Remarks by Director Michelle K. Lee at Massachusetts Institute of Technology

USPTO Director Michelle K. Lee

March 31, 2016, 3:00 p.m.

On the Front Lines of Promoting Innovation: 

From MIT to Washington, D.C.

It’s hard to believe over 30 years ago I arrived on this campus as a freshman, from California. In fact, I attended my very first lecture (for 6.001) right here in 10-250. My McCormick dorm-mates and I also treated ourselves to many a LSC movie in this lecture hall after working hard all week on problem sets! I said it last year on stage at SXSW, where I was sworn in as Undersecretary and Director of USPTO, and I’ll say it again now: My time here at MIT was a rewarding experience that informs my work to this day. And I can still program in more languages than I speak, something I take pride in as I lead our nation’s “innovation agency.” I owe some of those great experiences to Professor Hal Abelson, who was good enough to give me the opportunity to work with him at what was then called the Artificial Intelligence Lab—and now you probably know as “CSAIL.”

I don’t think either of us expected my career would take the trajectory it did after MIT. I know I certainly didn’t. But right around the time I was finishing my graduate thesis at the AI Lab—on “Summarizing Qualitative Behavior from Measurements of Nonlinear Circuits” —the courts were considering a very important case, Apple v. Microsoft, addressing the copyrightability of a graphical user interface.  I realized, courts were effectively making an important policy decision—with real implications for the field of CS. This led me to understand that, as technology inevitably evolved, there would be an increasing number of interesting technology policy issues for which the existing laws would have no clear answers, and that someone with an understanding of science and law could help shape the outcomes in a manner that could benefit innovation and society.

I had always been passionate about innovation since I was a little girl. Like all the dads on the street I grew up on, my dad was engineer and my mom a scientist. I saw first-hand the tremendous economic opportunities created by technological innovation and entrepreneurship. It was not uncommon for dads (back then it was mostly dads) on the streets of SV to invent a clever bit of technology, file for a patent, and obtain venture capital funding to build a company to bring the patented invention to the marketplace. Some of those companies succeeded, and some did not.  But for those that did, a small number of them revolutionized the world and the way we live. So I always knew that I wanted to be a part of promoting innovation.  And when it became clear to me the importance of the role that law plays in promoting innovation, I knew I wanted to apply my technical training to addressing unsolved legal and policy issues for the benefit of our society. So rather than getting my PhD in Computer Science at the MIT AI lab, which I was on track to do, I took a detour to pursue my law degree at Stanford. Hopefully Hal has forgiven me for this transgression.

What I discovered in law school, and as a young IP attorney representing innovators and entrepreneurs, is that my time and effort at MIT—when I was pursuing a very different kind of career—proved tremendously valuable as I pursued the then-somewhat unconventional path of law. I will say that my MIT education helped me (1) to define problems more carefully,  (2) to foresee the impact of legal positions taken on innovation, and (3) to pay closer attention to data, which I’ll have more to say about shortly. And my technical background certainly didn’t hurt when I went to work for a young company called Google, and became its first Head of Patents and Patent Strategy. In the span of eight short, very busy and exhilarating years, my team and I helped build the company’s portfolio from few handfuls of patents to over 10,500—including patents for such game-changing innovations as Google Maps, Google Earth, YouTube, and driverless cars. Collectively, those experiences were a profound lesson not only in the value of my MIT education—which I still value and use to this day—but more broadly, in the power of intellectual property and its vital role in promoting innovation and economic growth.

Today, that role of IP in promoting innovation is more important than ever. It wasn’t so long ago that the most valuable assets of companies were often tangible—like their plants, warehouses, and inventory. Today, however, the most valuable assets of many of our country’s most valuable companies are increasingly intangibletheir inventions, algorithms, processes, designs, and brands—their intellectual property. And because these intangible assets can be more easily copied, the system we use to protect and enforce IP rights is critical to protecting investment and American competitiveness. It’s a fact our economic competitors recognize, appreciate, and mention often in my meetings with them. This point was driven home to me last May, when I traveled to Beijing and had the privilege of meeting with one of the most senior members of the Chinese government, Vice Premier Wang Yang. 

During that meeting, he emphasized China’s desire to strengthen IP protection and enforcement—not just because their trading partners are requesting it, but because China views it as necessary to their desired transformation: from a low-cost, manufacturing-based economy of inventions developed elsewhere, to an innovation-based economy with technologies developed in China that provide products and services higher up on the value chain. Put another way, China wants an economy more like ours, in which innovation and IP plays a greater role than low-cost labor. Indeed, the entire U.S. economy today relies on some form of IP. IP-intensive industries support at least 40 million jobs and more than one third of our gross domestic product. Here in Massachusetts, about 22% of all jobs are in IP-intensive industries, placing it in the top five states for such employment. This should come as no surprise to MIT alum, about 30% of whom have filed for patents!

A paper just issued by one of the USPTO’s Edison Scholars further underscores the significant impact that patents have on startups. The paper reported that approval of a startup’s first patent application increases its employment growth over the next five years by an average of 36 percentage points, and boosts its sales growth by 51%.  This research makes clear why it is a priority of the USPTO to make sure innovators and startups fully understand how IP can be used to support their business goals. 

We want more startups to succeed because their success benefits all of us. But we also know that recently many startups (and big companies) have seen the power of patents from the other side—as the subjects of abusive patent litigation.  By this, I mean, e.g.,, when an entity brings meritless infringement claims against others sometimes preferring patents that are broader and vaguer to assert against entire industries with the sole purpose of extracting settlements just under the cost of defending against the lawsuit, hoping that the alleged infringer doesn’t look too closely at the merits of the claim.  But, abuse can also happen on the defendant’s side, when a defendant draws out the time and expense of a baseless defense, hoping the plaintiff has neither the will nor resources to pursue its claims.

Either way, abusive patent litigation can have a devastating impact on innovators, especially startups, since the cost of patent litigation can often exceed the amount of its latest round of funding. USPTO recognizes that there is a cost to society if we issue patents we should not have, just as there is also a cost to society if we don’t issue patents that we should.  This tension speaks to a fundamental balance that I believe the patent system must achieve between (1) incentivizing innovation and (2) allowing others to make improvements to existing technologies.

As to the first part of the “balance” – incentivizing innovation,  the better we do our work at the USPTO, the better we are to quickly provide protection so that innovators can continue with the important work of further research and development, knowing they can recoup the rewards of their efforts and investment. The second part of that “balance” involves making sure we issue patents correctly and clearly, to permit and incentivize follow-on innovation. Patents that have clear language provide meaningful notice as to what is covered by a patent and what is not. It is important to signal to competitors what areas are left for further innovation, whether they need to take a license, or if they should attempt to “design around” the original patent.  Such “design arounds”, or follow-on innovations, are a key part of what the patent system is meant to encourage.

Consider the example of the laser. It began in 1960 with a patent for a MASER (not a LASER), standing for Microwave Amplification by Stimulated Emission of Radiation. This patent issued to Charles Townes and Arthur Schawlow—two distinguished members of our National Inventors Hall of Fame. Over the next three decades, there was an outpouring of invention in this space as innovation spurred more innovation—from the ruby laser, to the laser printer, to lasick eye surgery. All told, more than 55,000 patents have been granted in the U.S., stemming from that first MASER—generating billions of dollars, and improving lives in countless ways. Making patents clear—so that we foster more innovations like the laser, and their important follow-on improvements— was a goal of our country’s founding fathers as they designed our IP system. 

But how to make patents clearer is a difficult problem.  The U.S. has been issuing patents for about 225 years, and the history of American innovation tells us the USPTO has, overall, done a pretty good job. But we can always do better.  So it’s been a privilege for me, as a public servant, to bring my experience in technology, law and business, and my MIT education, to bear on how to do better.

Now it would be easy for anyone in my position to simply talk about improving patent quality, because improved quality sounds good. But I see my mission as the leader of America’s Innovation Agency to drive change that is both good and sound. The key to good and sound 21st century decision-making lies in easy access to, and wise use of, data. We have decades – no, centuries—of patent data at the USPTO. So I challenged a small internal team to act like a “start-up” and develop some new ideas on how to use the vast reserves of data the USPTO gathers to help solve some of the agency’s age old challenges. It’s what the companies in the Silicon Valley, from which I come, have been doing a lot of recently.

The first “use case” my team decided to address with a big data approach was “consistency” of patent examination. A quick look at the volumes of information we handle will suggest to you why that’s an area perfect for this type of approach: We have about 8,300 patent examiners, each handling numerous cases; we receive about 600,000 applications each year; and we have about 1 million applications open and pending at any given time. So we have lots of data, and a need to find the best way to ensure that all our examiners are applying a complex set of laws as accurately and consistently as possible. Our team is currently developing algorithms to identify trends in patent examination and areas for improvement, which will result in further targeted training including potentially by technology area. Even with this, we can still do more with our wealth of data—for example, supplementing the examiner’s manual searches of prior art with automated searches using artificial intelligence techniques.

Having spent nearly a decade at Google and having worked at the MIT AI Lab, I know the PTO can do more to leverage the recent advances in automated search and AI technology to help identify the most relevant prior art quickly and efficiently, leading to a better quality patent. So you see, Hal, my fine training at the MIT AI Lab was not for naught!

Secondly, the USPTO has vast reserves of scientific knowledge contained in the millions of patent applications we receive. If you think about it, the USPTO might very well possess the world’s largest repository of innovation since innovations across the globe are often filed at the USPTO given the importance of the US market.  This is no accident—it’s part of the fundamental bargain in the Progress Clause of our US Constitution:  In exchange for disclosing your invention to the public—including what it is and how it works—you get exclusive rights for a limited period of time. In this sense, the USPTO has been in the business of “open data” since the very beginning of our agency’s history, when President Washington signed the First Patent Act into law in 1790. But while this information was always “open” in the sense that issued patents were available to the public, it wasn’t always easily accessible. For most of our history you had to flip through a bunch of paper documents to find what you needed. In fact, our stacks of patents were filed in what we called “shoes,” which refers all the way back to the man who was, in effect, our first patent examiner, Thomas Jefferson, who kept patent applications in shoe boxes.

Times have changed—fortunately, quite profoundly. And, at the USPTO, we are trying to harness those changes to make the scientific disclosures that have always been publicly “available,” actually available in a far more meaningful and easy-to-use manner. Shortly we’ll be releasing APIs to our patent data for the first time.  Anyone with even the most basic programming experience will be able to explore our data according to their own interests, curiosity and business needs. This makes it easier for innovators—from researchers to entrepreneurs to well-established companies—to mine this data, helping to inform where they spend their limited research and development resources, and providing a much more detailed view of the competitive landscape than previously available. If you think about it, one of the first things an innovator does after an invention is file for a patent, so the USPTO data can provide very current information of the competitive landscape.

In addition to releasing APIs, we’ll be launching the USPTO Open Data Portal, a developer’s hub –  establishing a shareable, and “social” platform, for anyone in this community to showcase unique ways they’re using our data, and combining it with the world’s vast array of datasets, such as economic and geographic data. This unique forum will not only leverage the power of the crowd to unlock our data to answer questions about trends in technology and innovation but it also functions as a forum so users of our hub can provide input to the USPTO on other types of data sets we should release.  This sharing, and feedback loop, will allow others—including the USPTO—to extract new value from our data reserves. This will allow all players in the innovation ecosystem to have better information to make smarter decisions. And this just doesn’t sound good, it is good and it is sound. And, we could really use the kinds of fresh ideas and insights that the MIT community excels at, so do check it out and let us know what you think. 

Which brings me to the final thoughts I’d like to share with you today, before we begin our Q&A session. Fifty-five years ago, a native son of Boston asked all of us to do more-offer more of ourselves in service to our shared future. President John F. Kennedy challenged us: “ask not what your country can do for you, ask what you can do for your country.” His call to public service still echoes today. It’s timeless. Now, as then, we still need Americans willing and able to get involved.  To consider public service, if even just for a time.

I was drawn to public service after a 25-year career in the private sector because I wanted to give back to a country that has given much to me and my family. In an increasingly global and competitive economy, we cannot take for granted that the United States will remain at the forefront of innovation. Taking on the leadership role at the US Patent and Trademark Office  has given me an opportunity to help, in a small way, ensure that our country remains at the forefront of innovation by having policies, programs and procedures that encourage invention. 

And it’s not just me—this administration has, in fact, promoted the need for those with scientific and technical backgrounds like yours and mine to serve in government and help inform and guide public policies. One example of this in action is the Presidential Innovation Fellows Program, which brings some of the most talented technologists from across the country into the government for one-year stints, working directly with leaders to address some of the biggest challenges facing our government. For example, it was Presidential Innovation Fellows from such companies as Google, Microsoft, startups incubated at YCombinator, and even Athena Health based just a few miles away in Watertown, Mass, who ultimately pitched in and helped resuscitate when it was facing enormous challenges. 

This kind of public service offers opportunities not only to work on some of the most interesting and important issues of our time, but to have a positive impact on our country and the world in ways that you sometimes don’t get in the private sector. It’s not just about bringing a specific scientific or subject matter expertise to the discussion—although that doesn’t hurt!—it’s about bringing different approaches to solving problems.

So I invite those of you who hadn’t thought about working in public service to consider it – now and in your future. And, while I’m still Director of the US Patent & Trademark Office, I’ll continue to apply the lessons I learned here at MIT toward supporting an intellectual property system designed to ensure that the greatest innovations are yet to come. Thank you.

#   #   #