that the base of a given idea I have had, many others support it in the form of knowledge. If any idea I have had, the scope of that idea being those in support, what have I created that I can claim ?, a wish ?. The patent is not the problem nor the danger, only in the “wrong” hands, such scopes of developemnt to become more Genies by patent in the future that heads to the question: Have all owers of all ideas used as base in scopes of developemnt within the Genies, less the language tricks in patent claims, been given credit ?. The problem is the PTO for years now is a race, cross the line first and you win …sad.