Cloudy with A Chance of Patents

I held off writing this article for a long time out of sheer respect for my former dear team members and colleagues. Now that said blocker is gone after the news yesterday, I thought that the time is finally here. I mean after this sad, yet unsurprising development:

https://www.iam-media.com/copyright/microsoft-shifts-its-focus-part-its-ip-team-heads-exit

A lot of people out there are making noise out of patents in relation to the cloud, big data, artificial intelligence, block chain and what not. Yet, most of those areas will be nearly irrelevant from patent perspective, as they pertain to cloud computing with all that it entails. I will below try to explain why this is, for reasons that have nothing to do with artificial intelligence and block chain, and everything to do with the patent system. And, frankly, its blatant shortcomings.

If you digest what I am saying below, you will also understand some of the reasons for why Microsoft has now fully wound down its patent licensing activities. Because it makes perfect business sense - for them. And, therefore, is certainly the right business decision - for them. Crucially, it will still not make sense for everyone else to emulate them with the reincarnation, if you will, of "The Burning of the Ships" from the already distant Marshall Phelps era.

SKI TRIP DISCUSSIONS

My oldest daughter is now 13 years old. For some reason, she has always been crazy about back country skiing. Waist deep powder and 40 plus degree steep slopes is something she loved already at the age of seven. Well, neck deep for her. When I was living in the Seattle area and working at Microsoft doing patent licensing, she was always, and I mean always, willing to go with me to any of the Washington state ski resorts for some powder runs. Now, those drives can be boring, so she sometimes made me tell her "war stories" about my job. Basically IP licensing, monetisation and litigation. She'd listen intently and insist on the next story with the snow crowned domes of the peaks around I-90 or highway 2 surrounding us. I suppose this happens with most parents, whether you are a carpenter, police, nurse or IP licensing executive. Kids look up to their parents, until they grow up I suppose.

At some stage she got to a point, where she said she wanted to have a career in IP, which outright horrified me as a father. Because, as her father and knowing what I know, I would definitely not want any of my kids to embark on a career in IP. I will below try to explain why. Now, she's luckily over it and, taking inspiration from Mr. Crabs in SpongeBob ("I like money!"), wants to become a businesswoman. Which I think is a far better and much more future proof choice than anything around IP, not to even mention other areas of law. Certainly has the character for that with her love of extreme sports and all kinds of competing. Yet, paradoxically, it was the peak patent market between 2008 - 2018 that allowed me to become a businessman. For which I remain very thankful.

THE PEAK PATENT MARKET: 2008 - 2018

To me, IP is still mainly about patents, and a little bit about trade secrets and data (and, no, I do not generally mean personal data with that) also. Trademarks, where I started my career, I believe is, to be blunt, a bit on the simple and boring side. Certainly a stable and solid area for long careers, if you want to feel like an expert and being in control already early on in your career. Just that not a lot has happened or should happen in that area of IP lately. Copyright too is not exactly happening, outside the licensing of copyrighted software. I do believe that we are essentially nearing a situation, where people are paid or otherwise compensated for watching content - rather than the other way around. The world is just so full of content and getting even fuller, while the time remains limited as ever. I even understand that big names in music are increasingly dependent on merchandising and concerts for income.

Now, I am not saying that a career in copyright also outside software licensing would not be a stable, steady job. It probably is. But neither copyright nor trademarks have seen the fascinating action patents saw over the two past decades. Really with the global patent licensing and litigation market peaking somewhere from about 2008 to 2018 or so. I was blessed to be in two great patent licensing jobs in that unique era, first at Nokia and then at Microsoft. Really like a breath of fresh air for a person with a legal background.

PATENTS ARE GETTING "DUSTY"

One of my own moments of "revelation" came when I, as a computer gamer since the late 1970s, heard that, given the advances in connectivity, 3D graphics rendering for games was moving to the cloud. That sort of makes sense as the resource drains are immense, just that I thought that the latency inherent to communications would prevent that from happening. Latency, for a computer gamer, often means an embarrassing death in an on-line game. Even outright glitching. But if such a super time sensitive thing as 3D graphic rendering can be done in the cloud without excess latency from the connectivity used to beam the end product to gamers, then basically everything save for sensors and client device side connectivity can, and therefore will, disappear into the cloud so to speak.

This is why, as the world and technology progresses, the patent system is getting outdated and, frankly, outright "dusty". Here's what I mean. In the increasingly globalized, hyper connected and cloud computing centric world, the patent system suffers from the following fatal flaws:

1) way too slow;

2) issues with detectability;

3) territoriality; and

4) dependency on the nation state.

I will handle them each at a time, based on more than 10 years of experience of trying to monetize patents in the cloud - without any real success with that - while succeeding beyond the wildest expectations outside the cloud for almost twice as long as that.

WAY TOO SLOW

It takes, at best, normally at least 3 - 5 years to get a patent granted. But 5 - 10 years is not uncommon either. And I have seen many cases where it takes more than 10 years. In some countries, like Brazil, that is actually the norm. Now, the term of patent protection is 20 years, so if it takes more than a decade just to get a patent, the world may have moved on from your invention. And with all the technical progress in the world and the ever increasing patent filing numbers, the patent system is, if anything, getting even more clogged up and slower over time. Not least because of what some individual nation states, or to be blunt really mainly just China, are doing. More on that below.

Now, it is actually still not that common for the world to move that fast. If you invent something groundbreaking today, the odds are that it indeed takes 10 - 15 years for it to be deployed at commercial scale, which commonly entails things like reliability, low manufacturing/maintenance cost, miniaturization, low power consumption, supportive regulation and so on. So, the world of technology, even still today, moves generally so slow that almost everything being of a technically groundbreaking nature out there on the market was invented at least 15 or more years ago. So far, so good. Looks rosy.

Now, crucially, while this may seem to solve a problem, it actually also introduces a much bigger problem. I know a guy, with a patent, who basically invented the application store (think Apple's AppStore or Google's Play) way in the nineties. While he could technically describe it in his patent, the language used in the patent of course sounds archaic or even flat out dorky almost 20 years later. Combine that with how prevalent application stores had become by then and you will understand that he did not stand a good chance of prevailing in litigation. But he actually did, once against the mighty Apple, at a grand scale. But failed against another company later. Mainly cause the language in the patent sounded outright dorky 20 years later. I have many other similar stories. Think about it, if you now invent something, how likely is it that you can, sitting here now, describe it in a manner, which still sounds relevant and compelling in 2040? You probably cannot, nobody can. Every patent is, in a nutshell, a 20 year time capsule into the future. And time travel has never been easy, as we know.

Finally, the way patents are intended to be used, as exclusionary rights, is through patent infringement litigation. That is to say that, if the licensing discussions fail, the patent holder's "product" is essentially the right to try to enforce the patent before a court of law, sometimes subject to certain constraints such as those arising out of so called (F)RAND licensing commitments. The problem with patent litigation is that it is slow in most places. Any outcomes faster than 2 years are rare and 5 years is not uncommon globally in the first instance. Add some years to that for appeals and you will get the picture. Not only are patents awarded super slow, but they are also super slow to use.

Actually, the window for using most patents in litigation is just a few years. And, even though the world and the technology progresses around us, patent litigation remains as slow as ever in most places. And is getting slower in some places. Patent litigation is of course also a very high friction, contentious situation. It is increasingly hard, or even downright impossible, for companies to be engaged in "patent wars" lasting a half a decade or more against other companies. This holds true even against competitors, not to even mention suppliers and customers or potential customers. Too high friction and way too confrontational for many - and also lasting for way too long often. I have lived through such long-lasting "patent wars", including the largest one ever, so trust me that the impact on organization are immense and supremely stressful.

So, suffice to say that, in this day and age in particular, that patents being so slow and reaching over a few decades makes them vulnerable to the passage of time. Far more so than simpler IP rights like trademarks, and copyright up to a limit, which generally age far better. After all, Coca Cola is still the same brand, only stronger, and the Beach Boys still sounds the same. But patents from the 1990s may make you chuckle now in 2019.

ISSUES WITH DETECTABILITY

Detecting patent use and, therefore, infringement absent a license, is becoming increasingly challenging. I have spent so much time being frustrated about this in a cloud context. As much of the computing and technical progress moves into the cloud, it also moves into happening behind armed guards. If you want to see what happens in a cloud data center, you may have to travel to Eastern Washington state, which basically looks the same as Nevada in places, and try to find the relevant data center. The next thing you will find out is that it is surrounded by armed guards, fences and barbed wire. You will not get in alive, uninvited. To be fair, sometimes the cloud hosted services expose their functionalities in online documentation or through the interfaces they expose. But, as I explain below, these can be changed in microseconds, if necessary.

What is particularly challenging with the cloud is that the hardware and software stacks they run in data centers are almost without exception highly optimized, customized and proprietary. That is to say that, if you want to buy the server hardware utilized in some cloud data center, you most likely cannot, as it is not for sale openly for you to get it for testing. Same even with the software. While much of the cloud is at least notionally open source based (think of the infamous LAMP stack = Linux, Apache, MySQL and Python), even the underlying core software stacks are actually highly customized, optimized, proprietary forks. Again unlikely to be available to you, not to even mention the higher, customer specific and proprietary levels of software, which are virtually never available for you to inspect and examine. Not even in the binary form, which would be generally useless anyhow.

Herein lies a great irony. Not only is the cloud in many ways killing the patents, it is even more so killing the open source movement. This because most open source licenses attach all kinds of obligations, including having to provide access to source code, to the act of "distribution", while the cloud data centers do no such thing. The customized software sits there on customized hardware, behind armed guards and barbed wire, without being distributed. And thus, in effect, becomes not just closed source, but even outright secret. The total opposite of what the open source people imagined. Oh, the irony with that!

In common law countries, you may have have some recourse to get document discovery to get clarity on what happens in a cloud data center. But way before that, you need to be able to have some theory of infringement and some justified business case for suing in the first place, each based on detecting infringement. And if you for some reason succeed with all that, as I explain below, the infringement can switch to another country or cease in mere microseconds. Slippery! The same essentially holds true for civil law type orders to secure evidence in IP matters through the involvement of police and/or court bailiffs. Equally unlikely to work in the cloud.

So, to make a really long and complicated story short, in today's increasingly cloud connected world, you are unlikely to be able to detect the infringement of many types of patents fast enough, reliably enough and compellingly enough. Put differently, while the commercial bargain, from the perspective of economics, is that you share your ground breaking invention openly with the world in exchange for a 20 year exclusive right in the same, the other side of that bargain has changed. You may not be able to know at all, who, where, when and at what scale are using your invention. So, the bargain is increasingly dead and other types of IP and quasi-IP, such as trade secrets, are taking the stage instead.

TERRITORIALITY

While there are some global and regional types of "patent application bundles", patents remain, for now, territorial and their effects limited to the area of one particular nation state. This is so very, very limiting in this hyper connected, globalized day and age that it is almost laughable. The only effort to go beyond that has been EU's project with the so called "Community Patent", which has taken decades and has been complicated by the Brexit most recently. Said EU project for the same has been so tortured that I have not even bothered to follow it lately. In part, also because the EU "Community Patent", while a step in the direct direction, will still be way too limiting and territorial for the cloud computing age. Here's what I mean.

I first bumped to this in about 2008 or so. We had a perfect infringement read, or so we believed, on something that we exceptionally could even detect in the adverse party's cloud data centers, as they carelessly exposed their interfaces and functionalities in a thoroughly and openly documented manner. Their primary cloud data center was somewhere in the UK and we could expect to get a litigation "hit" against it in 2 - 3 years to force them back to the negotiation table. The downside? They had a known backup cloud data center in, drum roll, Ukraine offering all the same functions. So, if we'd get a "hit" after years of hard work, intensive litigation and, say, 5 - 10 million pounds Sterling in litigation investment, they had a "get out of jail card". In less than a second, they could move all their workloads globally to Ukraine, as they did from time to time anyhow for mere maintenance reasons. So, all those years of high risk investment and work could be negated in less than one second. We of course had no patents in Ukraine and, as patents require solid rule of law to work, it would have not made much difference even if had them there. And, after Ukraine, they could set up a new data center relatively fast anywhere else also.

Another funky example is distributed infringement. Patents commonly involve many claim elements, the practice of which can be distributed between various cloud data centers. As an example, the geographical location processing can be done in the UK, the advertisement data may be done in the US based on some AI algorithm and those two can be combined with other elements performed in China. So the core steps of some patents can be practiced in several countries without any of them practicing all of them. This is not uncommon. In fact, it is more of the norm as the cloud computing is commonly distributed across multiple companies, who just happen to have domicile and operations somewhere - unless of course they rely on outsourced cloud computing data centers in various other geographies, as is also very common. Actually, the way the same patent is practiced in a distributed manner for the ultimate benefit of someone does often change over the course of same day, as the daylight and, correspondingly, the human activity varies across the globe as a function of the time of the day locally. Good luck with figuring that out as a patent holder. Slippery! The doctrines of indirect and contributory infringement are not advanced enough in many cases to construe infringement on individual claim elements being practiced in multiple, and maybe varying countries depending on the time of the day.

This is a glaring problem with patents in this day and age. They will lose in the global game of "whack-a-mole". Slow, complex and "dusty". Dependent on courts and years and years of complex, high-expense litigation. While in the cloud massive workloads can be moved from one country to another or even to another continents in mere microseconds. Just hit a button, and millions of instances of some cloud app workload are revved up 10,000 miles away and all the associated Internet traffic is directed there instead. There are even cloud computing services offered in "no country" that is to say on ship vessels outside all territorial waters. And even in the space in satellites. To add insult to injury, if you do not feel like moving the workloads somewhere else, often the alternative is to roll out a relatively simple design around, making some minor change that gets you around the patent. This too can take mere microseconds to roll out as massive scale. Rather than recalling millions of products, you just press a few buttons.

So, yet again, the patents lose out. Too slow, too complex, downright "dusty" even. THe Internet is global, without borders, while the patent system is extremely local and regional. That is to say territorial. And really most optimized for the tangible products of the gilded age of industrialization from the late 1800s to late 1900s. But not agile and quick enough for this day and age, not to even mention being future proof. Something should change, but with the way globalization is being retracted today, unlikely to happen anytime soon. Certainly not feasible to even entertain the idea of global patent system in our present age. Quite the opposite with the way global geopolitics are going.

DEPENDENCY ON THE NATION STATE

This is such a complex topic that it warrants its own article. Suffice to say that the patent system has always been dependent on the nation states. It is the nation state that awards patents through its patent offices. And it is the nation state that enforces them through its court system and court bailiffs. This has always been the case and, in and of itself, is not that peculiar. There are many other areas, where the nation state is equally deeply involved. Take the real estate system as an example. Registered parcels of land are created, as units of property rights, by the nation states, who also enforce the rights and obligations arising from the same - not to even mention controlling the use of the same through zoning etc.

However, China has basically crossed the border here as well. They opened up to the global markets belatedly after decades of being confined to their own markets. They then found themselves many decades behind others in technological progress and, consequently, being a huge net payor in the patent system. I believe that they were so irritated by this and disliked the tables being turned against them so much, that they chose to drive the global patent system to the wall. Really to illustrate its weaknesses. To make a point and to send a message. The main weakness they exploited was that of government subsidies, the domain of anti-trust law through the doctrine of state aid, but generally not regulated by IP law and its global conventions. They openly came up with 5 year plans, in a classic communist manner, deciding to come up with millions of patents over a very short time. Like other communist 5 year plans, this too materialized, as was decided, when the Chinese government poured vasts amounts of money in tax subsidies for patenting. In a centralized communist system, you tend to get what you decide - whether it makes sense or not. That's the whole problem.

Things got so out of hand that, on the average, China domiciled companies make a profit for filing a patent in China. Let that sink in. A profit, not a cost. What obviously happened was that millions of patents emerged in a few short years in China. As filing for patents was profitable activity in and of itself. Of course, large and hard working as China may be, they did not invent as much stuff in a few years as, say, the US had invented in 50 years before that. Obviously not, quite the opposite. But the vast sea of government paid and awarded junk patents is now slushing out there, some of it even starting to find its way to developed markets such as the USA and the EU.

Thinks about it. In China, the R&D is done by huge companies, which like Huawei, can be 99% government owned and controlled. Then they get tax subsidies from the government to file patents for what they invent as extended parts of the government. And, through the court system, it is again the government, who ultimately enforces such patents financed by government money to begin with. So it's government, government, government: inventing patentable inventions, filing patents, awarding patents, using patents, litigating patents, licensing patents, being the licensor, being the licensee, enforcing patents and paying for it all. Nothing that the government does not do. Let that sink in. In keeping with the latest 5 year plan from the Communist Party.

What a mess! And what I believe to be yet another existential threat to the global patent system. What's worst about it is that everything that the nation state controls is subject to politicized central control. And if they do control "everything" with patents, as they do in China, then "everything" in patents is also subject to politicized central control and, therefore, political interference. Which can only lead to great waste and failure. As is obviously happening over there. Sad, really.

CONCLUSION

So, net net, what is the conclusion? Patents will remain relevant for anything implemented in tangible items such as phones, cars and the like. Something that you can buy, something that comes in a box, something that can be torn down and have its inner workings examined. Something tangible you can buy, even if just a pill of medicine for us to swallow - or analyze. But as technological progress moves increasingly to the cloud, this will lead to only things like terminal side connectivity, sensors and some limited computing happening in the client devices. The real progress will be mainly elsewhere, that is to say in the cloud, and not in the client device. Computing becomes ever more important, but increasingly vaporizes into the cloud, where patents will not work at scale any time soon.

Of course, patents will also remain relevant for relatively niche, smaller industries such as the pharmaceutical industry. Yet, even the pharmaceutical industry patents are very vulnerable to what I say above about the role of the nation state and political interests of and even outright interference by the nation states. Putting it all together, it is a fact that, for a host of reasons, the nation state based patent system is already painfully outdated for the modern age and becoming increasingly so. This will mean that a career in IP is unlikely to be as attractive as it was before over the past few decades. Would certainly not want my children to embark on a career in IP now, based on what I know. 

And, if you internalized what I said above, you will easily understand, why it makes perfect business sense for Microsoft to ramp down its patent licensing. It was very, very clear already by 2016, you know. After all, they are THE computing company, more impacted by the cloud than almost anyone else - overwhelmingly for the good. I do believe that theirs was the correct decision for them. Given how well Microsoft's cloud business is booming and given how dysfunctional the patent system is for computing and, above all, given how much friction out-bound patent licensing creates, it would be downright bad business for them to continue on their old course. As smart people, they will not. But, again, it may not necessarily make sense for many others to follow their lead to, yet again, "Burn the Ships".

Don't get me wrong, I still love patent licensing, monetisation and licensing. I do it a lot and, blessedly, these days get to work mostly on sensors for self-driving cars and connectivity more broadly. But how I feel about things is of no consequence, obviously. Now, maybe one day we will have some unified global patent system or some other global harmonization such that the much awaited "pendulum swing back" will truly to start to happen. Until then, in the world of IP, we will see the rise of trade secrets and data (and, no, I do not generally mean personal data). Trademarks and copyright will remain the same old, the same old, as always. There will remain work with patents too, for sure, but the growth era is gone.

At least until the patent system catches up with the real world one day, I hope, as a longtime pro-patent person. For now, with our existing grossly outdated patent system, when computing goes to cloud, patents don't follow. Simple as that.

Yet, in the meantime innovation continues unabated. Cause it's far bigger than just patents.