Sunday, 4 September 2011

OnLive Founder and CEO Steve Perlman Opposes the S.23 America Invents Act

The patent system adds the fuel of interest to the genius of invention.” – Abraham Lincoln 1858.

We are at a major decision point in our country’s economic future. Big business is again trying to kill the financial incentive for American inventors by further destroying the American patent system upon which innovators and inventors rely to get the start which creates the small businesses which create the jobs which our economy needs so badly.

To small entities, the patent system is the engine that fuels and sustains our creativity, and which has done so for 220 years since it was established in 1790. Virtually every major business today, both large and small has dealt with issues of patent protection. It is ironic that many of the very big businesses that are now trying so hard to kill the patent system and eliminate that “fuel of interest” for small businesses relied on that fuel, that patent system, to become big. Steve Perlman, OnLive CEO and former VP of Microsoft, and the genius behind much of its software gives a super explanation of how S.23 would have likely stopped him from causing Microsoft to get the protection it needed to become big. How ironic that Microsoft now leads the call to pass S.23.

How S.23 Severely Disrupts America’s Unique Process of Invention

Steve Perlman, President & CEO, Rearden, OnLive and MOVA

I am an American inventor. I hold over 100 US patents with over 100 pending. My inventions are found in most computers and mobile devices and many TV devices sold today as well as in the production of major motion pictures with advanced visual effects, such as the latest Harry Potter movie.

The America Invents Act S.23 has passed the Senate. While much has been said through the four sessions of Congress the Bill has spanned, almost all discussion has been conceptual with very few, if any, real world examples of how patents are actually used in the US for the invention and development of the major innovations that form the backbone of US economic growth.

The following provides real world examples of how patents are currently used in the development of actual US inventions and how the core provisions of S.23 would disrupt this process, particularly for startups and other small entities that are the primary drivers of innovation in the US.

In terms of background, my inventions include QuickTime, WebTV, MOVA Contour, and OnLive, and span a wide range of fields, including video, audio, animation, special effects, 3D imaging, wireless, alternative energy, semiconductors, optics, material science, mechanical systems and medtech devices.

In addition to founding 8 US-based startups backed by my patents, I was a Principal Scientist of Apple, and a President of Microsoft. My work is found in all Macs, iPhones, iPods and iPads, most PCs, all of Microsoft’s TV-based products and in many TVs, cable/satellite/IPTV set-top boxes, and video games.

My inventions have supported investment of hundreds of millions of dollars into my startups, have resulted in billions of dollars of revenue and profits, and the creation of thousands of US jobs.

All of my companies have been pure “practicing entities”: my patents have been used purely to back the products my companies have developed, never sold or licensed for royalties.

The Process of Invention in America Today

Most of you have seen movies with computer-generated faces, but it wasn’t until Brad Pitt’s reverse-aging face in A Curious Case of Benjamin Button that you saw a computer-generated face that looked completely real, winning the film an Academy Award® in Visual Effects. One of my startups invented the technology, MOVA Contour, that made this possible. The invention process we went through is uniquely American, and is the same process used for most of my inventions (as well as for inventions of Edison and the Wright brothers, among others) that literally could not be accomplished outside the US, given the patent laws of other countries. If S.23 becomes law in its current form, this uniquely American invention process will cease to exist.



To understand why, I’ll walk you through the following picture, showing the process we used to invent MOVA Contour.

A large part of invention is trying out a vast number of ideas (such as Edison with thousands of light bulb filaments, or the Wright brothers with many wing shapes). When we set out to invent MOVA Contour, we came up with dozens of approaches to precisely capture the human face in motion and explored each of them until we ran into a dead end. Each of these initial inventions is shown as a black box on the left side of the picture. The tan line ending from each shows how far we explored each invention. A dot shows where we hit a dead end.

In the upper left, you can see tan arrows where the 2 most promising approaches were combined and then led to 7 more secondary inventions, which unfortunately all led to dead ends. But, one of them led to a Key Insight, shown in a dashed green arrow, that led to the rethinking of another initial invention (with a green arrow going to 5 boxes) into 5 more secondary inventions. One of these secondary inventions combined with another initial invention led to the first practical prototype, shown as a green circle with the word “Success!”.

At this point we had a working prototype that showed the basic idea worked, but not a practical product. Another 7 inventions followed for Practical Refinements, of which one led to a commercial product. For a Complete Practical System, 2 other adjunct inventions were needed (special lighting and makeup), and each of these resulted from testing several inventions, and selecting one. This entire process took about 5 years of intensely-focused R&D (this diagram shows only part of the work).


In total, about 100 inventions were conceived over the 5 years of development, but the only 6 inventions were actually used in the final system, filled in green, and those are the only inventions for which we filed patents. The reason we did not have to file patents on the other 90+ inventions is because the US is a “First-to-Invent” country and so long as we carefully document each invention, we maintain priority to the date of conception. This gives us time to determine which inventions are needed for the product, and it also gives us time to get a working prototype before we file patents so that we have something to show when we seek venture funding to cover the cost of the filings (the US system allows a 1-year grace period after disclosure to file a patent). Lastly, by only filing patents that matter, we minimize distractions to our key engineers and scientists in working with patent attorneys.

If the US were not a “First-to-Invent” country, and instead was a “First-to-File” country, then the process of invention would be completely different. To preserve defensible priority, every one of the 100 inventions would have to be filed as a patent immediately upon conception (which is why inventors throughout the world refer to “First-to-File” as “Race-to-the-patent-office”). Also, before a disclosure to investors (who rarely will file non-disclosure agreements) could be made, the patents would have to be filed, so venture dollars could not be used to file the patents. And lastly, the key engineers and scientists would be constantly working with patent attorneys to explain every idea they come up with as soon as it is dreamed up and have far less time to do development.

It typically costs us $20,000-$30,000 to obtain a commercial-grade patent. As you can imagine, in a First-to-File country, as a startup, we could only file patents on a small fraction of the inventions at the time of conception.Further, in the case of MOVA Contour, the inventions that looked the most promising at the outset turned out to be dead ends. Had we filed patents on them, it would have been wasted money, while the inventions that mattered would not have been patented at all, potentially making it impossible to fund the company. It is no surprise that the US is by far the leading nation in the world when it comes to startups and, since its earliest days (when “First-to-Invent” was established), has been known as a mecca for invention.

Like MOVA Contour, most of the products I’ve developed in my career simply could not have come into existence in a “First-to-File” country. And even with the financial resources I have at this stage of my career, it is still the case that the distraction to key engineers and scientists (myself included) in filing vast numbers of shallow and largely pointless patents would dramatically reduce our productivity in what are already intense multi-year developments, let alone further burdening the patent office.

S.23’s switch from “First-to-Invent” to “First-Inventor-to-File”

In dropping “First-to-Invent”, S.23 proposes the US switch to an obtuse “First-Inventor-To-File” concept which is “First-to-File” muddled with an immensely complex set of priority rules unlike any patent priority system ever used before in any country. It is far worse for small entities than conventional “First-to-File” because of the highly complex and obscure priority rules. Switching to “First-to-File” would be bad enough, as detailed above, but switching to a completely different system introduces immense complexity and risk. It does nothing to “harmonize” with other countries, and most certainly does nothing to reduce uncertainty of priority. Under “First-to-Invent” there are only about 200 “interference hearings” a year to resolve priority disputes, out of hundreds of thousands of patents filed. In short, “First-to-Invent” works exceptionally well in actual practice. Why risk changing it?

Even if the US were to adopt conventional “First-to-File”, it still would achieve little in harmonizing with other countries. Each country or region has its own statutes regarding what constitutes priority, prior art, what is patentable, etc. And, beyond that, by far the largest cost in foreign filings is language translation since patents are so technical and precise.

If “First-(Inventor)-To-File” is made into law, it most certainly will face constitutional challenges. I’ve provided legislative staffers with a detailed list of all scholarly articles on both sides of the issue going back the last 20 years. The constitutionality of switching to First-to-File has been a popular subject of analysis over the decades by constitutional scholars and the overwhelming consensus is that it is unconstitutional, and among scholars in the last decade, this is almost a universal view. The first Congressional Research Service (CRS) reports on the America Invents Act stated First-to-File was constitutional, yet cited a reference that concluded the opposite, so over the last 4 years, I’ve been pressing them to provide some scholarly reference to support the report’s assertion of constitutionality. Finally, last year they conceded there was no recent scholarly reference supporting constitutionality, so they added a reference to an authorless web page on a small law firm’s website supporting that conclusion. This arbitrary, authorless web page is all that is supporting the constitutionality of First-(Inventor)-to-File in any document before Congress. I urge you to take a look at the CRS report and verify this for yourself.

Lastly, it is important to point out that under First-to-Invent, a company with extensive resources can choose to practice First-to-File, by simply racing to the patent office as soon as every invention is conceived, eliminating any need to keep records of invention conception. Meanwhile, a company with limited resources can still utilize First-to-Invent, only filing patents that matter after funding is obtained.

Under First-(Inventor)-to-File, the reverse is not true: all parties, regardless of their resources, must adhere to “race to the patent office”. This places small entities at an enormous disadvantage to large entities. Indeed, it is a key reason why no other nation has anything close to the number of startups as the US.

S.23 and the Massive Patent Backlog

The core issue facing the US patent system today is the USPTO’s massive patent backlog.

Consider a real world example: A key patent for one of my startups just issued in December of 2010. It was filed in 2002, 8 years earlier. This patent was not even examined for 5 years. In the fast-moving world of high-tech, few technologies are so fundamental that they are even relevant 8 years later, and certainly it is impossible for small entities to utilize a patent for funding or closing key partnerships with such long pendencies.

While S.23 has a provision for the USPTO to increase its fees, it has no provision for the USPTO to retain its fees, which is the core issue. As recently as last October, Congress yet again took away the USPTO’s surplus fees, leaving it yet again underfunded to address the backlog. Increasing the fees absent the ability to retain them only serves to further burden inventors and small entities.

When patents are delayed past their product’s marketable window, the patents are largely useless to small entities, despite the enormous cost associated with filing them. One of the few ways to monetize such patents is to sell them to patent aggregators (pejoratively called “patent trolls”) who then may use the patents to retroactively sue companies who have introduced infringing products during the long pendency.

Despite this massive backlog, S.23 proposes to add a Post-Grant Review period after a patent issues to allow further challenges against that patent than those already allowed. Even setting aside the significant problems for small entities that Post-Grant Review introduces, given the current extraordinarily long pendencies facing small entities, we can’t afford any further delay to the certainty of the enforceability of a patent. With a large percentage of my patents taking 5 to 10 years to issue, we already have had to abandon several significant products because they were unfundable without the certainty of an issued patent. Until the patent backlog is addressed, tacking on more time for potential Post-Grant Review challenges will simply cause more products to be canceled and likely hand more patents to patent aggregators. We can’t afford any provision that delays patent issuance at this until the backlog is gone.

Patents Exist Primarily to Support Small Entities

If patents did not exist, established large entities would be the most able to survive, utilizing other forms of market power to defend their incumbency. Small entities with disruptive technologies, on the other hand, would simply not exist. Anything they would develop could be readily cloned by large entities, who could easily out-market them, and so no investors would ever fund the small entities.

S.23 is a Bill whose provisions dramatically favor large entities over small entities. Already, given the massive patent backlog, for many new products, the market window passes long before patents issue, placing large entities at an enormous advantage over small entities. Giving large entities even further advantage until the backlog is resolved will just make matters worse.

But worse than anything, S.23 undermines the uniquely American process of invention. It casually sweeps aside established paradigms that have successfully fueled the engine of innovation in America since its founding, and demonstrably fuels it today. We don’t want American invention to be like that of other countries. We want America to continue to be the world’s mecca for invention.

After 4 sessions of Congress considering the America Invents Act, not a single practicing inventor has ever been allowed to testify before the Senate, and only one inventor has testified before the House (and he strongly opposed the Bill). It should come as no surprise S.23 is so disconnected from the reality American Invention.

I urge you to oppose S.23. Focus instead on a much simpler Bill that would allow the USPTO to keep its fees and bring the massive patent backlog under control. After the patent system is functioning normally again, we can sit down and determine what changes would make it better. It is essential that practicing inventors from small entities are a part of this process.

More info about the opinion of OnLive Founder and CEO Steve Perlman on the America Invents Act S.23 is here:


SOURCES: TPL GroupBurdlaw IP Blog.

1 comment:

  1. Why not let MegaCorp have their steenkeen little patents and just not buy from them? Don't buy from anyone who has more than you. If my inventions ever make me any money, that is a philosophy I will try to live by as practically as I can.

    ReplyDelete