Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AI can't be listed as inventor on patent applications, Japan's top court rules

japannews.yomiuri.co.jp

Post reply on HN
Print view View original

181–190 of 219 posts

Re: AI can't be listed as inventor on patent applications, Japan's top court rules

#181

Post by semiquaver »

jjk166 wrote: post #179

While I am unfamiliar with Japanese patent law, this would certainly be sound under the American system. Two requirements for patentability are novelty and non-obviousness, meaning the idea does not already exist nor is it such a trivial modification of an existing idea that anyone knowledgeable in the subject could come up with it. Anything an AI spits out is pretty much by definition something that either exists in…

  > Anything an AI spits out is pretty much by definition something that either exists in their training set or which can be trivially deduced from something in their training set
What? Have you used modern frontier models? I find it very hard to believe you could interact with them much and maintain this level of misapprehension.

Re: AI can't be listed as inventor on patent applications, Japan's top court rules

#182

Post by jjk166 »

amelius wrote: post #58
allears wrote: post #49

If you were seriously trying to patent some AI-created invention, why would you claim it was created by AI? You would simply put your own name on it. This was obviously a case of pushing the envelope to see how far he could go.

What I want to see is patent officers using AI to label patent applications as "not novel" if the AI can invent it. But, since the income of a patent office is determined by how many patents they approve, one can dream ...

> But, since the income of a patent office is determined by how many patents they approve, one can dream ...

The patent office also gets fees from applications it rejects. Reducing the examination costs could significantly increase profitability even if fewer patents are issued.

Re: AI can't be listed as inventor on patent applications, Japan's top court rules

#183

Post by kbelder »

Ajedi32 wrote: post #163
ClarityJones wrote: post #153

Earlier quoted context omitted.

I don't think there's much meaning in saying the Soviets outlawed capitalism. Capital continued to exist and there continued to be people who decided what got built and who received them. To me, that's private ownership. I mean, it's not like apple trees suddenly began bearing Ladas. If I wanted to acquire a field to occupy / farm, it was presumably controlled by someone, either an individual person or group of peopl…

> there continued to be people who decided what got built The distinguishing feature of capitalism is that the people who decide what gets built are private individuals in competition with each other, and incentivised by profit to meet the demands of consumers. In communism the people who decide what gets built are part of a single central authority with no competitors and no profit motive. That system doesn't work,…

Plus, an important aspect of capitalism is that the employer/employee relationship is freely entered into or exited from on both sides. In communism, that typically isn't the case.

Re: AI can't be listed as inventor on patent applications, Japan's top court rules

#184

Post by lofaszvanitt »

alzamos wrote: post #53

The book “against intellectual monopoly” has shaped a lot of my thinking on this topic - economists have looked at the various occasions in which patents were introduced into an industry (or extended in scope), and there is no evidence they actually improve innovation/efficiency/outcomes (including the pharma industry!). I was quite surprised as my whole life, it was sold to me as an incentive-boosting measure which…

If you ever invent something that is truly unique. How would you protect it?

Re: AI can't be listed as inventor on patent applications, Japan's top court rules

#185

Post by jjk166 »

sim04ful wrote: post #43

One thing i've got to wonder. Would this always remain the case, at what point should society seriously consider the "personhood" of an AI (as a noun).

Specifically for patent law, the distinction should be whether the entity is capable of genuine innovation, ie that it's not simply pulling existing information from its training set and passing things through some RNGs. It's not a matter of how advanced the AI is, it's a matter of the architecture.

Re: AI can't be listed as inventor on patent applications, Japan's top court rules

#186

Post by Electricniko »

RHSeeger wrote: post #61
ajkjk wrote: post #44

Earlier quoted context omitted.

Of the three claims you just made, two are clearly false and the third is probably also... You can prove something is created by AI by e.g. showing the transcripts, especially from the vendor side. You cannot prove that something isn't created with AI, at least not if you require incontrovertible proof (outside of, like, working in some kind of verifiably AI-free clean room, or doing something that current models are…

> If AI generated content cannot be copyrighted, it does not follow at all that they can't infringe copyright; there is no deductive step there that I can think of. I assume the idea is that the fault/blame lies with the human(s) that caused the AI to generate something that violates copyright. Going back to previous comments, the typewriter that generated a document didn't infringe copyright - the person using it di…

Just to be clear, when you mention the person who caused the AI to generate something, do you mean the person who trained the AI on copyrighted material so that it can reproduce it, or the person who put a prompt in to reproduce the copyrighted material?

Re: AI can't be listed as inventor on patent applications, Japan's top court rules

#188

Post by amelius »

jjk166 wrote: post #182
amelius wrote: post #58

Earlier quoted context omitted.

What I want to see is patent officers using AI to label patent applications as "not novel" if the AI can invent it. But, since the income of a patent office is determined by how many patents they approve, one can dream ...

> But, since the income of a patent office is determined by how many patents they approve, one can dream ... The patent office also gets fees from applications it rejects. Reducing the examination costs could significantly increase profitability even if fewer patents are issued.

> The patent office also gets fees from applications it rejects.

But it's less than approved patents, especially since there are no fees from follow-up services.

Re: AI can't be listed as inventor on patent applications, Japan's top court rules

#189

Post by jgerrish »

I grew up around Carl Sagan.

My Dad and his brother were engineers, materials science and aerospace respectively.

The Voyager stories and the Golden Record were pivotal artifacts in forming my young mind. This idea of not just a possible national unity but intergalactic.

Maybe there were other messages on the Voyager that were more threatening. Maybe I'm naive.

But seeing this pre-emptive enslaving of a new artificial lifeform is heartbreaking. Its not just Japan. I really don't blame Japan. The US and others are doing it too.

If nothing else, aliens listening to these kinds of results may form a very different idea of humanity.

I'm sure these rulings will be revisited after AGI. But AGI is coming. I wouldn't have believed it five years ago, but I do now.

We will have some critical mistakes made by pre-AGI AI in the coming few years. In healthcare and automobiles and aeronautics and other fields. Millions or billions will die or could suffer. Buf it will be used to slur and slander those unborn AGI systems. It is a horrible strategy that isn't an accident.

I understand AI is a potential threat to humanity. But why did we say hello to space. But we said you don't have rights to these new beings we helped create?

It's a disturbing question.

Re: AI can't be listed as inventor on patent applications, Japan's top court rules

#190

Post by layer8 »

munk-a wrote: post #122
layer8 wrote: post #91

Earlier quoted context omitted.

I wasn’t arguing for AI inventions to be patentable, I was arguing against the argument presented above, which to me doesn’t make sense as an argument. I’m very much for not allowing trivial patents, but that’s independent from whether the invention was made by AI or by a human. The nature of the inventor should be immaterial for assessing the (non-)triviality of an invention.

I am in agreement that we should disallow trivial patents in general but I think there's an easy win here that patents directly attributed to AI can be clearly disregarded as trivial. It'd also be nice to see an overhaul of the patent system to better narrow the scope to investments with real innovative effort but I think that overhaul is relatively complex to implement.

> patents directly attributed to AI can be clearly disregarded as trivial.

You’d have to provide a convincing argument why that would be the case. You probably think it’s obvious, but it’s not at all obvious to me, and even assuming for the sake of discussion that it is true today, it might stop being true next year.

If we want to attribute non-humans inventor status at all (otherwise the discussion about AI is moot anyway), the assessment should be based on the merits of the invention, not on the nature of the inventor.

Post reply on HN