The Software Patent Problem: Not Emphasizing the Technological Contribution of the Innovation

Mark Lemley

On August 28, 2014, I had the opportunity to speak with Professor Mark Lemley on the record. Lemley and I share the opinion that Alice v. CLS Bank represents a significant change in the law relevant to software patents. To my surprise this truth is not understood or  appreciated by many in the patent community.

If you review the Alice form paragraph rejections from the USPTO, the reality that the USPTO is withdrawing Notices of Allowance and issuing Alice rejections, the latest decisions from the Patent Trial and Appeal Board, and a handful of post-Alice Federal Circuit cases the landscape is extraordinarily adverse to software patent applicants. Still there are those who protest and say that little or nothing has changed despite the objective reality facing applicants and patent owners.

Against this backdrop of disbelief I thought it would be useful to walk through the analysis with Professor Lemley. What follows is the final segment of our interview. To start reading from the beginning please see The Ramifications of Alice: A Conversation with Professor Mark Lemley. For our continuing coverage of the Alice decision please see our Alice archive.

[Interviews]

QUINN: Yes, I think so, too. What part do you suppose the fact that these judges largely did not grow up during the era of the Internet plays in their decision making? The reason I throw that out there is because it seems like there is just this completely false idea that coding software is a trivial process. And it’s not. Even Justice Kennedy during the oral argument said that “any second year engineering student could code this up.” Well, I don’t know about you but practically every time I open my computer I see a critical warning that tells me that I have to download this or my system’s will explode. Shouldn’t that nearly universal experience be a clue to somebody that maybe this isn’t as trivially easy as the judges think it is?

LEMLEY: I think that’s right. And part of the problem is you’ve got some old language, there’s old Federal Circuit language that let people get away with very little disclosure in the spec by saying that writing code was “necessarily a mere clerical function to a skilled programmer.” And that strikes me as crazy. But the Court seems to have come out and had that in its mind. It didn’t help that the lawyer for Alice was asked by the Supreme Court, well, couldn’t somebody code this in a weekend? And he said yes. So it may be that this was a particularly bad case for the patentability of software generally. And I think that’s been another lesson in the 101 cases. Part of the Supreme Court’s view and the outcomes is they just view the 101 issue differently than the Federal Circuit does but part of it is they’ve gotten some pretty lousy claims up there to review, right? And so they look at those claims and say well this sort of thing shouldn’t be patentable. But their opinions are not limited to this sort of thing. And it’s hard to draw a distinction without putting in front of them a case that really does look like it’s got real computer technology in it.

[JV-3]

QUINN: I think up until— I would have totally agreed with what you just said up until this case because I think that the claims in this case were good. This was by and large the way that you construct software claims. I mean I was reading some of IBM’s patents that deal with Watson and they’re really constructed the same way. I think the construct is a good construct whether or not the underlying technology comes through is a different question. That’s again problematic because I think you could read the Watson claims, a lot of them at least, and you get done and you’re like well what really is new here? But then at the end you have what is essentially a Star Trek level computer. The claims were no doubt written that way because that was how the industry was told to write claims to get them to pass muster, so the rules of the game have changed mid-stream it seems. And I don’t know what the answer to this is other than through all these changes in the law, additional layers of disclosure need to go into the spec in a way that doesn’t come back to haunt you. You just cannot hide the innovation anymore both in the claims and in the spec.

LEMLEY: Yeah. People have been writing claims that don’t emphasize the technological contribution of the innovation. And I think that’s part of the problem. And I think if we can write claims that really highlight the technological contribution then the Court maybe is going to be inclined to view those differently. And more favorably.

QUINN: I hope so. I mean basically right now what I tell people with software it’s not pretty, it’s not the end of the world moving forward, but it’s gotten a lot more expensive because we have to do a lot more in the spec to have any hope. And the reality is that the law is likely to change a couple times before any patent applications that you would file today would even get examined. But if I’m correct, it will be too late to apply when the law changes to whatever the next test is, or whatever the more developed version of this test will be. So you have to take your best shot now, give a solid technical disclosure and hope for the best.

LEMLEY: And that may well be right. Although I’m hoping that the other thing Alice has done, of course it’s a little bit too soon to tell, but in the conference in Seattle Judge Rader made reference to the Federal Circuit having staged a bit of a rebellion against Mayo. And I think the days of that rebellion are over. Judge Rader is no longer on the court. You’ve clearly got a majority of Federal Circuit judges who applied a more restrictive test. And I think the cases you’re seeing after Alice suggest to me that the Federal Circuit doesn’t have any continuing appetite to try to fight about this anymore. There have been four patentable subject matter cases go to the Supreme Court in the last four years and they’re not going to fight this battle any more. So if that’s right, while we’re certainly going to see a bunch of cases trying to draw this line, I don’t think that patent practitioners can hope and say well I’m just going to wait it out and the Federal Circuit will come back and make this all broad again. I’m just not sure that’s going to happen.

QUINN: I don’t think it’s going to happen either. But at the end of the day the reality is the Supreme Court’s just not following the law because these different sections of the statute are concerned with different concepts. They seem like they’re interested in just shoving it all into 101 and it’s disconcerting. But what the Supreme Court doesn’t seem to be wanting to take away with 101 the Federal Circuit seems to take away with 103. That’s probably another conversation for another day though.

One of the things I definitely wanted to ask you about is where do you see this going in Congress if anywhere? And the reason I bring that up is I’ve started to raise this as an issue, and I’m starting to hear others talk about it. With what’s going on both in the biotech world and in the software world with respect to patent eligibility, without even getting into all the 103 stuff that’s going on right now, patents are being devalued at an enormous pace. Many software patents are simply no longer worth the paper that they’re written on. And a lot of patents in the biotech space are the same way thanks to Myriad and Mayo. At what point do these companies have to notify their shareholders that maybe billions of dollars need to be written down because their patents aren’t worth anything? I know this is a Sarbanes-Oxley issue. I don’t think there’s a right of private action under Sarbanes Oxley, but do you suppose at any point in time that will cause CEOs to clue in to what’s going on here and that might force Congress to take a look?

LEMLEY: That’s an interesting question. I think part of the problem, frankly, is we’ve had, especially in the software world, we’ve had a lot of over claiming. So if you look at the history of the patent system you see some pendulum swings in different directions, right? And we go from really strong protection and maybe too strong, all the way back to really weak protection that’s too weak. We’re definitely on the backward swing of the pendulum away from strong protection right now. And one of the things that I’ve always kind of tried to push for is what you really want is actually to have the pendulum stop swinging so far in either direction. You’d like to keep it somewhere near the middle. And the problem is I think that Section 101 is just a really coarse filter to try to distinguish the good from the bad in software or anything else. And so the kind of impact of too many people asserting some dubious patents or patents that they’ve over claimed maybe the backlash is bad and a bunch of legitimate patents are gonna get swept into the invalidation. John Allison, Dave Schwartz and I have a comprehensive new study of patent litigation outcomes that we’ve just completed. We looked at the industry and technology differences in litigation outcomes and some of it’s what you’d expect, but some of it is fascinating. Pharmaceutical patents do pretty well. Pharmaceutical patent owners win more than half of the time. Software patent owners do pretty poorly. They win only about 13% of the time. But the thing that really surprised us, and I think is kind of alarming is the single industry that did worst in patent litigation was biotechnology. Biotechnology patent owners won 8% of the time that a case went to final judgment.

QUINN: Wow.

LEMLEY: For an industry like biotech, which I think really does depend pretty heavily on its patent portfolio, that’s an alarming statistic.

QUINN: Yeah, that’s very alarming because as you know most of those companies need investors for 10 or 12 years if they’re going to have any hope of ever succeeding.

LEMLEY: Right.

QUINN: And investors like patents, particularly in that space, because they’re investing so much money. If you don’t have a competitive advantage how are you ever going to grow as a small biotech company?

LEMLEY: Yes. I think that’s right.   We need to do something about the problem of nuisance value litigation and over claiming patents, which I think has been a real problem, but we also need to make sure that people who have real inventions can get effective protection for them. And the difficulty with something like 101 is it’s really hard to look at those and distinguish them and say okay, only the good patents are going to survive this filter and the bad patents are going to be weeded out.

QUINN: You know we’re entering I think a dark time for patents and I don’t know that anybody really knows what’s going to happen on the other end. I just know that patents are important to an awful lot of different companies for an awful lot of different reasons. And the decisions that we’ve made repeatedly throughout our history is we want to have a strong patent system and right now it doesn’t seem to me like we’re heading in that direction. And this is at a time when the rest of the world largely is adopting stronger patent rules and if you look at like Bayh-Dole, for instance, dozens of countries are trying to adopt a system like we have. It just seems like we’re going backwards at the exact wrong time.

LEMLEY: Well, the question to me is how far you go backwards, right? I do think we had some excesses in the last 15 years or so and the patent litigation explosion by patent trolls has been part of that. I think that’s in large part what the Supreme Court has been responding to. The key is to not go too far in the other direction.

QUINN: I agree. In this instance I just don’t see an awful lot of judicial restraint. We study judicial restraint probably our first day of law school in many different areas so it doesn’t seem fair. You practically beg for answers. Please answer this question, you know what the important question is we’re asking, don’t dodge it, give us an answer. But now it seems like the Courts today are answering these questions about litigants that aren’t really in front of them based on these generalized concerns. I’m not going to deny that there’s been excesses in the patent litigation arena, there is certainly has been and I think there is real, genuine evil in the patent troll market with fraudulent demand letters, which is just nonsense.

LEMLEY: Right.

QUINN: But by the same token I would be so much more comfortable handling these cases under 102, 103, and 112 because like Judge Newman says, if you handle technologies under 101 you’re potentially forestalling an entire avenue of innovation and preventing an entire industry from developing.

LEMLEY: I think that’s right. The problem is 101 is an attractive tool to people who are confronting those nuisance suits because it’s early and it’s easy, right? And it doesn’t require going through a lot of the facts, discovery, and work that can cost a lot of money if there were better mechanisms to try to get at those weak suits earlier and more cheaply you know then people could use those mechanisms. But the problem is that I’ve got somebody who shows up and says it’s going to cost you $3 million to get the summary judgment on your obviousness claim and probably courts aren’t going to grant summary judgment and they’ll send it to trial anyway. So why don’t you pay me a million dollars to go away. Everybody’s decision is to pay million dollars.

QUINN: Yeah. It is a mess all around.

LEMLEY: Yes.

QUINN: But I do think I would like to try requiring those people who are bringing a patent infringement lawsuit to file it with a much more specifically pleaded complaint and one that probably also includes detailed claim charts. Because if you’re a serious litigant you’ve got that stuff before you file anyway.

LEMLEY: Yep.

QUINN: All right. Well, I really appreciate you taking the time to chat with me. And it’ll be interesting to see where we go from here. So thanks a lot, Mark.

LEMLEY: Okay. Take care.

 

Share

Warning & Disclaimer: The pages, articles and comments on IPWatchdog.com 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 IPWatchdog.com.

Join the Discussion

17 comments so far.

  • [Avatar for anonymo]
    anonymo
    January 26, 2015 12:10 pm

    I’m months late to the party on this, but one of the amicus briefs argued that the claims at issue in Alice could be implemented in seven lines of code, and put some actual code in the brief. See pp. 5-8 of the brief linked below.

    http://www.americanbar.org/content/dam/aba/publications/supreme_court_preview/briefs-v3/13-298_resp_amcu_pk-ada.authcheckdam.pdf

    I’m far enough removed from my CS classes that I can’t tell if the argument’s any good. It seems to underscore Prof. Lemley’s point: “It didn’t help that the lawyer for Alice was asked by the Supreme Court, well, couldn’t somebody code this in a weekend? And he said yes. So it may be that this was a particularly bad case for the patentability of software generally. And I think that’s been another lesson in the 101 cases. Part of the Supreme Court’s view and the outcomes is they just view the 101 issue differently than the Federal Circuit does but part of it is they’ve gotten some pretty lousy claims up there to review, right? And so they look at those claims and say well this sort of thing shouldn’t be patentable. But their opinions are not limited to this sort of thing. And it’s hard to draw a distinction without putting in front of them a case that really does look like it’s got real computer technology in it.”

    It’s unclear whether the Public Knowledge amicus brief motivated Justice Kennedy’s question at argument or not. I’m too far removed from my CS classes to evaluate Public Knowledge’s argument. But if it’s no good, Alice’s lawyer had a chance to say so at argument. He didn’t, which suggests that it’s a good argument.

  • [Avatar for Gene Quinn]
    Gene Quinn
    September 12, 2014 11:51 am

    Fish Sticks-

    I think the over claiming problem is a 102, 103 and/or 112 issue, depending upon the context. Sometimes you see claims where you wonder how anyone could have thought they would pass 102 muster. Then other times the claims seem quite clearly obvious, and still other times the claims go well beyond what is fairly described in the specification.

    I personally have no problem writing claims in the first instance that I believe will pass 102 and have 112 support. Then add from there and let’s have the obviousness discussion during prosecution. This strategy really only works, in my opinion, if you are going to submit a claim set with varying claim scope. If you write to get past 102 and then every dependent claim depends on claim 1 I don’t think that is a wise strategy.

    -Gene

  • [Avatar for Fish Sticks]
    Fish Sticks
    September 10, 2014 04:21 pm

    What is “over claiming,” Gene?

    103? 112?

  • [Avatar for THM]
    THM
    September 10, 2014 12:32 pm

    For those thinking a “technological contribution” is sufficient to guarantee patentability or, in the alternative, that it is the contribution itself which determines patentability should look to the north where a contribution-approach adopted by the Canadian IP office was smacked around in a muddy field full of glass shards by the Federal Court in respect of the Amazon one-click patent. On appeal, a much saner approach was adopted by the FCA, but the system still struggles with the handling of software patents & business methods. Now there is (generally & precariously) a distinction between a computer problem (one that solves a problem with a computer) – patentable – and a problem merely addressed by the use of a computer for convenience – non-patentable.

    It’s interesting to watch how the two systems react. If we ever truly take on an SPLT, there is sure to be trouble.

  • [Avatar for step back]
    step back
    September 9, 2014 01:01 pm

    Moreover, there was a benefit to suppressing comment disclosure on this blog site. With less talk about patents and patent law, “innovation” was no longer stifled and all those people who wasted their time here were “unleashed” to innovate away in unstifled volume and quality. Needless to say, history has proven the pundits right. Just take away the pesky patent seekers and a cornucopia of innovation magically appears by sheer market forces,

    Let us count the many ways it has happened here while IPWatchDawg was down:

    1. ___________
    2. ___________
    3. …..
    Anybody?
    Bueller? Bueller?
    Anybody?

  • [Avatar for step back]
    step back
    September 9, 2014 12:55 pm

    You know Gene, you could have asked Professor Lemley to help you with fixing the problem. After all, how hard can debugging a mere abstraction be? Just say the magic words: Bug Begone and wallah, problem solved.

  • [Avatar for Gene Quinn]
    Gene Quinn
    September 9, 2014 12:23 pm

    Everyone-

    FYI. I believe I have rectified the comment problem and see why all comments were being sent to spam. Comments should be appearing immediately except in rare cases where they are being held due to the spam filter capturing them.

    -Gene

  • [Avatar for Rick G.]
    Rick G.
    September 8, 2014 05:57 pm

    Lemley : “the single industry that did worst in patent litigation was biotechnology. Biotechnology patent owners won 8% of the time that a case went to final judgment.”

    I don’t think this says much about the value or validity of “biotech patents” generally. It might say something about the types of biotech patents that are infringed and taken to final judgment in court rather than reaching settlement, or it might say something about the interests of the entities on either side of the litigation.

  • [Avatar for Fish Sticks]
    Fish Sticks
    September 8, 2014 04:01 pm

    Gene, can you give us some examples of claims that “emphasize the technological contribution of the innovation” Thanks

  • [Avatar for Anonymous]
    Anonymous
    September 8, 2014 12:18 pm

    Paul wrote:

    “If there are cracks in the structure of fact before the Court, then the law will not be adequate to paper those cracks over and the decision will be reversed. I express no concluded view as to the outcome, merely identify a factor which in my opinion will have a powerful influence on the outcome. And identifying that question does not require legal knowledge, merely an understanding of how judges’ minds usually work. That changes less than you might think in travelling from London to Washington.”

    The problem is that 101 is being increasingly applied right out of the gate in response to a 12(b)(6) motion – note that Alice was a 12(b)(6). Rader was trying to push back against that trend with the Ultramercial decision, but that decision is going down too. There simply are no “facts” that can save most of these patents and limited procedural mechanisms to get whatever “facts” you may want to push before the Court.

  • [Avatar for EG]
    EG
    September 8, 2014 10:02 am

    “People have been writing claims that don’t emphasize the technological contribution of the innovation.”

    Gene,

    Sorry, but that statement by Lemley is a non-sequitur. The claims are supposed to define what the invention is, not what its “technological contribution to innovation” is. That’s judged by measuring the claimed invention relative to the prior art. Also the “technological contribution to innovation” of the claimed invention determines its patentability under 35 USC 103, not whether it’s patent-eligible subject matter under 35 USC 101. Lemley’s view is what the Royal Nine adopted in Alice, so it’s no wonder the Royal Nine have thoroughly screwed up patent-eligibility under 35 USC 101.

  • [Avatar for Ron Hilton]
    Ron Hilton
    September 8, 2014 09:38 am

    At the risk of sounding like a broken record, the real problem in my opinion is the historic low quality of software patents and software patent examination by the USPTO, with computer science professionals having been excluded from the patent bar in general. Even though subject matter eligibility is not the right tool to invalidate such patents, it is the most convenient tool for the technologically unsophisticated courts to apply. Novelty and non-obviousness are much harder to prove/disprove. For a long time I have advocated a more rigorous examination process based on crowd-sourced prior art search as a prerequisite to litigation. By allowing low-quality patents to be litigated, we have endangered a whole class of technology in terms of legitimate patent protection. The patent bar has been shortsighted to resist a higher standard of patent examination, for software patents in particular where domain expertise has been sorely lacking in the profession.

  • [Avatar for Paul Cole]
    Paul Cole
    September 8, 2014 05:37 am

    I could not agree more: see my posting on this blog when the first Federal Circuit opinion was handed down

    https://ipwatchdog.com/2012/07/15/cls-bank-v-alice-corporation-reveals-a-house-divided/id=26433/

    In comment 15 to that posting I added:

    “Why, in a few words, was this a good invention? If there is no easy answer to that question then the present decision is at risk of reversal. Especially in a non-expert court such as the Supreme Court which has in recent years seen a parade of weak patents coming before it and from the decisions handed down has evidently been less than impressed. It is not attractive for a judge to support an apparently worthless patent on a technicality, and if that is how they perceive the inventions in these patents, the Supremes will be inclined to point out the similarities between this case and Bilski and hold the claimed inventions to be unpatentable. If there are cracks in the structure of fact before the Court, then the law will not be adequate to paper those cracks over and the decision will be reversed. I express no concluded view as to the outcome, merely identify a factor which in my opinion will have a powerful influence on the outcome. And identifying that question does not require legal knowledge, merely an understanding of how judges’ minds usually work. That changes less than you might think in travelling from London to Washington.”

    Everything said in that comment proved true.

    What Alice ought to have been able to say before the Supreme Court is: “We made a marvellous invention. It achieves X. Nobody before could do that, and is of technical and practical utility. Everyone in the banking industry uses our technology. Those bas***ds over there stole it, and now they are trying to throw out our patent for a trivial technicality.” Compare that with what they did say in oral argument, including the weekend coding concession, and weep.

  • [Avatar for Curious]
    Curious
    September 7, 2014 10:13 pm

    “If you review the Alice form paragraph rejections from the USPTO, the reality that the USPTO is withdrawing Notices of Allowance and issuing Alice rejections.”
    … and I have several notice of allowances I need to report out from last week (in the “software” arts). The sky is NOT falling.

    “And that strikes me as crazy.”
    Excuse my criticism, but that doesn’t hold a lot of water with somebody with a degree in political science. While programming can be complicated, those or ordinary skill in the art are capable of a lot — I don’t even want to get into the vast number of computer programs that help programmers program (e.g., the programmer comes up with the high-level concepts and the computer program generates the source code).

    “It didn’t help that the lawyer for Alice was asked by the Supreme Court, well, couldn’t somebody code this in a weekend? And he said yes”
    Patentability is not based upon “how difficult is it to practice?” If it was, every “simple mechanical” patent would be thrown out.

    “Shouldn’t that nearly universal experience be a clue to somebody that maybe this isn’t as trivially easy as the judges think it is.”
    Not quite. Most software patent applications, even if functionally described, are fairly easy to implement. The problem is that sophisticated software probably has hundreds if not thousands or tens of thousands of separately patentable inventions embodied therewith. Windows 8 (just the operating system and not the programs that run on top of it) has somewhere along the order of 30-80 million lines of code. Get one thing to work correctly isn’t that hard … getting it to work correctly with everything else is another story.

    “People have been writing claims that don’t emphasize the technological contribution of the innovation”
    Claims are written to distinguish the invention over the prior art – not to describe the invention.

    “There have been four patentable subject matter cases go to the Supreme Court in the last four years and they’re not going to fight this battle any more”
    Two biotech cases (one with poor claims) and two cases dealing with the financial industry. This is nibbling at the edge – not a strike deep at the heart of software eligibility.

    “We need to do something about the problem of nuisance value litigation and over claiming patents”
    Nuisance value litigation has nothing to do with 101. As for “over claiming” … whatever that means … 102 and 103 have always been the best tools for getting rid of patents that are not valid because of prior art.

    “I just don’t see an awful lot of judicial restraint.”
    I think the Alice decision was very restrained – the wrong decision, IMHO – but very restrained. SCOTUS essentially said this looks like Bilski and thus it is an abstract idea … and BTW, adding machine elements don’t make your claim directed to a machine. As I have noted in the past, SCOTUS could have done a lot more with Alice but they didn’t — that is restraint.

    “Because if you’re a serious litigant you’ve got that stuff before you file anyway.”
    Except with computer-implemented claims, you really don’t know what is going on under the hood. You can make pretty good guesses, but many companies don’t let you look at their source code just for the asking.

  • [Avatar for NWPA]
    NWPA
    September 7, 2014 06:43 pm

    What is scary is that Lemley feels pretty confident that he has a total victory meaning that he believes that Alice gives the Google judges enough latitude to invalidate essentially all software. I for one won’t give up the fight.

    Pretty amazing how the law changed. No legislation, but judicial activism where the judges who are ignorant of science just know that these claims are bad. I guess their reason ’cause we feel it man.

  • [Avatar for NWPA]
    NWPA
    September 7, 2014 06:32 pm

    Lemley >>over claiming patents,

    I would like Lemley to point to some “over claiming.” I think one of the problems is that with people like Lemley we have someone that has profoundly changed our patent laws for the worse and it comes from someone with no science background and no experience in patent law but litigation. I would also point out that Lemley has said he wants to get rid of the “problem” of software patents and has much to do with the interpretation of 101 at the SCOTUS.

    Still, what strikes me about all this is the continued misreprepresenations of the technology. Functional claiming in software is fine. It is not only fine it is what software engineers do to specific software and it how software is described in textbooks. The idea is that there are so many known ways to solve problems that they are functionally described to encompass the known solutions. Now, for the ignorant judges on the FEd. Cir. and SCOTUS who have no foundation in science at all, this explanation just makes their eyes roll. And people like Lemley exploit this and fabricate nonsense theories. But, let’s be clear that all of this has been a giant PR job with Lemley front and center.

    It was interesting that Lemley as much admits that 101 jurisprudence is not patent law but a good vs. bad test that has been put at the front of 102, 103, and 112.

  • [Avatar for NWPA]
    NWPA
    September 7, 2014 06:23 pm

    Lemley>>>Yeah. People have been writing claims that don’t emphasize the technological contribution of the innovation.

    That is total nonsense. The fact is that patent attorneys write claims that reflect how one skilled in the art thinks about these inventions. Lemley is playing a little game with this. What tech aspect was not emphasized? You know, the best indication of how the claims should look is right on the internet. Just look at one of the sites for outsourcing software. The spec looks an awful like the claims that Lemley says aren’t technical.

    And this functional nonsense is outrageous. Viewed from one skilled in the art (which I am), a functional claim is fine under 112 as long as I would know how to write the code to implement it without undue experimentation. And that is exactly how it is treated in computer science textbooks and everywhere else in computer science.

    Lemley tell us how you are protecting software in your start–up? And tell us how software has no structure which you wrote in one of your papers.