Your concerns and suggestions are all completely valid and I am completely in favour. That’s said, I’m not sure if you’ve written or analysed any pareents before.

Aside from the suggestion of the reduced exploitation limit, the others are very difficult in practise to realize.

The biggest problem is that patents don’t actually effectively protect anything today. The underlying causes as I see it are:

1. Technology has become far too advanced that it is extremely difficult for any review process to be sufficiently thorough in a practical time frame: A patent needs to be “new”, novel. It’s becoming very difficult to prove that any invention is genuinely new. An imperfect analogy is colours. How many new colours can you find up with? Is this black or that pink really new? The concept of pink or black has been around for quite a while.

2. Proven product in the marketplace: very rarely if any, does a patent translate into a product in the market. A patent is a ‘method’. Not a design (which is a copyright). An example of an indirectly commercial method might be a faster or more accurate way to measure temperature in a production process that is not yet implemented (e.g next gen ASML technology). For a new technology line say, a 2nm semicon fab, the technology is broken down into its fundamental pieces and patented separately to

2a. Make it practical to review and patent specifically (a patent must be specific). Otherwise the tech for a 2nm line will simply be thousands of pages long with thousands upon thousands of individual claims. No one would ever read or assess that.

2b. Distribute (protection) risk. ASML isn’t the only semicon tech company. It could be that similar technology is already existing that overlaps slightly. Instead of risking that the entire patent be denied due to a small overlap, only a few overlapping patents can be taken off and the rest protected, without needing to reinvest the very significant resources involved in rewriting the whole thing. This also illustrates why parents have to be specific. Otherwise it becomes a little subjective as to how much one patent infringes upon another.

3. Patent trolls are imho a (visible) exception to the system. They need to be sorted out post haste. But then again, the line between patent trolls and strategic holding of patents can be fuzzy: eg. ASML could be developing a roadmap for a 1.5nm process but finds this patent or two from elsewhere for another industry that is pretty key to their upcoming tech and fears it may become a weak point and they can’t find a good alternative method. They may choose to buy said patent and company and close it to hold the patent (before someone else buys it and uses it to get to 1.5nm first) to ensure their 1.5nm process can be sufficiently covered. This might look like trolling behavior because ASML isn’t going into whichever industry they nabbed the tech from.

4. A patent by definition requires that the secrets be shared. I have a huge problem with this as well, because the language used is very unnecessarily verbose. (Similar to T&C’s). A patent requires that a person ‘sufficiently skilled in the art can reproduce the outcome of the claims given the information in the patent’. It’s just that most of the time, the general public (or reviewers) won’t have easy access to whatever they are talking about. ‘method to ensure femtosecond laser periodicity falls within target parameters with 8 sigma confidence’

5. “Nebulous concept” I like this, because this really grabs at what I also think is the problem with patents today. Patents today are treated as ammunition. Several companies can hold patents for a folding phone. Because there’s more than one way to fold a phone. Now they can fight in court. But to the public, a folding phone is a folding phone. Technically. There can be any number of hinges to achieve the same function.

Even if only one hinge was possible, because of the required specificity of patent claims, another patent writer could side skirt the coverage by introducing an intermediary gear and say that it makes the opening smoother or whatever.

A patent must be:

• a method

• novel

• specific

• not be obvious to someone trained in the art (it can’t be something that is intuitive to an expert)

• some other requirements I forget, I invite you to look it up. I won’t post a link as they are plentiful and I don’t want to colour your source.

Edit: perhaps most importantly, a patent must solve a problem.

Because patents have become:

• difficult to properly review (either by office or ‘peers’)

• don’t effectively protect an invention because it’s usually possible to sideskirt them

• become far too verbose in their language

• technology today is too entangled to separate coverage, dependency and hierarchy (it works if everyone has only one ancestor but we have many, and share ancestry with other unrelated people) - even though there’s a runout limit of 20 (or 10) years

• complexity is exploited (key example is medicine, changing this functional group or that process to improve some miniscule performance or other to protect the entire medicine)

It’s very difficult to have the current (or new, overhauled) system do what it was originally intended to do: protect a novel invention for a limited period of exclusivity: it can’t be effectively protected; is rarely truly novel; limited period can be extended with loop holes.

I think the one thing to change that could go far is similar to what’s been done in the material safety space for producing SDSes. Previously with EU safety requirements, the regulatory board assessed the safety of each new chemical or material. But today. Every day there are new filings. To take the load off, the assessment of safety (!) is today offloaded to the filing company and also opened to everyone in the registry (iirc free to become a member as a company) to review the safety of. After some period, the SDS is filed (with blanks where safety data is not available or relevant). This could be done with patents - just throw it out into the open for critical review by everyone in the registered industry. It would be in their interest to critically review and honestly challenge the filing or forever hold their peace. There could be problems as well with this of course, especially with regards to fair use and not using the pre-filing to hasten their internal RnD.