I don't think the only, or even primary, value of patents is to encourage invention. I think they're valuable for allowing a separation of concerns between inventors and implementors. When you think of patents in that way, it makes total sense for them to be treated as transferable assets.

I think it might be -

"To promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries."

Nothing there about it being a transferable asset to be treated as property. Is an exclusive right for a limited time for authors and inventors.

[edit] Otherwise it would presumably read: "by securing for limited Times to Companies and Investors the exclusive Right to the respective Writings and Discoveries of Authors and Inventors." Which doesn't scan nearly as well.

-- What happens when somebody wants to buy your startup?

The model I am thinking of is that if the stated intention of the monopoly is for the benefit of inventors as people, then for one thing, those individual inventors are the best qualified to decide if they want to continue the monopoly, which does increase risk for companies, however we are talking about patents and they are supposedly there to empower inventors rather than corporations and this incentivises corporations to be a lot nicer to their employees.

Also, the originating inventor should also always be in a position to negotiate fees for their patent from parties that licence it, including their own employer, this however should be subject to fair dealing laws with those they choose to licence to so they can't suddenly price out a competitor that they have already chosen to licence to or bankrupt their employer or anything like that.

You are taking away options from the creator, most notably the option to sell his stuff. Please don't call that giving him options.

the originating inventor should also always be in a position to negotiate fees for their patent from parties that licence it, including their own employer

There are usually multiple inventors. If 3 people invent it, do each of them get a 1/3 vote on it? Do all 3 need to agree on a license? Do each of them get 1/3 of the claims? Can any of the 3 license it individually (which means that the price will naturally fall to $0 since you can always take your business to one of the other two)?

You are creating a brand new problem that was solved long ago. The solution is for those 3 inventors to assign or sell their invention to a company that they form. The company owns and licenses the patent, and those 3 inventors can sell or trade their shares in the company as they wish.

This desire to interfere in trade is very odd. If I have something worth $100 to me, but $500 to my neighbor, the ability to trade creates wealth. I'll sell it to my neighbor (a price of anywhere from $100 to $500 is a Pareto improvement) and then we can each move on with our lives, instead of trying to work out some license or lease agreement.

I am removing the option to sell rights, not stuff. In other words I am calling for the rights to be statutory and non-transferable.

Removing the option of rights being wholly transferred to another party is the removal of the option to remove all future options and therefore, in absolute terms, can actually have the practical effect of giving inventors more access to other options over the lifetime of their patent.

This desire to interfere in trade is very odd.

Given that patents are an issued right to interfere in trade, I see nothing odd in interfering in the right to trade a right to interfere in trade.

[edit] I completely agree that this creates a different set of messyness, but I think that patents by their very nature create messyness and that this set of messyness is smaller and more dealable with than the current insanity.

Also, I would have to say I am far from convinced of the general need for patents at all, although I would definitely concede that they have been useful in some specific instances.

If A is sub-optimal and underpeforming its theoretical envelope how do we A/B test?