“All Prior Art” algorithm won’t stop bad patents or actors

Stopping new inventions dead in their tracks by making them “automatically” unpatentable might sound like a good idea to some.

However, programming a computer to disclose an endless combination of dubious inventions is more effective at stopping bad patents in theory, not practice.

At least that is what three experienced patent attorneys told me recently. Publishing millions of pieces of supposed prior art will not stop bad patents (really, unpatentable inventions)  from being issued or make legitimate invention-ownership less difficult to discern.

More Style than Substance

All Prior Art is a project that attempts to “algorithmically create and publicly publish millions of pieces of possible new prior art, thereby making the disclosed concepts unpatentable.”  All Prior Art is trying to take highly obvious ideas out-of-play. Unfortunately, preventing patents from being granted, even those that do not deserve to be, is not a simple fix.

In order to be admissible, prior art must be enabling; i.e. the disclosure must describe how the invention can be produced — not something that All Prior Art addresses.

The brainchild of Alexander Reben is a self-described engineer and artist with a background in robotics and applied math.  Mr. Reben’s random configurations have been described to me as “a mishmash of questionably inventive ideas.” Even if there are millions of combinations, it is highly unlikely that they will serve to invalidate a perspective patent, let alone one that an NPE may be enforcing.

All Prior Art may be more style than substance. Sure, inventions that do not meet the appropriate tests of novelty should be prevented from issuing or invalidated if they are enforced (or, maybe, even if they are not). Even if this scheme were to work, it would weaken the prospects of many inventors, universities and SMEs, and dissuade investment in innovation.

Missing Description

The theory behind Mr. Reben’s invention is that if a computer can spew out enough combinations of prior art — i.e. publish millions of pieces of data that describe inventions — the subject matter disclosures will preempt patent issuance. It’s just not that simple.

The problem is that for any prior art reference to serve as an invalidating disclosure, it has to be enabling. And the fact that no human being has come up with the combination is a pretty good argument against enablement.

The means it must be described so that a “person of ordinary skill in the field of the invention can practice the subject matter based on the reference, without undue experimentation.” (Sanofi-Synthelabo v. Apotex, Inc., 550 F.3d 1075, 1082 (Fed. Cir. 2008).

When it comes to All Prior Art’s prior art, the “description” is missing, according to Bloomberg BNA’s Tony Dutra.

“The virtually infinite number of combinations will generate over 99 percent dreck,” chides Mr. Dutra. “But, like the ‘infinite monkey theorem’ that predicts a monkey hitting typewriter keys at random for an indefinite time will almost surely, eventually type a given text, Reben believes his project will generate at least some combinations that someone, someday might try to patent. It’s that someone [Reben] aims to stop.”

Patent the Universe

The tech media, notably the DailyDot, encouraged by with the idea of that patents can be neutered, has taken an ill-conceived idea and made it worse with a misleading, inaccurate headline: “One man is trying to single-handedly create every patent imaginable.”

A sister website All The Claims is attempting to achieve a similar objective, but with the use of claims and a more verbose alternative.

“Hard to Take Seriously”

“Bruce, this approach is hard to take this seriously, especially with an admission that most of the inventions generated will be nonsensical,” one patent attorney wrote to me. “Pertinent excerpts from the Patent Office Examiner guidelines illustrate that:

‘In determining that quantum of prior art disclosure which is necessary to declare an applicant’s invention ‘not novel’ or ‘anticipated’ within section 102, the stated test is whether a reference contains an ‘enabling disclosure’… .’”

“To be invalidating prior art as to make a meaningful contribution,” a former USPTO examiner, inventor and Fortune 100 Chief Patent Counsel told me. “If utility is missing, it will not be seen at invalidating. Prior to be taken seriously by an examiner or in litigation needs novelty and utility, or specific application.”

“There are billions of combinations, but so what? There needs to be novelty and utility to qualify as prior art. This scheme might be well intended bu ti hurts the little guy more than the big tech player. I doubt that courts will take it seriously or that automated generation of inventions will have a meaningful impact.”

Trivial Variations

“Is Mr. Reben really attempting to automate the generation of trivial, random variations of existing stuff?” asked another patent attorney. “It is already difficult to patent trivial variations, as they are ‘obvious’.  The term ‘automated generation of patentable’ is kind of an oxymoron. This approach  should be titled ‘automated generation of trivial variations of existing inventions’.”

Non-Patents

So for now, at least, bad patents (really, non-patents) will continue to be issued by the USPTO and other major patent offices in large numbers because they lack the resources to prevent them.

Inter Partes Reviews and other mechanisms will continue to try to eliminate or to help fix erroneously issued patents, and litigation will remain the only reliable, albeit, costly way to determine if a patent is valid and infringed.

If it were easy to solve patent problems algorithmically with a PC, mathematicians would be higher paid than patent attorneys. That’s not likely to happen anytime soon.

Image source: ttconsultants.com