Buzzwords are used to disguise patents on algorithms, but in-depth analysis would expose them for what they really are

Pendulum does not swing back; buzzwords just move back and forth

Summary: The dawn of ۤ 101/Alice, which in principle eliminates almost every software patent, means that applicants find themselves having to utilise loopholes to fool examiners, but that's unlikely to impress judges (if they ever come to assessing these patents)

THE USPTO will continue to grant software patents in the foreseeable future, but that does not mean that these patents will be able to cause much damage. Why not? As we shall show later today and tomorrow, PTAB smacks down many of these patents. It's an invaluable mechanism of quality control, akin to oppositions and appeals at the EPO.

One might ask, "why are software patents granted after Alice?"

The answer is simple. There are tricks. The EPO and other patent offices too have tricks. Those are usually designed to bypass examiners' guidelines -- the sorts of guidelines that matter a lot less to courts which assess past court cases and underlying evidence, such as prior art and expert testimonies. Knowing that the courts are hostile towards software patents, many potential plaintiffs (patent holders) will not even bother suing. And that's a good thing.

This post concerns few of the aforementioned tricks, which exploit loopholes. Many of them are nowadays buzzwords, which help dodge ۤ 101/Alice (at least at a superficial level). At the EPO they like to use terms like "technical effect" or "device", but in the USPTO it looks like "Artificial Intelligence" (AI) is currently one of the favourites because the corporate media resurrected that hype. Almost any algorithm can be framed as "AI" as it's a rather nebulous concept. We previously wrote many articles about other buzzwords, such as "cloud", not to mention the old "over the Internet", "on a computer" and so on.

Finnegan, Henderson, Farabow, Garrett & Dunner LLP, a very large law firm, is still all about buzzwords in patents. Without even delving into the underlying granularities, the headline alone is rather telling: blah blah blah Artificial Intelligence blah blah.

Wow. Must be innovative because "AI" is supposedly "hot"! Granted! Yesterday Watchtroll wrote about passage of some patents in the "self-driving space," arguing that it "delivers on Didi’s commitment to invest in artificial intelligence capacity."

Whatever!

I already wrote some algorithms related to this (self-driving tools) and the only "AI" in it tends to be some classifier trained on an image set to help segment an unseen image (or long sequence thereof). That's hardly innovative. It could be made to work several decades agp and in fact there were working implementations a long time ago; they just lacked sufficient computing power.

Here's what Finnegan says in relation to "AI" and ۤ 101:

In addition to €§ 101 concerns, AI in medicine raises questions of inventorship and ownership in patent law. The US patent system only recognizes individuals as inventors,38 not companies39 or machines.40 But with AI, it may be the machine that is taking the inventive leap, not the human programmer. Recently, both Google and Facebook have seen AI develop its own language to perform the assigned tasks, eschewing known languages in favor of a more efficient means of communication.41 As the use of AI grows in medicine and the life sciences, it is more and more likely that the AI will be the entity taking the inventive step, drawing new conclusions between the observed and the unknown. Indeed, current AI systems develop their own code as a result of the system’s training.42 If that is the case, the United States Patent and Trademark Office (USPTO) and the courts will have to decide whether the current Patent Act encompasses computer-based inventors, and if not, who among the humans responsible for the AI should be considered an inventor.43 The list of possible human inventors includes the AI software and hardware developers, the medical professionals or experts who provided the data set with known values or otherwise provided input into the development of the AI, and/or those who reviewed the AI results and recognized that an invention had been made.

Examiners ought to be reminded that "AI" just means algorithms and patents on algorithms are annulled by ۤ 101. Here's an example of computer vision patents that have just been granted by the USPTO. This article says: "The last patent includes foreground motion detection in compressed video data with software that can tell the difference between background and foreground features in compressed video streams."

That's pure software. Surely they know these are worthless after Alice? Or maybe they delude themselves into thinking otherwise? In relation to an Olympian called Vincent Zhou there was coverage some days ago that said: "One is a 28-year-old from a blue-collar home in Scranton, Pennsylvania. The other is a 17-year-old son of Chinese immigrants, two computer scientists, who hails from California."

"She owns numerous software patents," it said further down. Well, too bad they're worthless now, eh? Here's another new example of patents on software, this time from LINE. Again, these patents are worthless after Alice. Why are they being granted? As we shall show in a separate article, few grants are even being challenged; those that do typically perish (PTAB overturning examiners' determinations).

Here's another software patent. "GBOX develops all software both inhouse and with international subsidiaries," says the release, "and has been awarded 5 provisional patents for its technology."

How many of them (if any) are even worth anything?

"With Valentine's Day upon us, one would rightly suspect that there is already an abundance of patents and patent applications related to online dating software," lawyers' media said some days ago. But software patents are worthless now. They themselves call it "software". Do they conveniently overlook the issue? Don't they try to disguise it by calling it something like "technology"?

"Blockchain" is another term that we often see used in relation to software patents. That's just a tired new loophole that software patents proponents love to exploit. It's an algorithm. And watch the China envy:

China is leading the world in blockchain patents: incoPat published the 2017 Global Blockchain Patent Ranking (top 100) applications for invention-, utility- and design-patents. See: http://www.iprdaily.cn/news_18252.html pic.twitter.com/DZLTnkuXdw

Well, China -- unlike the US -- actually permits software patents, so there might be nothing wrong about this. There's something wrong with the policy, sure, but not with the application thereof.

For the record, we're not against patents that aren't on algorithms. We're very picky in selecting what to criticse. Here, for instance, is a press release about a new patent settlement over bar code readers (not software, no problem). It says:

Honeywell (NYSE: HON) today announced that it has reached a settlement with Code Corp., a company that manufactures bar code readers, to settle Honeywell's claims that Code infringed certain Honeywell patents related to bar code scanning technology.

The scanning techniques tend to involve sensory aspects that are hardware-side, not software-side heuristics. The projection and reflection of infrared lights for instance.

Thankfully, as time goes by we see fewer software patents slipping through the sieve. Does that mean that the USPTO will stop granting software patents altogether one day? We doubt it. But the number of lawsuits over algorithms will decline sharply unless something radical happens (like PTAB getting squashed). ⬆