💊 PATENTS AND INCENTIVIZING INNOVATION
Tuesday brought decisions from both patent cases argued in front of the Supreme Court this term, Oil States Energy Services v. Greene’s Energy Group and SAS Institute v. Iancu. At issue in both cases was the process of inter partes review (IPR), which was instituted in 2012 after the passage of the America Invents Act. The process allows a competitor to bring call for an IPR, which, if institute, initiates a trial-like process engaging both parties. The USPTO has the sole discretion in reconsidering and potentially invalidating the patent at issue.
In Oil States v. Greene’s Energy, the question before the Court was whether the adjudication of IPR petitions by the USPTO is an exercise of the “judicial power” that under Article III of the Constitution can be exercised only by courts. The Supreme Court’s resident patent expert, Justice Clarence Thomas, wrote the opinion, holding that IPR does not violate the Constitution. Thomas wrote that because patents are a public right and the IPR process is simply a reconsideration of that grant, Congress has the authority to conduct the reconsideration. There has long been an academic argument about whether patents are public (e.g. statutorily created) rights or private (e.g. common law) rights, with Justice Thomas falling on the former side of that debate. But, Thomas was careful to articulate that the decision should not be misconstrued as suggesting that patents are not property for purposes of the Due Process Clause or Takings Clause.
According to Court documents, the IPR process has been used to invalidate some 1,300 patents since 2012. Tech companies are largely please with the decision, as it allows them to continue to use the IPR process to challenge patents instead of having to undergo long, drawn out litigation against alleged patent trolls.
But, pharmaceutical and biotech companies are more worried about the ruling, fearing it may open the door for competitors to more easily (and cheaply) invalidate patents through IPR. This gets to the difficulty of treating two fundamentally different industries (tech/internet companies and pharm/biotech companies) with different business models the same when it comes to intellectual property.
For modern tech companies, potential returns are much greater. Modern internet companies are driven by network effects that can generate massive returns to scale for innovation. They’re able to spread fixed costs over a massive user base that can scale quickly. Meanwhile, Moore’s Law (yea, I went there) implies that R&D cost, on a per unit basis, continues to decline for computing-based R&D.
On the other hand, evidence suggests that biotech and pharmaceutical R&D is only getting more difficult and more costly. Some have even referred to drug discovery as operating under a “reverse-Moore’s law,” and while machine learning may facilitate the drug discovery process, this promise is largely unfilled as of yet.
The Constitution gave Congress the power to “promote the progress of Science and and useful arts,” but the reality is spurring innovation requires different tactics in different industries. We can quibble about how strong patents should be in the biotech and pharmaceutical industries, but the heavy capital investment required to innovate in these sectors necessitates some statutory protection. But, the case for government-granted monopolies (patents) for software innovations is much more tenuous.
While we wait for the merging of software and biology (or, for software to eat biology), it’s important to recognize the different capital structures at play for different industries and structure incentives accordingly. We don’t need the government to incentive investment in the next “one-click buying” patent, but we do need to incentivize continued commercial investment in biological R&D. Government-granted monopolies are an impure means to an end (innovation), and it’s important to critically evaluate when and where they should actually be issued.
The second, less exciting opinion held that when the USPTO institutes an IPR process to review an already issued patent, it must decide the patentability of all claims the petitioner has challenged.
🎧 SPOTIFY THE HITMAKER
On April 24, Spotify announced its new free tier with a few major changes. First, it’ll recommend music to users on the fly using its machine learning. Second, and more importantly, it’ll let free-tier users listen on-demand to whatever song they want, as many times as they want, if the song appears on one of Spotify’s 15 personalized discovery playlists. These playlists account for about 750 tracks in total.
As I wrote after Spotify filed its F1, its success depends largely on its ability to cut out the music labels, whose content currently accounts for 87% of content streamed on Spotify. With the overhaul of its free tier, Spotify is attempting to mint new hints, allowing the 90 million users of its free tier to listen to certain songs, selected by Spotify, as often as they want. Spotify’s success depends in large part on its ability to serve as a platform for new or up-and-coming artists, helping users discover new artists while at the same time helping new artists get discovered. It’s a beneficial relationship for both listener and musician, with Spotify sitting in the middle taking a cut of the transaction. Providing free users access to unlimited listening of select songs drastically increases Spotify’s ability to be a hitmaker.
Apropos the patent discussion above, note that while Spotify has patents related to its machine learning and recommendation algorithms, these may not be essential to its business. The fact that it has 160 million total users makes the cost of acquiring these patents relatively trivial. Either way, the data Spotify continues to gather on its users is much more valuable (and protectable) than any patent it may or may not have acquired.
🎬 NETFLIX THEATRES?
From one subscription media business to another. Reports have emerged that Netflix is looking for ways to get into the movie theatre business. While it may not be interested in Mark Cuban’s Landmark Theatres anymore, it sounds like Netflix wants to get some of its productions in theatres so it can qualify for awards. Jeff Bezos has long been obsessed with garnering awards with his Prime Video service, so it seems he’s not the only one with a complex (the Prime Video section of his shareholder letter is basically one long sentence about how awesome Prime is because of all the meaningless awards its various shows received).
Of course, Netflix (and to a lesser extent, Prime Video), are successful illustrations of the “laddering up” strategy that Spotify is now pursuing. That is, they built businesses off of licensing copyrighted content from TV and movie studios, building a war chest to invest in their own original content and cut out the middle man.
But, there may be differences between music and video that make this strategy harder for Spotify. Namely, old music still holds a ton of value, while old TV shows and movies are less valuable. I love to bump the Beatles, but only watch I Love Lucy when I’m home during the holidays and my mom makes me.
🦅 Bird’s Eye View. In China, facial recognition technology was used to spot a man in a crowd of 60,000 concert goers.
💡 From the Printing Press to the Internet. A chart showing productivity from 1440 printing press to today. Other academics have studied causes of the productivity slowdown since 2004, illustrating it’s more than just a “measurement problem” (Business Insider, Marginal Revolution).