Back in 2013, I wrote a stubby post to create a link to audio of an interview with physicist Richard P. Feynman, in which he describes how he came to be named as inventor on U.S. patents. Since that post is getting a number of readers recently, I thought to take a look at it, and as seems usual, I found some comments forming. So here goes.
First, another link to the interview, posted by Muon Ray on YouTube. Feynman tells the story in Surely You’re Joking, Mr. Feynman! in the chapter titled “I Want My Dollar!” (pp. 162-65) in my Norton paperback edition.
Short version: Feynman is working at Los Alamos during WW2 when he gets a note from the lab’s government patent officer, a Captain Smith, asking for inventions involving nuclear energy, to be patented by the US Government. Feynman later meets Smith at lunch and tells him that his note was “kind of crazy” to ask the scientists for *every* idea–“There are so many ideas about nuclear energy that are so perfectly obvious, that I’d be here all day telling you stuff.” Smith responds with–“Like What?” Feynman rattles off nuclear powered submarine, airplane, rocket, and amped up nuclear reactor as an electrical power station, and, as he was leaving, says there are “a million ideas.”
Three months later, Smith calls Feynman. Sub is taken but the others are yours, he says. Smith has apparently prepared patent applications on Feynman’s handwaving descriptions of what Feynman has called *obvious* ideas. Well, maybe Feynman doesn’t have “ordinary” skill in the art, so what he thought obvious doesn’t count. When Smith asks Feynman to assign the inventions to the U.S. Government, the agreement says that Feynman will receive $1 ($30 equivalent in today’s dollars) for each assignment. Feynman calls this a “dopey legal thing.” But there’s no $1 forthcoming, so Feynman insists and eventually Smith pays Feynman from his own pocket. Feynman goes off and buys cookies for the lab. “I got a prize! A dollar for my patent!” Everyone goes nuts submitting their invention ideas to Captain Smith and demanding their dollar. Smith goes nuts trying to find a way for the Government to pay all the inventors their dollar so it wasn’t coming from his own money. Feynman later gets a call from a big airplane company looking to build a nuclear-powered airplane and found his name on the core patent. They offer to make him director of their development lab. Feynman declines.
Let’s work through this.
First, there is no record I can find that Feynman ever got any patents issued to him. Steve Sether posted a note on Stack Exchange in 2019 that pointed out that he couldn’t find any record of Feynman patents, and proposed that Feynman was mistaken or made the story up, or the patents never issued, or patents issued but there was a spelling error or the like which makes it hard to find. All good thoughts. One commenter, going by the handle “simplicio” suggests Feynman’s patents could have been classified–by Feynman’s account they were after all dealing with nuclear energy during World War 2, while Feynman was working on the bomb. simplicio also mentions a bibliography entry for a personal communication that suggests Feynman “did do some brief post-war work on atomic aircraft….” Another commenter, Bill Streifer, who has been doing research on Manhattan Project inventions, points out that some of those inventions remain classified or redacted:
The Manhattan Project patented everything. Until a patent was published, it remained between the inventor and the US Patent Office. Once it was published, it became unclassified and anyone could view the patent details. And those details were very detailed. Regarding how a patent for an aircraft engine could possibly be classified, they ALL were. We’re talking about a nuclear engine. Over the years, there have been many nuclear engine patents, they weren’t as “obvious” as “you bolt it to a rocket.” Nuclear engines have many parts and each part was classified for many years after the Manhattan
Here’s the list of Manhattan Project patents that Bill Streifer obtained by FOIA. No inventor names here. But with about 40 inventions per page, and 41 pages of them, we are looking at 1500 or more patents–filed between 1942 and 1946, mostly, and many issuing in the late 1950s, some in the early 1960s, and at least one as late as 1965. Poor Captain Smith–if he didn’t arrange for government funding all those $1 token payments for assignments, he would have been out a good deal of personal cash.
I expect that Feynman has the general details correct and his applications were classified and any patents that issued were placed under a secrecy order by the Patent Office. That has happened at universities more recently. It’s like the work was never done, and no one is allowed to disclose even that there is a secrecy order. So Feynman may be right in saying that he has no idea where his patents are.
On the chance that Feynman’s invention thinking might have resulted in patents, I went to the USPTO. Searching pre-1976 US patents there is not so easy to do. Public Search supports only searching by patent number pre-1976. So I went through the Index of patents issued from the United States Patent Office (the Smithsonian has the series but for 1961, which the Internet Archive has) for the years 1942 through 1965, and there’s no Richard Feynman listed as an inventor. But using Google Patents, I found an issued patent (Serduke 2,770590; filed 1950, issued 1956) that cites a patent application with Richard Feynman as the inventor (application no 43,993–I think this may be a misprint); filed August 1948). So there was at least one Richard Feynman patent application. But there is oddness here. If that application did not result in a patent, then how is it that Serduke knew of it, and of an application filed by Fermi, in 1950? Applications weren’t published then. And how does Serduke know about Feynman’s application if it is classified? Or, why is Feynman’s patent classified, but Fermi’s–also listed as prior art and co-pending with Serduke’s application–not? It’s possible that the Feynman application Serduke cites as prior art has some connection with Feynman’s idea to use enriched uranium and beryllium oxide to create an electrical power plant. So it goes.
Bill Steifer posted a later note on 1/2/2021 that suggests there may indeed be Feynman patents:
By the way, I’m “this close” to locating Richard Feynman’s “lost” patents. But it might take several months or more. Once I do, I’m going to write an article about it, perhaps for the American Intelligence Journal where I’ve been published often. I’ll also write an “ANSWER” here
So far, that answer hasn’t been posted, and I don’t see an article in AIJ. But there’s hope.
Let’s say Feynman is right, and there were three patent applications listing him as inventor, and patents issued, but perhaps it all was classified, and Streifer would you believe got this close to finding them. In the 1940s, when the U.S. Government took in patentable ideas (for which it required assignment or in some cases simply asserted ownership by operation by law) and issued itself a patent, the practice was to make the invention available to the public at no charge and perhaps also without the bother of licensing. The $1 that an inventor received for assignment was the only money that the inventor would ever receive by way of licensing the patent.
But an inventor might receive money through other channels. Look at Feynman–the big airplane company in California finds his name (apparently) on a patent and offers him a big-time position in their company–money and influence abounding. And what’s the enabling bit for this offer? That the company doesn’t have to negotiate for rights with the Government, doesn’t have to pay, doesn’t have delays. Doesn’t have uncertainties regarding access. Amazing, in its way, given all the flustering about “certainty of title” to patentable inventions made by university IP folks ginning up support for Norman Latker’s inside job to change federal government patent practice, resulting eventually in the Bayh-Dole Act. The government got title. Certain title. Only when folks like Latker started messing with trying to leave title with “federal contractors” like universities so they could grant exclusive licenses that Latker couldn’t grant in his role of patent counsel for the NIH that things start to get “uncertain.”
The federal government policy of using the patent system to publish government-owned inventions to the American public while reserving the prospect of obtaining foreign rights for the government to consider was formalized by Investigation of Government Patent Practices and Policies: Report and Recommendations of the Attorney General to the President, a report prepared by David Lloyd Kregeer (and referred to in some of the less careful academic literature as “the Biddle Report” though A.G. Francis Biddle got fired by Truman in 1945 before he could write a lick and Tom Clark was the new A.G. (and later, Supreme Court justice) when the report was completed, but what the heck). Here’s a link to the Report at Hathitrust, The Report issued in 1947. Kreeger wrote an article that summarized the Report–“The Control of Patent Rights Resulting From Federal Research“–published in Law and Contemporary Problems, vol 12 (Fall 1947).
The core of the Report is that yes, the federal government can issue patents to itself, but for a number of compelling reasons, federal agencies should not assert patent rights against American citizens and companies, should not get involved in patent litigation, should not grant exclusive licenses or take a money interest in licenses. Kreeger makes great arguments for these policy recommendations. Government should not play favorites. Government should not have a money interest in patents and should get its funding from Congress. Government should not sue its citizens and companies for practicing inventions made in federally supported work. Government should not give companies a right to sue as proxies for the federal government, backed by the federal government.
And pause to consider the potential for malice in the federal government taking out patents on its inventions in other countries–using those countries’ patent laws to extend U.S. Government control of new technology into those countries, and perhaps excluding all practice in some country for a time, or playing favorites there, or creating a market for a U.S. company monopolist to control that country’s market and access to technology. Sure, maybe that sounds clever, but geez, the malice that comes along for the ride. Be feared, but not hated. Folks should know their Machiavelli if they are going to go this route.
In effect, the Report recommended that the federal government should use the patent system to publish inventions, identify inventors as having made them, and work the opportunities to enable inventions where federal agencies have a need or where it is within their mission to promote use. That sort of thing is directly in line with the Constitutional premise of the patent clause–to promote the progress of the useful arts, where there’s a good argument that “progress” is akin to “dissemination” rather than, say, creating thousands of little pimples of private monopoly to shake down or thwart makers and doers. It’s one thing for inventors to decide to do such things on their own. It’s another for federal agencies to say “there are not enough pimples so we will make even more and that for sure will advance innovation through government research and development.”
Combine the Report with Vannevar Bush’s argument in Science the Endless Frontier (1945) to open up federal classified research as much as possible for public use after the war, and with the confustication of federal agencies in trying to make their rush to embrace research result in “technology transfer” (as it came to be called by the mid-1960s) and discovering that just because they patented stuff didn’t mean anyone in American industry cared, and you’ve got the concoction that led to federal patent attorneys like Latker pushing Bayh-Dole. If federal research is largely irrelevant to industry, well, then, make the fools pay by trying to grant exclusive patent licenses to folks ready to try to prove them wrong or to shake them down–which means, never granting non-exclusive licenses until you have failed for a decade or more trying to find a cooperative monopolizer. There’s a built-in bureaucratic delay in access that the patent and exclusively license folks just won’t talk about. Their code for it is that technology transfer is “hard.” Well, yes, with practices like that, it necessarily will be.
For Professor Feynman, however, things were still in the glory days of government open access, and the benefit for Feynman–if he wanted it–was that folks with business propositions could find him, that he was linked publicly and directly to the ideas surrounding certain applications of nuclear energy even if that was not an area he in which he would ever think to publish. The take away: even with free access to federal patents, inventors still might have opportunities to profit (if that’s what they want). Or, more directly, perhaps, because of open access, inventors get way more opportunities to profit (if that’s what they want). It’s just that whatever opportunities come their way, the profit won’t take the form of a share income from a patent license. Not demanding payment to use an invention does not mean an inventor doesn’t profit–it just means the inventor doesn’t profit via patent licensing. To think all else is darkness is just foolish.
Translating ideas, even Feynman’s ideas, into social benefit does not appear to be driven by withholding practice by means of patent positions. At least not for federal government patent positions. The role of federal ownership of patents is not to produce more pimples of patent monopoly, just now ones that have effectively excluded inventors from access to their own inventions–but rather to create broad access and to facilitate that access with opportunities for inventors to work with those who would test or acquire whatever has been invented. In the mid-1960s the federal government tried to encourage technology transfer by setting up regional technology marketing operations and give these mighty sounding names and hiring transfer specialists to curate lists of inventions and try to get company folks interested, but somehow it just didn’t work. Like well yeah I would guess not. But having failed at that bungle, the federal bureaucratic torpedo searched for a new target to acquire. Must be the bit about not excluding others. What good is a patent if it isn’t used to exclude all others? Exclude even the inventors! Somehow, someone should see a money incentive in shaking down an industry with the threat of a patent, and by giving favored access to that sort of someone, why then society will benefit from all those federal research dollars that otherwise were being spent on stuff that other than an occasional frozen orange juice weren’t that particularly relevant.
But Feynman and his fellow scientists were motivated by money–not licensing revenue but rather that $1 for each patent assignment. That was all that was needed to get scientists at Los Alamos to submit piles of ideas for patenting. Those $1 hits then could be used to buy cookies and shared all around. No, it’s not mocking the patent system or the government’s interest in getting and publishing new technology ideas by means of the federal patent system but rather reflects a playful spirit, a sort of game to get a dollar from the Government in return for publication of their ideas. The $1 becomes play money, to buy treats. The patent connects something new (and available to all) with inventors’ names. Benefit and opportunity if any follow routes other than patent licensing (which doesn’t work all that well anyway–for universities seeking to license exclusively, maybe they get 1 deal in 2,000 inventions to go commercially successful, or one deal every five years or more at a “top” university. That’s 1,999 dead ideas with no access for every one that might have commercial success, even if not “lucrative” success. What a deathtrap for anything new.)
In a way, then, the Los Alamos scientists were and weren’t motivated by money. They could not care about money from licensing their inventions–that simply was not going to happen. They cared about getting their dollar out of Captain Smith. That sort of money. But somehow, inventors should be motivated by the prospect of 1 chance in 2,000 to get a tiny bit share of a university’s tiny bit share of payments five or ten years later from a patent license, during which time the inventor has no prospect of practicing anything about the invention while a maybe a company eventually shows up to futz around with “development” of a product practicing some tiny bit of what has been claimed. It’s nuts. I’ve never seen a university pay their $1 to inventors upon assignment. As Captain Smith tried to explain, “it’s just a formality” of expression–even when it is not. Maybe universities should pay out $30 per patent assignment upon signing. Maybe half a dinner out would be more motivating than overselling the prospect of making life-changing income from letting the university take a patent position that excludes all others–including the inventor.
Do you think that university scientists will do better science if they believe the technology transfer pitch and think they will have a money interest in the public use of their results? Will they publish better, more accurate, even candid scientific papers, practicing even Feynman’s standard for full disclosure honesty as the obligation of science? And if you do think scientists will do better science and innovating if they have a serious financial interest in subsequent use of their work, then wouldn’t it be better to have university scientists stand to get 100% of whatever money is available, and let them pay out whatever they choose to owe to anyone who helps them get to that golden bowl?
By way of contrast, it seems that the money motivation argument is more directed at everyone but for the inventors. In a devious twist, it appears the thrust of the institutional claim to hold patents is that society needs more patent monopolists, and these will burst out, if not erupt, if there’s an offer of money to be had from holding the right to exclude all others. Or is it the universities that are motivated by money from patent positions–motivated to create more patent licensing operations, and spend more money on them, in the hopes of getting a big payout, even if only once every few decades while holding back literally thousands of inventions.
In a way, many of Kreeger’s arguments regarding the federal government and patents also apply to universities. Universities might own patents, but they should not be in the business of suing citizens or companies or setting up a patent monopoly operation (which they call a “start up”) that actsnas a proxy for their interests. A university cannot act as a regulator of research behaviors at the same time it has a financial interest that might lead its administrators to twist their regulatory powers to advantage their hope of big money. Universities should not play favorites, should not get in the way of practice, and all that.
It would appear that money motivation at universities–the University of California did this in a big time way in 1960 when they set up their own patent licensing operation–was to get money for research from patent licensing. After all, that’s what Research Corporation was doing (but differently, but hey bureaucrats tend to understand only bits of what is)–so why couldn’t a university do it without Research Corporation and keep even more money for themselves–you know, for research? Well “for research” ended up being, really, “for research administration” which is just (in the administrative twist) just a part of “research” generally, no? and thus the money motivation comes down to this: a university will be motivated by the hope of money to spend more money on research administration and technology licensing services.
The patent monopoly argument, in practice, reduces to the idea that the prospect of big money motivates universities to spend more money on research administration and patenting. Preventing ideas from getting used immediately, so the argument goes, is essential to getting ideas used at all. It doesn’t make much sense. And I don’t see how from this nutty idea we get better science, better technology transfer, better technology change. It’s like somehow to solve the problem that not a lot of research sponsored by federal agencies in the 1960s was relevant to the American public the answer would be to create a ton of university licensing operations taking in every idea, patentable or not, and holding them for a prospect of 1 in 2,000 commercial success in the best of times.
Feyman thought the applications of nuclear energy were in the millions–millions of ideas. The question is, who ought to be thinking practically about those ideas? Federal patent attorneys? Technology transfer officers? Patent monopoly folks? Will federal research become more relevant because every patentable bit of it gets patented by some university or federal lab and held for an exclusive license that mostly never shows up and even when it does mostly comes to nothing (other than extending the time the public is denied access)?
In effect, holding anything up for the hope of an exclusive license (to create that patent monopoly) gives all the companies not getting that exclusive license or not wanting it or not wanting to be saddled with a prying relationship with a university incentive to design around federal and university work. The university patent practice works to make federally supported work even less desirable and less relevant! Where that work was often passively not relevant before, now it is actively not relevant–folks can’t practice even if they wanted to, unless they agree to be that pimple. So they ignore, design around, undermine, wait it out. That’s the effect of university and government monopoly positions. More licensing offices doing worser jobs (because “it’s hard”), more inventions behind paywalls that never pay. Nothing like what Feynman says he experienced.
It comes with some bit of irony that Los Alamos National Laboratory has a “Richard P. Feynman Center for Innovation” that appears to be their new packaging of their patenting and licensing operation:
The Richard P. Feynman Center for Innovation is at the crossroads of research and market realization, providing a gateway to transform groundbreaking discoveries into impactful solutions. Our mission is to fuel entrepreneurship and commercial ingenuity among scientists and engineers, paving the way for innovation-driven growth. From safeguarding intellectual property to forging dynamic collaborations, the Feynman Center provides comprehensive support at every step of the journey.
Back in the day, when LANL was still part of the University of California, I participated in a two-person multi-day site review of the LANL technology transfer effort. I should have suggested handing out $30 per invention disclosure!