Hacker News Reader

Comment by Animats | Hacker News Reader

I'm always amused at the Silicon Valley attitude towards "patent trolls". Figuring out some way to make people slave for below minimum wage - fine. Obtaining a monopoly and then raising prices - fine. Using a monopoly in one area, such as an app store, to keep out competitors for your own products - fine. Shipping crap that doesn't work - fine. But enforcing patent rights - that's bad.

View on HN

Wow. Straw man galore.

Edit: feel like I should expand one this.

None of the opinions you mentioned are consistent across HN, nor are they supported by all of the same people. Not even a majority of HNers support most of those opinions. In fact, I think the only one of those that HNers do tend to support in majority is "But enforcing patent rights - that's bad."

I don't think even the majority of agree that enforcing patents is wholly bad. It's a bit more nuanced than that. Some of my personal problems with patent enforcement of the top of my head: (1) it often isn't encouraging creators to keep creating (the patents being enforced have been sold or assigned away from the individual) (2) patents are being used as a stock pile for insurance against other companies suing them (3) non-practicing entities (patent trolls) are generally the ones bringing law suits (4) many patents we hear about appear low quality (obvious, prior art, do it on a computer/network, etc.)

After seeing issues like the above over and over again it feels like the system isn't working to do what it should. I'd love to see patent enforcement to help fund the work of inventors and push the overall state-of-the-art forward, but it really feels like that's not what's happening right now.

All of the "problems" you state are valid, but targeting patent enforcement to fix the problem is like trying to stiffle a food poisoning epidemic by changing the way people eat instead of fixing the source of the problem which is in the food prep.

Patent trolls are succesful because of the millions of worthless patents issued (i.e. scope that actually does overlap prior art, such that people think they have broader protection than they actually have). The other reason is that people are filing patents (and getting them granted) based on concepts that seem viable in principle, but which they haven't actually distilled to practice (ahem - Magic Leap).

All of these issues can be greatly mitigated by reinstituting (1) the rule for inventors to actually demonstrate the technology they're patenting to the USPTO.

1. https://www.uspto.gov/custom-page/inventors-eye-what-are-cur...

You're equating HN and Silicon Valley there, and I'm not sure that's right. I also don't think the "Silicon Valley attitude" is the opinion of every person who lives in the area, just those who run it.

For all the complaints we see on HN about, say, Uber, it's still very much on top in SV and held up as a huge success story and model for other companies.

For the record, I think the OP was being hyperbolic. But there is an element of truth: Silicon Valley (rightly) criticises the things that hold it back, but is largely quite happy to stay silent about things it takes advantage of.

> enforcing patent rights

Software on patents were not a thing to begin with, if you were around 20-30 years ago. It's a recent invention. So yeah, it's totally worth challenging, especially the super broad patents used by patent trolls.

Correct and one of the key causes is an understaffed and under trained USPTO staff. Even if you could train them to understand the (constantly changing) state of the art in ML learning, it could take years to truly determine if there was prior art or not.

I still think the way to solve the patent issue for software is that you make it's time limited to 5 years and make them non transferrable. Unlike patents on physical things, it is much easier to make something similiar that just works around a single one of the claims. Another thing we probably need is that any patent dispute that goes to court requires a prior art check and "novel" reassessment and if it is found that the patent was never novel or prior art existed, then the patent is thrown out. This could be done as part of the USPTO office instead of the courts as a precursor to actual trial.

Sure it was. I held software patents 30 years ago. The first software patent was for SyncSort (1971), the first large sort algorithm to beat O(N log N).

The funding to fight against patent trolls isn't based on morality, it's based on profitability.

It happens to be the case that patent trolls are also immoral, but don't let that confuse you.

Patent trolls are extreme actors in the patent system. The spirit of patent laws is that it rewards & protect innovation. However the letter of patent law is open to interpretation and unfortunately, abuse. A patent troll, doesn't invent, invest or build. It spends all it's time attempting tot retroactively claim ownership of innovation, through acquisition of junk patents, building portfolios of them and using them to, effectively, extort companies.

The topic at hand isn't the general idea of 'patent rights': it's the specific case of patent trolls. Even if the things you mention were all really accepted and supported by HN/SV, I would still claim that patent trolls are a more clearly negative, less defensible, phenomenon.

I don't think anyone is actually against enforcing patent rights. The problem is so many patents involve obvious ways of doing things.

The explanation is that might is right and Silicon Valley has power.

Reply on news.ycombinator.com