Skip to content

The Most Dangerous Problem in Patent Law: Unreliability

The Most Dangerous Problem in Patent Law: Unreliability published on 21 Comments on The Most Dangerous Problem in Patent Law: Unreliability

What’s worse than a property system that is inefficient, inequitable, or delayed? Answer: A property system that is unreliable – one in which determinations of property rights are unpredictable, capricious, and incoherent.

Over the past decade, the reliability of the patent system has been critically damaged. No one can predict whether, or when, an application will be allowed. No one can determine whether an issued patent will be found valid or invalid at court – nor whether any higher courts will respect lower courts’ decisions. And no one knows what this turbulent body of law will even look like, twelve months from now.

This unreliability in patent law hurts nearly everyone: applicants, patent owners, competitors, patent examiners, the patent office, and the court system. The only parties that benefit from this chaos are academics, litigators, and speculators – sure signs of a deeply dysfunctional legal process.


1. Introduction

The most basic function of any property system is public notice: a centralized source of information about who claims what property. In patent law, public notice is its most basic cornerstone – it is even more fundamental and essential than a government-enforced right to exclude.

This public notice function of patents enables a consistent and dependable framework for resolving key questions. Let’s say that ten individuals – some patent practitioners, and some technologists unfamiliar with patent law – were given the following materials:

  1. A patent claim, with a description of the invention;
  2. Two prior art journal articles presenting some elements of similar techniques in the same field;
  3. A description of a competitor’s product; and
  4. 35 USC and 37 CFR, or a more basic description of the legal framework of patents.

Given this information (and an abundance of time and caffeine), each member of this group should reach similar conclusions on the following questions:

  1. Does the claim comply with the formal legal requirements – clarity, written description, enablement, statutory class / transitional phrases, etc.?
  2. Is the claimed invention technically distinguished from the prior art?
  3. Would a patent examiner allow or reject this patent claim?
  4. If the patent were to be allowed, does the claim fully describe the competing product?
  5. Would a court find that the product infringes? If so, what penalties would the infringer face?
  6. Would appellate courts uphold the decisions of the examiner and the trial proceedings?

It is impossible to overstate the importance of being able to answer these questions reliably – i.e., without resorting to litigation. A well-functioning patent system will yield clear answers to such questions in the great majority of scenarios. Yet, the current U.S. patent system falls far short of this goal. The fabric of patent law is shot through with uncertainty, of many types:

  • Inconsistency: Nearly identical applications will take different paths through the patent system, and will reach different outcomes.

  • Capriciousness: Critical legal decisions are based on the subjective sentiments of the reviewer.

  • Incoherence: Contradictory patent law principles enable reviewers to reach mutually exclusive, diametrically opposite conclusions based on the same facts.

  • Undependability: Reviewers do not consistently support the determinations of other reviewers. Higher courts routinely vacate the conclusions of lower courts and examiners; lower courts exhibit difficulty (and sometimes resistance!) to higher courts’ rulings.

  • Volatility: Today’s patent law principles are quite different than they were a year ago, and are likely to be quite different still a year from now.

Together, these factors undermine the public notice function of the patent system.


2. Sources of Unreliability

The unreliability pervades all three stages of the practice of patents: patent examination, patent assertion, and judicial review.

  • Stage 1: Patent Examination

    The unreliability of the patent system begins with examination. Of course, some unpredictability inevitable: the examiner’s search is expected to be broad enough to find references unknown to the applicant. But other causes of unreliability are unnecessary, unfair, and subject to misuse:

    • Discretionary legal standards. Over the past decade, many key legal tests of patent law have shifted from an evidence-driven technical analysis to evidence-free, subjective opinions:
      • 35 USC 101: Is the invention “abstract?”
      • 35 USC 101: What claim language sets forth a patent-eligible statutory class?
      • 35 USC 112(b): When does a claim sufficiently “set forth and distinctly claim the subject matter regarded as the invention?”
      • 35 USC 112(f): What types of claim language invoke the means-plus-function claim form?
      • 35 USC 102: What is within the “broadest reasonable interpretation” of a claim element?
      • 35 USC 102: When comparing a patent claim with a prior art reference, which claim elements will be given their due weight and recited meaning – and which claim elements will be disregarded, dismissed as immaterial, or overgeneralized?
      • 35 USC 103: When can references from different fields be combined to meet the claimed invention?
      These decisions are not resolvable according to evidence or objective rules. Rather, each issue is a “smell test,” based primarily on the reviewer’s opinion. Vast differences arise in their understanding and application of these concepts arise among the 9,000 USPTO examiners currently making such decisions.

    • Procedural uncertainties. Despite the voluminous detail specified in 37 CFR about the process of patent examination, several key procedural questions remain unclear – and are subject to discretionary application by examiners and supervisors:
      • Restriction requirements: Will an examiner restrict materially identical independent claims, just because they are in different classes? Or will the examiner restrict the dependent claims as showing different “species” of an invention?
      • The “RCE Hole”: After an applicant files an RCE, will the examiner promptly reexamine the application, or will it languish on the examiner’s “special docket” – sometimes as much as several years – before the examiner reviews it again? 1
      • Search scope: When the references of record have been traversed, how much additional searching must the examiner perform? Will the examiner opt to allow it, or will the examiner reject it again based on different references that might not be any better? Will the examiner address the elements that are missing from the traversed references by simply tacking on additional references – six, ten, twenty – to cobble together a new rejection like a ransom note? 2
      • Inexplicable delay: If the patent office simply does not want to allow an application, will it tie up the application indefinitely (such as under the the Sensitive Application Warning System (“SAWS”) program) – or will it simply refuse to examine it for decades?
      • Patent term extension and adjustment: When a patent issues, how much patent term extension or adjustment will the USPTO award, based on delays in the examination process? 3

    • Volatile examination framework. The legal framework for patent practice is shifting rapidly – both due to external pressures (patent reform and case law) and USPTO-led initiatives (new programs, both official and clandestine, that continuously adjust the rules of examination). And there is no end in sight to efforts to modify patent law, both outside and within. In addition to requiring significant effort from examiners and practitioners to maintain an understanding of the current law, this volatility routinely imposes retroactive rules upon previously filed patent applications – while restricting the ability of applicants and patentees to update their applications and patents to comply with such laws. 4

    As a result of these unpredictable factors, it is impossible to predict, with any reasonable degree of confidence, whether any patent application will be issued as a patent; what claim scope will be allowed; and how long or expensive the process will be. All three questions are controlled much more strongly by the capricious sentiments of the patent office than by objective evidence or the rule of law.

  • Stage 2: Patent Assertion

    In a well-functioning patent system, an issued patent should be easy to assert, based on a strong presumption of validity. Notwithstanding unusual circumstances or new information, both patent owners and competitors should expect a patent to be enforceable against products that fit the claims, as well as the available enforcement mechanisms. Business strategy and investment can then be adapted based on these expectations.

    On the contrary, the assertion of a patent today is fraught with peril, due to the enormous complexity in the enforcement of patent law. Consider the variety of issues that courts been opened – and left materially unresolved – in just the past few years:

  • Stage 3: Judicial Review

    In a well-defined property system, appeal and review mechanisms are applied sparingly to difficult questions that arise in the crevices between legal technical and principles. Decisions are generally consistent, which improves the coherence of the legal process, reduces unnecessary appeals, and avoids gamesmanship such as forum-shopping.

    But U.S. patent law is a hotbed not just of litigation, but of relitigation – not just because venues vary and consistency is very poor, but because deference is practically nonexistent. Consider:


    As a result of this tangled review process, many important cases have a long and tortured litigation history, sprawling over the entire court system and requiring repeated rehearings based on new legal standards and theories – such as Teva v. Sandoz; Classen v. Biogen; AMP v. Myriad Genetics (two trips to the Supreme Court); Apple v. Samsung (currently heading into its fourth trial), and eBay v. MercExchange (six years of litigation). These cases seem to be shuffled among the courts like hot potatoes, until the issues are so muddled and distended by re-re-review that it’s difficult to draw meaningful legal principles from them.

    The implications of these factors, for any particular dispute, is that decisions only matter once the options for review are exhausted. Until then, opportunities to challenge the decision of an examiner, judge, or court are not only open – but promising, so long as one can only cobble together a legal argument that’s not totally specious. This availability of repeated challenge diminishes the value of every opinion, and encourages protracted disputes – especially when the stakes scale into billions of dollars.

In view of this extremely imbalanced patent process, a peculiar irony arises over the courts’ stated mistrust of patent practitioners. To quote the Alice opinion: “This Court has long warned against interpreting § 101 in ways that make patent eligibility ‘depend simply on the draftsman’s art.'” 6 Yet, the overcomplexity of the patent process, and the pliability of the patent system to novel legal theories, has made patent law more dependent than ever upon the “art” of zealous patent representation.


3. Effects and Evidence of Unreliability

This abundant array of sources of unreliability in the patent system eradicates the public notice function of the patent system. No one can predict whether a published application will issue as a patent; no one can predict whether an issued patent will be upheld by the courts.

This unreliability critically harms nearly everyone involved in the patent system:

  • Patent applicants are unable to forecast the odds of receiving a patent, with good claims, in a particular time frame.

  • Patentees are unable to rely on issued patents – even after extensive investment to secure a patent and assert it through multiple layers of judicial review, only to have it invalided at the next level up – either because the law was unclear, or because the law changed unpredictably since the patent was filed or issued.

  • Competitors are unable to distinguish patents that are valid and must be respected, vs. patents that can be invalidated or avoided through the judicial process. Those who license or design around a patent may later see their resources wasted when a competitor successfully challenges it; while those who operate on a good-faith presumption of patent invalidity may nevertheless find themselves at risk of damages for induced infringement.

  • Examiners see their decisions reversed due to changes in the law, or the subjective views of a supervising reviewer who simply feels differently about a particular legal test.

  • The USPTO suffers a chronically poor reputation, due to the constant attack from a zealous body of reform advocates who are more interested in their agendas than a system of law that can be administrated with reasonable certainty.

  • Judges see their opinions reversed on review, often as part of a higher court’s particular agenda for reshaping the face of patent law.

  • The Supreme Court must spend resources attempting to state legal theories that are beyond its core legal knowledge, and waste time issuing opinions with few consistent principles, thus guaranteeing that the Court must pick them up again very soon to clarify further.

  • Small entities – sole inventors and small companies – are selectively excluded from patent enforcement by the enormous costs involved. While these parties can affordably win a patent, enforcing them is a different story, because the legal costs of enforcement are priced well beyond their budgets: the average litigation is between $3 and $10 million. This disparity creates a deeply inequitable system, where anyone can get a patent, but only the most well-funded entities can actually use them.

  • Only a few parties actually benefit from this chaotic environment of patent law:
    • Academics, whose often focus on ideology at the expense of process efficiency and practicality, and who bear neither responsibility nor the consequences of unreliability in the patent system;
    • Litigators, who enjoy increased demand for their services when disputes over basic issues require extensive litigation and re-litigation to resolve; and
    • Speculators, who profit by tearing down patent value.

Moreover, evidence abounds as to the damage to the patent system wrought specifically by its endemic unreliability:


4. Root Causes of Unreliability

Looking at we are compelled to ask: Why has patent law become this way? If uncertainty is deleterious to the entire patent system, why is it so pervasive today?

We can cite several reasons:

  • Lack of familiarity with patent law and technology. The Supreme Court’s zeal to reform patent law is not informed by actual experience with the patent system: the nine seated justices have zero technical degrees, and zero experience in patent practice or the business of technology. That this body of unelected officials, with no accountability to any constituency, has assumed the mantle of “protectors of the technology economy,” should be deeply troubling. (Of course, it has done so in response to the tepid responsiveness of Congress to perceived problems in patent law.) Similarly, consider the Electronic Frontier Foundation proposal to the USPTO to “require working code with every claim in a software patent” (apparently unaware that the USPTO has no authority to create substantive new requirements 8).

  • Poorly considered changes to patent law. “Patent reform” is occurring far too fast, far too haphazardly, and with far too little holistic projection of the consequences. Consider the patent reform proposal that tacitly lumps university technology transfer and startup companies into the definition of “patent trolls”.

  • Discretionary patent rules. “Discretion” is a legislative and jurisprudential crutch: it is used when a higher authority is insufficiently savvy to craft specific rules, and delegates such responsibility to a lower authority. Yet, starting with KSR v. Teleflex, the Supreme Court has gone wild with trading “discretion” for specific, objective rules that Court found too restrictive. The Court seems unaware of the extreme dangers of converting an entire property system to a “discretionary” one – but adamant in its intent to continue doing so.

  • Overabundance of self-interest. Those seeking patent reform appear too willing to sacrifice efficiency and reliability, in the form of a patent process based on objective rules, in order to tilt patent law in favor of their agendas.

  • The tyranny of the majority. Patent law change is being sought through force of consensus and lobbying, rather than responsible management by a disinterested organization. As a result, organizations that are underrepresented in this reform process, such as small businesses, find patent law increasingly inaccessible and unsuited to their needs. Even USPTO Director Lee’s efforts for patent reform seems focused more on “reforming” applicants’ behavior, than improving the quality of the USPTO’s measurements and internal processes.
Simply put, there is too little interest in wrestling the patent system into a form that serves particular agendas, and scant interest in making the patent system work efficiently for everyone. And this skewed perspective is damaging the entire technology sector and the patent system at large.


5. Conclusions

For those of us who perceive the value of the patent system – what shall we do?

We must raise public awareness about the desperate need to improve the reliability of the patent system. We must spread the word about the variety of sources and causes of undependability in the patent system.

We must proclaim and document the damage being done by these flaws, and the dangers posed by the steady conversion of intellectual property to a discretionary award system.

We must call out judges, justices, and legislators on the unsuitability of their careless tinkering with the fabric of patent law.

We must recognize unreliability as an existential threat to patent law: continued devolution into dysfunction and incoherence may give traction to more radical calls to abandon the patent system, or revert to a “registration” system. This problem is even bigger, and even more pressing, than where the threshold questions of patent eligibility and validity are set: we should prefer a more reliable system, to one where the standards are set according to our preferences, but in an unreliable manner.

In short: Unreliability is the issue in patent law today. Let us prioritize it accordingly in our continuing dialogue over patent law.


Notes:

  1. As I have previously noted, this “hole” in patent examination is not a glitch; it was explicitly created as a stick that examiners could use to punish uncooperative applicants. The USPTO’s comments about this docketing process state:
    The USPTO believes these initiatives and the new examiner production system will work together to reduce the overall pendency of applications. Applicants seeking early allowance of an application can help eliminate the need for the filing of an RCE by addressing all issues in the application early in the prosecution. If an RCE is needed, cooperative efforts between the examiner and the applicant will go a long way toward obtaining early action of the RCE. Of course, for those applicants who do not desire fast action of their applications, the placement of RCEs on the examiner’s Special New docket will afford them the flexibility to better control the rate at which their cases are handled by the USPTO.
  2. Note that this tactic is explicitly endorsed by the USPTO. MPEP § 2145 includes this curious statement:

    MPEP 2145: Consideration of Applicant’s Rebuttal Arguments

    V. Arguing About the Number of References Combined

    Reliance on a large number of references in a rejection does not, without more, weigh against the obviousness of the claimed invention. In re Gorman, 933 F.2d 982, 18 USPQ2d 1885 (Fed. Cir. 1991) (Court affirmed a rejection of a detailed claim to a candy sucker shaped like a thumb on a stick based on thirteen prior art references.)

  3. The amount of patent term extension or adjustment that is due to a patentee is a surprisingly difficult question to answer – with some key questions unresolved. The USPTO has been inexplicably hostile to patentees in making such determinations – and the process of disputing the USPTO’s decision is also rather difficult.
  4. The volatile nature of patent law squarely violates the proscription against ex post facto laws under Article I of the U.S. Constitution.

    This aspect of patent reform is interesting. Legislative changes to 35 USC and 37 CFR are typically introduced with a specific future date on which such changes will take effect – e.g., the change in patent term from “17 years from issuance” to “20 years from priority date” specifically occurred on June 8th, 1995, and the America Invents Act featured a host of carefully chosen implementation dates. Yet, even more important substantive changes, arising from USPTO procedural changes and court decisions, are imposed with no warning and immediate effect upon past, present, and future patents. A constitutional challenge to this whiplash tinkering with patent law is long overdue.

  5. And more importantly: How is this even remotely compatible with the notion of due process? Gene Quinn’s article on this incident expresses a suitable level of alarm.
  6. This is hardly the first instance of a patent court disparaging patent practitioners – White v. Dunbar (1886) included the cautionary statement:

    Some persons seem to suppose that a claim in a patent is like a nose of wax, which may be turned and twisted in any direction, by merely referring to the specification, so as to make it include something more than, or something different from, what its words express.

  7. I thought this sort of thing was blatantly illegal, and might be of interest to bodies such as the Securities and Exchange Commission – but, alas, it seems not to be the case.
  8. Of course, the USPTO learned a painful lesson on this topic in Tafas v. Dudas. It’s a shame the EFF was not similarly educated.

21 Comments

This is one of the most important and interesting papers I have read in recent years and a copy is retained in my permanent collection. As a primary source of key US references for European practitioners on these various issues it is invaluable. So firstly many thanks.

One of the major problems is the failure of the US court system, including the PTAB, to set a consistent, reliable and practical approach to the main issues of patent law.

For example, the approach to unobviousness has descended into chaos. New function or result has been downgraded into a secondary indicator. Field of endeavour is considered of marginal significance. Grotesque opinions such as that in In re Bigio where a hairbrush and a toothbrush were held to be in the same field of endeavour were handed down, and in In re Klein it took the Federal Circuit to reverse PTAB and hold that an shopkeeper’s drawer with a movable partition is not in the same field of endeavour as a mixing jug. Under the Graham test in the US, similarly to the Windsurfer test in the UK there is no legally prescribed standard, and subject to some prescribed preliminary enquiries each case has to be decided according to the evidence and the arguments of counsel. Nevertheless evidence comes in a range of predictable categories, and these are not all of equal intrinsic weight so that there is room to develop a reasonably systematic approach.

Before the EPO the technical problem test is applied consistently. The Examiners know the rules and so do the attorneys. There may be problems, but the overall system is very practical and works. And for those interested, new function or result is at the root of the technical problem test. References are not randomly combined: the primary reference from which the technical problem is defined has to be within the field of endeavour of the invention.

How are attorneys to write workable patent applications, and how are examiners to do their job if the legal framework within which they work is chaotic? Even if an examiner decides on a realistic and practical approach, how is he or she to know that they will be upheld if appealed? If the law becomes settled then patent examination can become more systematized, and hopefully the prospects of enforcement will become more certain. The risk of a late found, but devastating, Korean national application not found in the search tools available to the applicant and examiner remains and increases as technology spreads into a greater range of countries with diverse languages. The resources put into litigation are at lest one order of magnitude greater than those put into prosecution, so new arguments will surface and new arguments will be developed. But at least we should all be working in a legal framework we all know and understand.

Thank you for the kind words! I agree with many of your observations.

And for those interested, new function or result is at the root of the technical problem test. References are not randomly combined: the primary reference from which the technical problem is defined has to be within the field of endeavour of the invention.

Yes, I have certainly encountered this problematic type of examination – I even wrote about it in one of my earliest USPTO Talk articles.

Your comments lend support to my understanding that the EPO does not typically play these kinds of games. And I believe that there is a deep cultural difference between the EPO and the USPTO – about the acceptable level of persuasiveness of the examiner’s position – that is producing very different experiences. There will need to be a cultural shift within the USPTO examining corps to address his problem, which first requires exposing it – which is a central purpose of this blog.

Before the EPO the technical problem test is applied consistently. The Examiners know the rules and so do the attorneys.

I am curious about this part, because the research that I’ve done into the EPO’s “technical effect” requirement raises similar confusion, for largely the same reasons. For example:

…etc. Indeed, judges’ elaboration in the application of Alice seems to raise eerily similar problems as the EPO’s “technical effect” nuance – such as packaging undefined legal requirements like “technical problem,” “technical solution,” “technical features,” etc. If anything, persistent difficulties with the EPO’s “technical effect” requirement should have strongly discouraged the Supreme Court from the Alice framework.

On the upside: I guess the fact that the EPO and USPTO now have the same basic problem promotes international patent law harmonization…?

Excellent article David. Also excellent rebuttal of the critiques that missed the whole point of the article. However, I was reminded of my grandfather’s caution: when an intelligent person argues with fools it can quickly become difficult to determine which is which.

Every legal system ever invented includes significant variability in outcomes. There has always been patent litigation, and patents are inherently subjective because language is an imperfect vehicle to fully encompass inventions.

Nonetheless, every single problem you decry has two direct causes:

1) the costs of defending or asserting patents are completely out of line in a great many cases with the economic value of the invention. If the minimum cost to defend or assert a patent is $1.5M, then nothing with a market value of less than about $5M can be practically protected via patent. If inventions or alleged infringements add up to a few hundred thousand in value, the patent system can’t be reliably used to protect or recover.

2) Virtually all of the distortions in patentability and all patent jurisprudence springs from the mania to patent the unpatentable: namely software and method patents. If you magically remove every software and method patent from the universe, the whole system looks about a million percent more reliable and useful. Why are patents on software and methods an insane concept? Because software is the expression of ideas using a language, and we don’t patent that activity because any two factfinders can find reasonable disagreement about what they see. Business methods are simply ideas, with no meaningful technological anchor Once patent adjudication turns into literary criticism, which is what you see today, its chaos. Patents are for things, not ideas.

Also, to be sure, you show your own bias here- lots of worry about the poor small inventor who can’t assert, and no worry about the fact that to be accused of infringement is to automatically lose- maybe your whole business- and often for absolutely trivial infringement, if it even exists. Software and method patents are economic weapons of mass destruction that have managed to severely damage the system itself.

Wake up already….

If you want the problems to go away, ban software and method patents, all of them, now, or come up with a bright line test better than Alice. I have some modest proposals about that….

The costs of defending or asserting patents are completely out of line in a great many cases with the economic value of the invention…

According to who?

In our current system, that decision is left to the patentee, who has the best vantage point. Once the cost of litigation exceeds the cost of defending or asserting the patent, the patentee drops out of litigation, so that is a self-correcting problem.

Are you suggesting allocating that decision to accused infringers – maybe allowing “the patent is worthless” as an affirmative defense? This raises the obvious question: if the claimed technology isn’t worth anything, then why are you infringing it, instead of just designing around it?

Because software is the expression of ideas using a language…

“Expression” is covered by copyright. Patents cover functionality. These are different concepts.

Solutions to technical problems that involve a particular type of device are patent-eligible inventions. This was the point of Diamond v. Diehr: a solution to a problem does not become patent-ineligible simply because the solution involves some calculation – even if it uses a formula that’s centuries old.

Indeed, it does not even matter whether that solution is software or hardware. According to the Church-Turing thesis, every implementation of a Turing machine is functionally equivalent. A device with dedicated circuitry has the same functional power as a device with a processor and software. The distinction that you seek between these implementations is futile and artificial.

“The costs of defending or asserting patents are completely out of line in a great many cases with the economic value of the invention…According to who?”

According to plain logic. Regardless of the merits (or lack) of a given assertion, the fact that the minimum litigation cost is around $1.5M remains just that: an economic fact. If you have an invention worth a few hundred K or an infringement worth a few hundred K, it’s plainly uneconomic to use the court system for enforcement, for either side. Unless the costs of adjudication can be brought down to Earth, massive distortions and economic injustice are inevitable.

“This raises the obvious question: if the claimed technology isn’t worth anything, then why are you infringing it, instead of just designing around it?”

Because 90% of defendants in method and software cases had no idea they were infringing, and the infringing portion/value of the solution is vastly overstated by patentees. Oh sorry, that NEVER happens in the real world.

“A device with dedicated circuitry has the same functional power as a device with a processor and software. The distinction that you seek between these implementations is futile and artificial.”

Then why not just execute in hardware if you want a patent? Oh, because there IS a difference in the real world, isn’t there?

What does ‘artificial’ mean in this context? Everything about patents is artificial. What other human creations are covered by both patent and copyright?

We don’t patent literature because the world’s literature cannot be fairly or practically examined for each new claim, and because ideas can be expressed in myriad shadings. Nor can the world’s software be actually examined in reasonably repeatable ways. Why don’t we allow patenting of clever legal arguments? They are useful. They have structure. They may be novel and non-obvious.

Trying to control everything that happens under the sun with a grotesquely expensive and asymmetric legal regime is a very, very bad idea. Patents work well for machines, compositions of matter, and chemical or mechanical processes. You need to wake up, as I said above, to the problem in the patent system. Its not the reliability- its the inherent unreliability of patenting “methods” that amount to ideas and softwares that amount to ideas.

And who is engaging in futility here? Someone on the wrong side of a bunch of 9-0 USSC decisions maybe?

Regardless of the merits (or lack) of a given assertion, the fact that the minimum litigation cost is around $1.5M remains just that: an economic fact.

If litigation costs are your concern – you should definitely be on my side of this argument about inconsistency.

When the law is clear, the parties don’t need to resort to litigation to resolve a patent dispute: they agree on the likely outcome, and can operate on that basis without litigating it. Why would the parties spend $1MM or more, just to get the answer that they both expect? But when the law is vastly unclear like it is now, everything must be litigated – and re-litigated, and re-re-litigated – to get an actual answer.

Unless the costs of adjudication can be brought down to Earth, massive distortions and economic injustice are inevitable.

I couldn’t agree more, Martin.

But you cannot have it both ways. The patent system cannot simultaneously allow any reviewer to say, “I just think it’s ‘abstract’ / trivial / overbroad, etc.,” and to arbitrarily flush cases down the drain (only to be overturned by the next level in the review process), and also allow parties to reach an efficient resolution of these very issues. The pursuit of “patent quality” through this “smell test” technique has exacerbated inefficiency and wasteful litigation of issues that should be predictable.

Because 90% of defendants in method and software cases had no idea they were infringing

That’s not very surprising. Do you honestly expect defendants to admit that they knew about the patent at stake in an infringement trial? I’m guessing that 90% of people pulled over for speeding say that they weren’t.

It’s like this:
Top Android Executive Says Google Didn’t Copy Apple’s iPhone
Uh huh. If you ask Apple this same question, you might get a slightly different answer.

Then why not just execute in hardware if you want a patent? Oh, because there IS a difference in the real world, isn’t there?

Certainly. Software can be designed, built, and tested at a lower cost than custom-designed hardware. Software inventions can also be more efficiently incorporated into an existing device, and updated via patches, etc.

But at a technical level, when the invention is used to solve a particular problem, it typically does not actually matter whether you use hardware or software.

For example, an MPEG-4 codec can be designed as a dedicated circuit, or as a software module. But when a device decodes video using this technique, both implementations achieve the same result: they are functionally equivalent. Every logical invention that can be achieved using specially designed hardware, can also be achieved using specially designed software – and vice versa. This is a necessary outcome of the Church-Turing Thesis.

We don’t patent literature because the world’s literature cannot be fairly or practically examined for each new claim…

No, we don’t patent literature because literature is not functional. Patents protect functional inventions.

And who is engaging in futility here? Someone on the wrong side of a bunch of 9-0 USSC decisions maybe?

If Bernard Bilski or Alice Corp. came to me today with their respective inventions – or any other inventions resembling them – I would advise them not to file a patent application. I don’t have a problem with that.

However, there are a great many inventions that look nothing like Bilski’s or Alice’s inventions – that provide legitimate, technical solutions to legitimate, technical problems (like making devices run faster, or conserving memory, or extending battery power, or resolving queries more accurately, etc.) – but that are jeopardized by the emptiness of these decisions. Worse, they are not even ruled invalid; they are maybe invalid to some reviewers, but valid to others. That is not an acceptable outcome of a property system such as ours.

“If, as you argue, 95 percent of filings are invalid – a consistent system would reject 95 percent of them, at each stage. We can certainly argue about how to classify applications as “good” or “bad” – but as long as the system produces the *same result*, it is consistent.”

This is not what was argued. What I said was 95 percent of patent applications are written by patent practitioners licensed to practice before the PTO. They are responsible for the garbage going in and being handled unpredictably. I never said 95 percent of patent applications are invalid. Never said anything that even resembles that.

There are frequently patent applications that are issued without ever having a non-final office action rejection by an examiner. Why are there so few? Because patent practitioners write a lot of bad patent applications – applications that lead to a lot of unpredictability. Its like trying to fly a badly formed frisbee. It’s going to fly unpredictably sir !

What I said was 95 percent of patent applications are written by patent practitioners licensed to practice before the PTO.

Fair enough – duly noted.

Because patent practitioners write a lot of bad patent applications – applications that lead to a lot of unpredictability.

Those two statements are disconnected – and that is exactly my point.

In a consistent system, a “bad” application will be consistently rejected at each stage of review.

In an inconsistent system – such as ours – a “bad” application might be allowed right away; might be allowed after protracted prosecution; and might be duly rejected. And if allowed, the patent might be upheld by the court, or might be invalidated. And both decisions might be supported by a higher court, or might be reversed – repeatedly. The only way to know for sure is to litigate it to the exhaustion of all appeals, which is why cases like eBay v. MercExchange have been kicking around in litigation for six years – multiple trips to the Supreme Court, etc. That is a problem, totally independent of “quality.”

Do you see the difference?

David,

“Bad” and “good” are irrelevant. The important axis is valid or invalid, right? What BEEF seems to be suggesting is that the patent bar works very hard to make invalid applications appear valid, and that this obfuscation is a major source of unreliability.

One other note: your example of the re-exam was unfortunate. That article was from 2012, and apples patent was affirmed in the end.

What BEEF seems to be suggesting is that the patent bar works very hard to make invalid applications appear valid, and that this obfuscation is a major source of unreliability.

But this raises the question: “valid” and “invalid” according to… who?

The patent bar works very hard to secure patents for their clients’ inventions that will be allowed, and also found patent-eligible if taken to court. Hopefully you’ll agree that this is a good thing.

The patent bar also wants to be able to advise clients, with some degree of accuracy, about which inventions will be found patent-eligible, and which inventions will not.

Patent practitioners have no interest in filing applications for a client that are likely to be found ineligible: it is a waste of our time and effort, and it leads to a frustrated client – not a desirable result in a customer-service-oriented field.

The difficulty created by the current state of patent law is that, for a very large range of inventions, no one can predict eligibility. The only way to tell is to file the application, and see how the examiner handles it – and then how the PTAB handles it – and then how various courts handle it. And because the law is so malleable, all of these reviewers can reach different conclusions over the same patent.

In these opinions, judges express such mistrust over patent practitioners’ work, like we are trying to trick them with technology mumbo-jumbo into reaching the wrong conclusion. On the contrary – we are just trying trying to follow the law, by describing inventions in the ways that judges have advised us to claim them, so that the resulting patents are upheld. What exactly is wrong with that?

Your example of the re-exam was unfortunate. That article was from 2012, and apples patent was affirmed in the end.

I was referring to this story, where Apple’s design patent – which was allowed, issued, and ruled valid and infringed at trial… only to be invalidated on reexamination.

I’m not denying that the system (USPTO, Courts) contributes to uncertainty. I’m just saying you’re ignoring/denying the bar’s contribution to uncertainty.

Many of class 705 claims I’ve seen are drafted in a way that obscures the scope of the claims (but does not necessarily make the claim indefinite) and makes it difficult to search the art. It is entirely within the rights of the bar to draft claims like that. But it does not help the system produce repeatable results. More straightforward claims, focusing on the actual invention, and without obfuscating terminology would lead to more repeatable results at all levels of the system.

Many of class 705 claims I’ve seen are drafted in a way that obscures the scope of the claims (but does not necessarily make the claim indefinite) and makes it difficult to search the art.

Examiners are experts at claim interpretation, right? When I file a patent claim, I’d like to presume that the examiner can readily identify its scope.

When I draft claims, I write specific details where I believe limitations matter, and fewer details where I intend the claims to be broad. I expect details in the claim body to be regarded as limiting and given weight; I expect details not in the claims (but in the specification) to not be limiting. This is all pretty basic stuff, and I’m always baffled when examiners err on either side (usually by disregarding important details in the body of the claim, when analogizing my case with a reference that doesn’t match those limitations).

As far as intentionally “obscuring” the claims:

  • Some applications are difficult to understand because they are translations of foreign documents.
  • Some applications are difficult to understand because they were written pro-se, or by a new associate, or because the applicant or practitioner are not PHOSITAs (the old “I’m a physicist, but I understand software because it’s easy, so surely I can handle your (compiler optimization | database query processor | software architecture) invention” mistake).
  • Some applications are difficult to understand because the technology is legitimately hard. (I once received an eight-page technical article about a database processing technique that required an 80-page specification to explain at an acceptable level. Journal articles can be extremely dense, and can require a whole lot of domain-specific knowledge that examiners can’t be expected to have.)
  • Some applications use terminology that does not match the art, because the applicants were not aware of prior work by other technologists (or vice versa!)
  • Some applications use different terminology to reflect actual differences that the examiner isn’t picking up.

My point is: I have never encountered an instance of a practitioner intentionally obscuring a patent application. I hear about it all the time in these discussions – even from judges – and I yet to see a single instance that suggested “obfuscation.”

Besides: Examiners have a variety of tools at their disposal, to address claims that they don’t understand. First and foremost: Call me! If you, Mr. / Ms. Examiner, don’t understand my claims, then a jury might not either, and I want to work with you to clarify any aspects that seem fuzzy. Alternatively: Reject it under 112(a) or 112(b), or suggest an amendment as an informal objection, etc. Bottom line: Do something, other than just “punting” by guessing at its meaning, and allowing or rejecting it via the path of least resistance.

As far as I can tell, that’s a complete denial of any responsibility on the part of the patent bar.

While we both only have anecdotes, it is much more reasonable to argue existence from anecdote than to argue against existence by annecdote.

There are certainly bad practitioners out there, and bad patent claims. Whether or not they were intentionally obfuscated is immaterial. (But I do take issue with the frequent demonization of patent practitioners: in addition to being untrue, these kinds of ad-hominem attacks do not facilitate conversation.)

But my point remains that if an examiner is confronted with bad claims, the examiner has plenty of tools to force it into better shape – and rightly should. This is why we utilize and pay examiners, as gatekeepers of the patent grant.

Then you argument for why the bar is not responsible is based on all examiners doing a perfect job. That is not realistic. In reality, the response of the system will be a function of the inputs. Better claims would result in more predictable results.

Look at the problem this way: For those who perceive a problem and want to find a solution, which processes can actually be controlled?

Patent applicants cannot be “controlled.” It is a globally distributed group, with vastly different motivations and circumstances.

But the examining corps can certainly be “controlled.” It is a single organization, led by a director and a centralized policy group – the Office of Patent Legal Administration.

The wrong way to solve this problem is to instruct examiners to push down the allowance rate as a “quality” throttle. Jon Dudas tried this, and it failed catastrophically. You could take five patent applications, randomly choose one, and throw the other four in the trash: no guarantee of improved “quality” there.

The right way to solve this problem is… well, it’s described here.

(I think that our conversation has run its course – thank you for your contributions.)

This post is a blatantly false. You claim the sources of unreliability are patent examination, patent assertion, and judicial review. In fact, the source of unreliability is the patent application.

If the patent office is issuing poor quality, non-meritorious, unenforceable patents, its is because patent practitioners have written them. Over 95 percent of patent applications are prepared by licensed patent counsel. This patent bar loves to blame everyone but itself for the poor predictability of outcome of the patent system. However, they themselves are to blame for the poor predictability of input to the patent system.

These problems are actually unrelated. Let me explain.

Consider the entire patent system – the USPTO and the court system – as a black box, or a single component. Ideally, for any input – i.e., any particular patent application – the black box should produce the same answer, every time, at every level of review. That is:

* For a good patent application, the system should produce a positive answer every single time, at every level of review. The patent office should allow it; the PTAB should affirm the allowance; and the courts should find it valid. Same result, every time.

* For a bad patent application, the system should produce a negative answer every single time, at every level of review. The patent office should consistently reject it; the PTAB should affirm the rejection; and the courts should find it invalid. Same result, every time.

That is the definition of consistency and reliability. And this consistency should be completely independent of the quality of patent applications filed by the patent bar. If, as you argue, 95 percent of filings are invalid – a consistent system would reject 95 percent of them, at each stage. We can certainly argue about how to classify applications as “good” or “bad” – but as long as the system produces the *same result*, it is consistent.

But this is completely not what we are seeing from the patent system. Rather, the patent system is producing wildly inconsistent results at each level, for all of the reasons I’ve demonstrated above. “Good” patents are being variously allowed and rejected. “Bad” patents are being variously allowed and rejected. Both decisions are routinely reversed on appeal.

That is the problem that this article discusses. And it is completely unrelated to the quality of filings – each issue should be addressed independently.

The anecdotes above, demonstrating the unreliability of the system, have nothing to do with the “quality” of any filed application. Of course, the “quality” of the filing does not change as the patent moves through the system: it is the same disclosure and claim set – yet, determinations vary wildly as this static application moves through the process.

Leave a Reply to Paul Cole Cancel reply

Your email address will not be published. Required fields are marked *