CalTech wins $1.1B jury verdict in patent case against Apple, Broadcom (opens in new tab)

The CRISPR debacle is a funny story because on one side you have a public institution and on the other side you have a private institution.

However, both in reality are publicly funded whether that be through government grants or tax breaks on endowments. I think we should do a trade: an institution that in someway benefits from the public should get credit for the discovery but patent goes back to the public.

Edit: I realize there’s a chance that whatever I just said, there’s a small European country that already has implemented this idea.

What’s crazy is universities don’t fund research. That’s a more central issue. They both directly profit from research and want to then own the results of that research, it’s a parasitic relationship with funding agencies.

See the motivation to fund CRISPR research for a counterpoint.

I bet that the money that they get will largely go to lawyers, administration, and increased efforts around finding/litigating patents on current research.

I am extremely dubious that it will result in research that advances the state of science. Science flourishes best when you share results and hypotheses early and often, then get feedback/inspire others. But that type of communication threatens patents, so inspires people to put their research into black boxes until the legal paperwork has been filed.

Did CalTech's filing of this patent help Apple or Broadcom's engineers create their devices? I sincerely doubt it. This is just patent trolling, and Caltech should be shamed for it.

> But in this case, a research institution is getting more money, presumably to be able to fund further research.

Yes, research that can't be used for reasonable amounts of money, and therefore out of reach for startups.

> If I create something novel, I deserve a short term monopoly so I can compete against established corporations.

I agree, but I don't think the current patent system does this. I think the length of patents should be reduced dramatically, 5 years sounds reasonable.

It's crazy when you think about how patents last for 20 years, which doesn't really work with the current pace of technology IMO (as an example, the iPhone wasn't released until 2007). So you can't use anything Apple patented to make the ORIGINAL iPhone work until... 2027? That doesn't seem good for humanity to me.

If the lengths were reduced, and the requirements much more strict, I think the world would be a better place.

- https://en.wikipedia.org/wiki/Term_of_patent

Their problem was that other engineers didn’t actually use the novel tech the brothers invented in their airplanes. The basic idea of putting an engine with a propeller on a winged vehicle was not novel, it had been tried many times before. It was simply a matter of power to weight ratios and aerodynamics knowledge crossing the right threshold.

Their key innovation was wing warping for control, but the industry rapidly switched to flaps and ailerons, leaving the brothers’ main innovation behind. Their aerodynamic concepts were advanced, but again others pushed beyond their concepts very quickly.

So their technical innovations were brilliant, but not used extensively by later engineers.

You realize the Wright Brothers in particular are an absolutely horrible example of the patent system working right?

Their patent trolling is credited with setting back the development of the United States aviation industry for years.

https://en.m.wikipedia.org/wiki/Wright_brothers_patent_war

So really, the only benefit is the actual patent ending up in the public domain, at the cost of further slowing down the field when someone ends up with patent rights assigned andisfeeling particularly litigous.

I agree somewhat. The problem is that there are so many bullshit patents in the world that mostly just create a minefield for those trying to create new things.

A proposal (in “Radical Markets”) I agree with is to charge a property tax of, say, 2% for each year that a patent holder wants to keep the patent. The value of the patent is whatever the owner decides it should be but (to prevent claims of $1 values) the owner must sell the patent to the first entity willing to pay more than the declared value.

This, in theory, would ensure that only patents that were genuinely useful would be kept around rather than being used as tools to stifle innovation or rent seek for those that actually create great works.

If you invent a new technology as part of a grant funded by my tax dollars, it's much less clear.

The problem seems to be that the bar for "novelty" in software patents is extremely low. (Amazon's patent on 1 click ordering is probably my favorite example).

> If an individual has a novel idea they should be entitled to some of the money that a corporation makes using that idea.

Absolutely not. Suppose I, Mr Independent Inventor, come up with some idea and convince the patent office to grant me a patent, and then proceed to do exactly nothing with it. Five years later, Ms Employed Engineer working at ABC Corp comes up with the same idea, and they manage to turn it into a product, which becomes successful. I contributed nothing to their product. Why on Earth am I owed any money for their work?

This used to be the case until the passage of the Bayh-Dole Act.

https://en.wikipedia.org/wiki/Bayh%E2%80%93Dole_Act

Patents and code licensing rights are not related. You may release your code under GPL and if it violates some patent, the owner may sue you or whoever uses your code.

how do we know trade secrets would be as effective today?

Thank you. You're correct, I've wondered why this isn't mentioned more often, and maybe that's why.

However even after realizing that, I personally still consider a design patent about rounded corners to be on the same level of B.S. as a software patent, and still stand by my conclusions. But I'd be interested to know why I shouldn't do that, and if you have more information I'm honestly all ears

They sure them for multitouch software as well didn't they?

The America Invents Act was signed into law in September 2011. This was exactly the time to do something about patent reform. There was a lot of discussion about it among tech companies at the time, but little leadership and no unified message on anything beyond, "Trolls are bad." Congress was not likely to (and did not) revisit the issue for some time afterwards.

True. But barriers to market entry is also a prong for establishing monopoly power. So it cuts both ways.

It’s honestly really sad that even _professional_ journalists don’t know how to spell it correctly :-(

Pretty much any case requiring non-trivial evaluation of evidence is unsuited for a jury trial. Outsourcing the adjudication of justice to a dozen random people is stupidity on a par with deciding complex policy issues by popular referendum. There's a reason the non-Anglo-American developed world doesn't do it this way.

That actually refers to criminal cases, not civil.

You make it sound easy, but the patent office is severely overworked. Also, does the case just “go away?” Or do you have to file a motion to dismiss on the grounds that the patent is invalid?

That’s highly nontrivial. (And by the way, almost no one uses reexam anymore. It’s all about IPR these days.)

IMO you shouldn’t need experts; The jury should (in the first place) consist of people who understand the issue. Not people who are swayed by buzzwords.

That's not how trolling works. Your out of pocket cost for patent prosecution is maybe $20K or $30K per. If you want to litigate, you enter into a contingency or mixed-fee arrangement with a plaintiff firm and possibly also a litigation financier, depending upon the size of your pocketbook and your appetite for risk. Your lawyers build up a war chest by shaking down easy targets until it's big enough that they can lay siege to a big opponent. It's become more difficult to do this than it used to be due to recent developments in patent law, but it's still eminently possible and happens all the time. This is, in fact, how plaintiff litigation works in a number of other civil case types.

Your last point has me thinking - hypothetically if an individual were in this situation, could they sell 'equity' in the eventual settlement in order to take money off the table now, rather than go bankrupt from legal expenses etc and get nothing?

E.g. Sell 50% of the initially ruled $1.1B settlement for, say, $100MM.

Anything like this ever happened?

They’ve managed to be broke twice in the last 20 years despite receiving three of the four largest donations in higher education history. I wouldn’t put it past them.

"Serial concatenation of interleaved convolutional codes forming turbo-like codes"

The popular observation (truthful joke?) on campus is that they could give a full ride to 15 consecutive classes of undergrads with all that money...

Yup, I think is is called patent infringement indemnification, but I'm guessing lawyers always sue everyone with deep pockets no matter what.

Then there would be a loophole where you make an independent entity that “makes” something that shields infinite liability. It makes sense to go after all beneficiary owners and parties.