The patent office is about to make bad patents untouchable
eff.org
eff.org
Here is the new Groklaw mission statement:
Our Mission
Our mission is simple: to guide you toward safe, rewarding, and responsible crypto gambling experiences. We believe in transparency, player protection, and giving you the tools to make informed choices — whether you want massive Bitcoin bonuses, ultra-fast withdrawals, or niche altcoin gaming.
One marketer I know made his fortune (as in tens of millions) buying up dead sites, often for dissolved companies / organizations / etc., kept the design somewhat relevant to said entities, but filled the site with ads and referral links to gambling sites / credit cards / etc.
And as seen with Groklaw, it still works. Whenever I see some old semi-popular website call it quits, it just takes a couple of months until it becomes a landing or referral page for something crypto.
It's what happened with Adventure Gamers, for example. Decades worth of adventure game reviews and interviews still available -- now with a slight thematic shift where new articles are all about "reviewing" online casinos.
Every website eventually becomes a funnel for ad-based shadow companies.
"everything is so bad it's hopeless" or smth?
So much could be intercepted back then because of this. It wasn't until 2010 that various large services—including Yahoo Mail and Facebook—got a kick in their ass by a whitehat browser plugin that allowed anyone on the same network to trivially hijack session cookies of others, stimulating an adoption of HTTPS[1] during 2011-2012.
By the time the Snowden leaks occurred in 2013 the trend was heading toward encrypted-by-default and governments were having to adapt.
[1] https://threatpost.com/facebook-kills-firesheep-new-secure-b...
How would that actually work?
TLS runs on the client and the server. There's no "TLS magic box" in between.
Like let’s say you have a proxy server like Nginx on a server with a public facing ip address and then it also has access to a private subnet where your application servers are running. A visitor to your website’s browser make a secure https connection the nginx server where https would be terminated and then it would proxy traffic in plain http over your internal private subnet to the app server. And your are in a five eyes country where your intelligence services took it on themselves to follow the nsa or fbis instructions and plug a network device into those private subnets of all the big service providers inside their datacenters that is configured in something like a promiscuous way so it receives all the packets for any device on the network. Then those packets somehow end up in a big nsa datalake.. or something along those lines
(Or was, back then. These days you can probably collapse all of that into a single medium-sized epyc or something.)
I know where there are Sun V880s still running Oracle databases in a biggish cluster.
Their processor power, memory capacity, and storage capacity are exactly equivalent to a Raspberry Pi 4 with a biggish SD card.
We have come a long way.
If there was it'd be called Cloudflare :)
Let's not even forget the original motivation for patents, which was to increase knowledge sharing so that companies didn't keep their technology secret, and conferring an advantage on companies who chose to share their methods for others to copy and build upon. That's clearly a totally different outcome to when patent trolls are suing people for accidentally infringing a patent they didn't even know existed.
- measuring how well a patent office works by how many patents they grant (their purpose is to reject bad patents, it's 100% the opposite of the metrics often used)
- a standard where patents are often written with the most obtuse complex language as multi layered onions, instead of clear straight forward language
- no relevant consequences for most companies systematically deceiving the patent office about what they actually try to patent using artificial complex language, systematically fail to do the research into prior art and "happen" to overlook all prior art all the time
- patents require a certain level of innovation _in the aspects they patent_, but in practice patent offices rarely do enforce this. A solution anyone can trivially find if they run into similar "normal" problems isn't innovation. The abstract concept of applying a concept from IRL onto games (or similar) isn't innovation. A minor tweak(1) or trivial re-combinations of existing patents isn't innovation either. etc. etc.
- the patent office tends to be severely under-stocked, especially in context of AI making it easier to mass generate nonsense patents and non-clear-straight-forward language being the norm
- allowing people to not (relevantly) use their patent and still sue others for using it (which is directly harming innovation and really shouldn't be allowed)
- allowing absurd prices for usage permissions of patents (outside of certain especially fundamental/core patents, forgot how they are called), everyone should be able to buy _any_ patents usage right for a _reasonable_ price. Patents are for making sure companies doing innovation can re-coupe their investments, not for preventing marked competition and enforcing quasi-monopolies
(1): Like a minor tweak which doesn't require tons of research, don't make a fundamental difference etc. Like think
99% of modern software is complete shit, deserves no "my idea" protection, and will definitely not get any better with a higher barrier to competition. The real innovation these days is to make software that is usable, performant, and actually fucking works - and barely any company has the motivation and technical capability to achieve that. And when they do, as rare as that is, that simply cannot be copied.
Patent trolls hurt tech so thats now a good thing on new HN (now filled with normies like most tech companies these days). The enemy of my enemy is my friend sort of thing.
Nobody has any real values or beliefs anymore. We’re all just swimming in vibes.
Have you heard of the concept called projection?
Lamenting that everything is about AI, seems to be the thing of today.
If it was up to me I would probably eliminate patent enforcements against small entities completely. If they grow over some certain size then sure, let 2 big gorillas fight each other
https://www.sportskeeda.com/mmo/news-nintendo-vs-palworld-po...
From the perspective of policy makers, the biggest advantage of the modern patent system was to give US companies a leg up over foreign competitors, as we had the largest patent portfolios, better R&D, and legal frameworks we could use to our advantage. But this no longer works, as our biggest competitor understandably decided this wasn’t a fair bargain and has ignored IP rules to their own benefit. (Why nobody saw this coming is a mystery.) IMHO most patents now harm the economy more than they help.
Patents are a form of business insurance you can get on expensive R&D work. When you go into the market to sell your products, your marketing material and the function of your product naturally leaks how it works. Motivated competitors can then reverse-engineer what you did or otherwise figure out how to enter the market that you revealed was valuable. If they do it by actually out-innovating you, your patents don't really protect you, but if they just take what you're doing, the patent is there so you can claim the fruits of your labor regardless. In general, you cash in on this insurance contract by either contracting with lawyers to sue the infringers (usually on contingency or with litigation financing to remove the cash cost) or by selling your patent portfolio to someone who will sue the infringers.
In return for this insurance contract, you have to publicly disclose the details of your inventions in a way that a motivated party can read and understand (it's patentese, not English, but you can decipher it if you know the language). It also frees you up to publish internal details through other fora like scientific journals because this disclosure becomes pure upside (raising your company's reputation) rather than making you balance that upside against the downside of revealing information. Google is a good example of a firm that does a lot of balancing there: they publish a lot of their old, antiquated work while keeping the new stuff secret because it's mostly software and software is mostly unpatentable. In electrical engineering, small companies are much more likely to publish a lot more details on their newest devices because their devices are usually patented.
On HN, people mostly think about software patents. Many of these are stupid patents, and almost all of them have been invalidated through a decision called Alice Corp. vs CLS Bank. The digital shopping cart is one of these patents that Alice invalidated. By value, most patent litigation is about drugs, and after that you have things like computer hardware. The patent system really isn't for the HN crowd, and it really doesn't make sense for these software patents to be a thing - there's usually no expensive R&D to insure, and you have to write your patent before finding out if you have PMF.
The trouble is there are ways to nominally do the first one without really doing anybody any good.
You invent a way to improve efficiency by 10 points, they need to do the same to be competitive, so they come up with a different, incompatible way to do the same thing. Now your product isn't any better than theirs so you can't charge higher prices or increase your market share any more than you could without the patent. But now we've collectively had to pay to invent the wheel twice in exchange for no benefit.
Worse, suppose that the different ways of improving the product are compatible with each other. So now your product is 10 points better than what's in the public domain, but Megacorp's product is 20 points better because they invented two things. If you could combine all of them together it would be 30 points better and give you an advantage over them, but you can't use their patents. And you also can't profit from selling your product yourself, because 10 is less than 20 so everybody wants theirs instead of yours. Meanwhile they don't really need your patent for the same reason since they're going to get the customers either way, so they have you over a barrel in negotiations.
Then they get to underpay to buy your patent because they don't actually need it, but selling it is your only option to get anything out of it. Which in turn makes their product 30 points better than what's in the public domain and makes it even worse for the next little guy, who now can't compete even if they made a contribution worth 25 points.
> Like the legal standards are easier for the patentholder at USPTO or something like that?
It is much easier to get a patent than it is to use that patent successfully in legal proceedings. The bar that you have to clear to get it through the USPTO has never been particularly high.
So this is kinda, if you squint at it a little, just writing down what the status quo already is.
(As sibling mentions - if it did work, of course it's way cheaper to nip it in the bud earlier. You just can't really guarantee that you'll get enough expert eyes to get a reasonable decision out of the USPTO)
I would have to ask more to be able to explain more, but that's how things have been explained to me growing up.
Oh yeah, there are so many patents with obvious prior art, or that are so much "the obvious way to do it" (to anyone with expertise in the field) that they're inherently unpatentable, but which get through the US Patent and Trademark Office because the reviewers have no experience in software development, so they get fooled by language designed to conceal what the patent is really about.
I still remember a case from the early 2000's or perhaps late 1990's, which I can't dig up now since I don't remember enough keywords to get a search engine to find it for me. So no link, sorry, you'll have to rely on my could-be-missing-a-detail memory. But I recall reading a patent and realizing that under the complex language, it was essentially patenting scrollbars, which at that point had been around for years. But to someone who didn't understand programming terminology, they might have thought this was a novel idea, rather than a snowjob designed to hide the fact that prior art on this patent submission had existed for over a decade, making the patent submission invalid on its face.
There have been other patents which, when I read them, I realized were patenting something that had lots of prior art, but I don't remember them. Maybe someone has more examples. But that patent that boiled down to "clicking a mouse on a designated area to scroll a page" stuck in my mind for years. (At least two decades, in fact).
Yea that's exactly what I mean
One solution would potentially be to increase the USPTO budget by a lot. But... nobody does that I guess. So here we are.
And if we reduced the length of patent validity, then invalid patenta would expire sooner and wouldn't be quite as much of an issue.
There are large corporations (IBM, Qualcomm, Nokia, etc...) lobbying Congress and the Senate to restore software patents via the PERA bill, while the lobbying from small and medium software companies is very low.
A successful IPR will most likely need multiple forms of prior art. Each prior art must cover every independent and dependent claim of the patent in question, either through anticipation or combined obviousness.
Typical infringement contentions will not cover every claim and are therefore easier to defend.
The same thing happened with 3D printing a while ago. It only took off after the patents expired.
Patents are a pest. They're just another mechanism to pump money from below to above.
Patents are a way to make sure inventors are getting compensated for their R&D work and risk.
I do agree with your observation though - IMHO, the "exclusivity" period of a patent should last five years, and for the 15 years after that, patent holders should be mandated to license out their patent at reasonable pricing.
Hopefully it would come up with the patented idea and thus 'prove' it is obvious and thus not able to be patented. Then you could make different vintage LLMs and basically spam them at trolls to invalidate the patents.
...just a thought from a lurker
Would love to see it!
https://www.federalregister.gov/documents/2025/10/17/2025-19...
It would be nice to have some confirmation, but I'm assuming there was an extension.
Given that EFF and the comment form at regulations.gov cite Dec 2nd.
In any case, I'm filing my comment now, and encouraging others to do so as well.
Anyone who's ever been through any kind of patent process should understand just how egregious this is...
I think a better UI design would be to show that info before you press the button, but my main point is that they are apparently still accepting comments.
"Do not submit personally identifiable information through this form."