Bonum Certa Men Certa

EPO Throws DABUS Under the Bus... But for the Wrong Reasons

Not the applicant is the problem

DABUS meme



Summary: The hype and 'hot air' (or various buzzwords) surrounding algorithms which do "something clever" (not a new thing!) persist; DABUS is being rejected only because it's non-human, which means that a human can reapply with the same so-called 'invention'

THE European Patent Office (EPO) abandoned patent quality quite some time ago and this became rather obvious when Team Campinos/Battistelli -- a nonscientific bunch with ex-military officers -- took control of the Office, promoting things they haven't a clue about. They simply lack background and qualifications. Today's (or yesterday's) EPO examiners are far more suitable for management jobs than existing managers, who are handpicked for connections (nepotism). As the old saying goes, it's not about what you know but who you know...



The news about DABUS was covered here quickly (when the EPO came out with the statement). It's related to the issue of the EPO granting software patents in Europe under the guise of "HEY HI" (AI), albeit not so directly. In Europe, "HEY HI" is routinely used as an excuse for granting illegal patents. Less than 2 months ago this was even integrated into examination guidelines (in effect since). The same semantic nonsense had been leveraged to bypass 35 U.S.C. ۤ 101 at the U.S. Patent and Trademark Office (USPTO) -- another issue we covered here several times this year.

Since real journalism about patents is virtually dead the DABUS announcement was covered only by patent maximalists and their outlets. We know this for a fact because we've looked carefully until this morning. Who covered it and how?

"Since real journalism about patents is virtually dead the DABUS announcement was covered only by patent maximalists and their outlets."Well, the EPO mouthpiece IAM said (Joff Wild): "Whatever was decided once the first patent applications began to be assessed in which machines were specifically identified as inventors was going to matter, so in and of itself the EPO’s ruling is noteworthy. [...] Artificial intelligence is not going away, neither is invention by machine. Answers to the many IP questions this raises will have to be found – at legal, regulatory and strategic levels. And, of course, the subject is not confined to Europe, it is worldwide. If ever there was something for the major patent issuing authorities to be working on together, perhaps in concert with a WIPO expert group, this is it."

Sarah Morgan at World Intellectual Property Review wrote:

Late last week, the European Patent Office (EPO) refused two patent applications that list an artificial intelligence (AI) application as the sole inventor.

After hearing the arguments of the applicant, the EPO refused the European patent applications as they don’t meet the requirement that an inventor designated in the application has to be a human being, not a machine.

Dabus, the AI application that is named as inventor on both of the patents, is said to have designed a type of plastic food container and a flashing beacon light.


Law360, another outlet of patent maximalists, wrote this:



The European Patent Office announced Friday that it had rejected two patent applications that claimed the inventions at issue were created by artificial intelligence, ruling that under Europe’s patent laws, an inventor must be a human being.

The applications, which attracted worldwide attention when they were filed over the summer, marked the first-ever attempts to secure patent protection for inventions said to be entirely the work of artificial intelligence with no human involvement.

The EPO issued a brief statement announcing that it had refused the applications, which cover a beverage container and a flashing device to be used in search-and-rescue missions...


We loved this comment from "Do not pull my leg" -- an anonymous commenter who wrote that "artificial intelligence does appear no more than a hype, which will most probably end up like a deflated balloon. There is nothing intelligent in those machines whatever the applicant of both applications may say."

Here's the full comment:

Independently of the discussion relating to inventorship, it does not seem that the machine was so "intelligent" as the search in both cases has revealed very relevant documents.

In the case of the can, the mere connection of cans through their external profile is known. The only difference is that in the case of the application, the surface is a fractal surface. Whether this is inventive remains to be seen. As the application has been refused by the Receiving Section, we might never know.

As far as the light beacon is concerned, the whole invention seems entirely based on studies of the applicant himself. I would say if only the theory on which the applicant bases its application is proven that one could start believing what is going on. It would interesting if the applicant provides more than a “paper” invention and would show a real device working according to the claimed invention. To me this invention is nearing a substantial lack of sufficiency. As the application has been refused by the Receiving Section, we might never know.

What is striking as well, is that in both cases the notion of fractals come up. I do not think this is innocent.

When reading the explanations given about the way the invention was allegedly created, it is difficult to follow that “the machine was not trained on any special data relevant to present invention”, but a few lines higher is said that the machine. Either one or the other, but not both at the same time. The whole. A quick look at the references allegedly explaining the working of DABUS, at least US 5659666 has never crossed the Atlantic and US 7454388 has not led to a European Patent due to problems with Art 123(2). For the EP application corresponding to US 2015/0379394 summons to OP have been issued. Art 84 (if not Art 83) seems to be a major problem, so that we might also end up with problems under Art 123(2).

On the other hand, artificial intelligence does appear no more than a hype, which will most probably end up like a deflated balloon. There is nothing intelligent in those machines whatever the applicant of both applications may say.

They are only doing what they are told and if some self-perturbation of connection weights between neurons, like alleged in DABUS should all bring the desired result, this needs a bit more explanation.


This was posted in reply to a post from Rose Hughes of AstraZeneca, who filed this under "AI hype". So even patent maximalists realise it's nothing but marketing nonsense?

To Hughes' credit, she noted how (and where) they left a loophole:

It seems that the EPO can therefore neatly dismiss any need for detailed discussions on the morality of AI inventorship, so long as the AI inventor is not also claimed to be the owner of its creations. It has always been this Kat's view that the whole exercise of the AI inventor applications is slightly irrelevant given that what primarily matters from a legal standpoint is who owns an invention, and who should therefore be awarded the patent.

[...]

If Dr Thaler's appeal of the Receiving Office decision does indeed go ahead, IPKat will await the Statement of Grounds with interest (deadline for filing the appeal will be early 2020). Based on their submissions to the Receiving Office, it seems that the AI inventor team are lacking the legal arguments to overturn the decision. Moral and social arguments are unlikely to convince the Boards of Appeal. A request for a referral to the Enlarged Board can probably also be expected, but is similarly unlikely to be granted. None-the-less, the team have undoubtedly been successful in one goal, bringing attention to Dr Thaler and the team. Notably, we are only able to see the file history because early publication of the applications was requested. However, in this Kat's humble view, the whole argument surrounding AI inventorship is premature until the existence of an AI truly capable of a inventive act has been proved.


Appeal or no appeal, the person received plenty of publicity and could reapply with a real person's name. Maybe the whole thing was a premeditated publicity stunt all along.

What's at stake here isn't software patents, not directly anyway. When they use "HEY HI" to justify software patents it is a real issue however; sometimes not even that is necessary as massive corruption at EPO and an attack on the judges have both facilitated the granting of illegal patents such as patents on algorithms and only lawyers are happy, the litigation 'industry'. James Prankerd Smith (GJE) has just mentioned a European Patent pertaining to "algorithm [...] implemented on a computer" -- one that we named in passing before. The judges in this case aren't autonomous, so a real legal test would have to be outside the EPO:

This decision is certainly worth reading if you deal with inventive step objections of the form “abstract algorithm implemented on a generic computer” or the like. The Board of Appeal provides a helpful review of case law, and pushes back the frequent assumption that improved algorithms cannot give a technical effect. This decision could well be worth referencing when formulating a response to this type of objection.

The application relates to SQL database management, and claims a method of updating values in a data structure in a relational database system. This is a fairly classic case of an algorithm which could conceivably be abstract, but which is implemented on a computer in any modern industrial application.

The examining division originally refused the application under Article 52 EPC on the ground that it is directed to a purely abstract method. The Board of Appeal dismissed this ground, confirming the well-established low bar for patent-eligible subject-matter, with even “database system” being enough to give a technical element to the claim.

However, the examining division also sensibly set out an assessment of inventive step for an assumed computer-implemented version of the claimed method, in the expectation that the first hurdle of Article 52 EPC could potentially be overcome.

[...]

Although this was not a point discussed in the decision, it seems likely that it may be difficult to prove during prosecution that a particular feature has been motivated by technical considerations.

However, this can be significantly helped by identifying technical intentions and motivations at the drafting stage. In particular, by associating specific features with specific expected technical advantages in the specification as originally filed, the burden can be transferred onto the examiner to show that the feature lacks technical character.


For those who are in central Europe, this coming weekend there's a talk of interest. Benjamin Henrion has just mentioned "Discussion on Unitary Software Patents at 36C3 this Saturday" and the abstract says:

The Unitary Patent is the third attempt to impose software patents in Europe. Software patents are a threat to small- and medium-sized software companies that cannot defend themselves. The UPC (Unified Patent Court) is an international court made outside of the European Union (EU), which would have the last word over the question of software patenting. The Court would favour “patent trolls” which steal our jobs and extort money.


He recently organised a protest against this. We've seen no press reports however; does press coverage regarding patents still exist? Hardly. Many of the so-called 'articles' are composed by law firms directly or indirectly (the likes of Joff, their megaphone whom they pay to lobby).

Recent Techrights' Posts

Daniel Pocock elected on ANZAC Day and anniversary of Easter Rising (FSFE Fellowship)
Reprinted with permission from Daniel Pocock
Ulrike Uhlig & Debian, the $200,000 woman who quit
Reprinted with permission from disguised.work
Girlfriends, Sex, Prostitution & Debian at DebConf22, Prizren, Kosovo
Reprinted with permission from disguised.work
Martina Ferrari & Debian, DebConf room list: who sleeps with who?
Reprinted with permission from Daniel Pocock
 
Joerg (Ganneff) Jaspert, Dalbergschule Fulda & Debian Death threats
Reprinted with permission from disguised.work
Amber Heard, Junior Female Developers & Debian Embezzlement
Reprinted with permission from disguised.work
[Video] Time to Acknowledge Debian Has a Real Problem and This Problem Needs to be Solved
it would make sense to try to resolve conflicts and issues, not exacerbate these
[Video] IBM's Poor Results Reinforce the Idea of Mass Layoffs on the Way (Just Like at Microsoft)
it seems likely Red Hat layoffs are in the making
IRC Proceedings: Wednesday, April 24, 2024
IRC logs for Wednesday, April 24, 2024
Over at Tux Machines...
GNU/Linux news for the past day
Links 24/04/2024: Layoffs and Shutdowns at Microsoft, Apple Sales in China Have Collapsed
Links for the day
Sexism processing travel reimbursement
Reprinted with permission from disguised.work
Microsoft is Shutting Down Offices and Studios (Microsoft Layoffs Every Month This Year, Media Barely Mentions These)
Microsoft shutting down more offices (there have been layoffs every month this year)
Balkan women & Debian sexism, WeBoob leaks
Reprinted with permission from disguised.work
Links 24/04/2024: Advances in TikTok Ban, Microsoft Lacks Security Incentives (It Profits From Breaches)
Links for the day
Gemini Links 24/04/2024: People Returning to Gemlogs, Stateless Workstations
Links for the day
Meike Reichle & Debian Dating
Reprinted with permission from disguised.work
Europe Won't be Safe From Russia Until the Last Windows PC is Turned Off (or Switched to BSDs and GNU/Linux)
Lives are at stake
Over at Tux Machines...
GNU/Linux news for the past day
IRC Proceedings: Tuesday, April 23, 2024
IRC logs for Tuesday, April 23, 2024
[Meme] EPO: Breaking the Law as a Business Model
Total disregard for the EPO to sell more monopolies in Europe (to companies that are seldom European and in need of monopoly)
The EPO's Central Staff Committee (CSC) on New Ways of Working (NWoW) and “Bringing Teams Together” (BTT)
The latest publication from the Central Staff Committee (CSC)
Volunteers wanted: Unknown Suspects team
Reprinted with permission from Daniel Pocock
Debian trademark: where does the value come from?
Reprinted with permission from Daniel Pocock
Detecting suspicious transactions in the Wikimedia grants process
Reprinted with permission from Daniel Pocock
Links 23/04/2024: US Doubles Down on Patent Obviousness, North Korea Practices Nuclear Conflict
Links for the day
Stardust Nightclub Tragedy, Unlawful killing, Censorship & Debian Scapegoating
Reprinted with permission from Daniel Pocock
Gunnar Wolf & Debian Modern Slavery punishments
Reprinted with permission from Daniel Pocock
On DebConf and Debian 'Bedroom Nepotism' (Connected to Canonical, Red Hat, and Google)
Why the public must know suppressed facts (which women themselves are voicing concerns about; some men muzzle them to save face)
Several Years After Vista 11 Came Out Few People in Africa Use It, Its Relative Share Declines (People Delete It and Move to BSD/GNU/Linux?)
These trends are worth discussing
Canonical, Ubuntu & Debian DebConf19 Diversity Girls email
Reprinted with permission from disguised.work
Links 23/04/2024: Escalations Around Poland, Microsoft Shares Dumped
Links for the day
Gemini Links 23/04/2024: Offline PSP Media Player and OpenBSD on ThinkPad
Links for the day
Amaya Rodrigo Sastre, Holger Levsen & Debian DebConf6 fight
Reprinted with permission from disguised.work
DebConf8: who slept with who? Rooming list leaked
Reprinted with permission from disguised.work
Bruce Perens & Debian: swiping the Open Source trademark
Reprinted with permission from disguised.work
Ean Schuessler & Debian SPI OSI trademark disputes
Reprinted with permission from disguised.work
Windows in Sudan: From 99.15% to 2.12%
With conflict in Sudan, plus the occasional escalation/s, buying a laptop with Vista 11 isn't a high priority
Anatomy of a Cancel Mob Campaign
how they go about
[Meme] The 'Cancel Culture' and Its 'Hit List'
organisers are being contacted by the 'cancel mob'
Richard Stallman's Next Public Talk is on Friday, 17:30 in Córdoba (Spain), FSF Cannot Mention It
Any attempt to marginalise founders isn't unprecedented as a strategy
IRC Proceedings: Monday, April 22, 2024
IRC logs for Monday, April 22, 2024
Over at Tux Machines...
GNU/Linux news for the past day
Don't trust me. Trust the voters.
Reprinted with permission from Daniel Pocock
Chris Lamb & Debian demanded Ubuntu censor my blog
Reprinted with permission from disguised.work
Ean Schuessler, Branden Robinson & Debian SPI accounting crisis
Reprinted with permission from disguised.work
William Lee Irwin III, Michael Schultheiss & Debian, Oracle, Russian kernel scandal
Reprinted with permission from disguised.work