This week on IPWatchdog released I speak with Josh Harlanwho is the founder and managing partner of Harlan Capital Partners, a private investment firm based in Palm Beach, Florida, focused on asset-based investments, including investments backed by intellectual property rights and the media, and in particular the monetization of streams. income recovered by intangible assets. You can listen to our full discussion on monetizing revenue streams backed by intangible assets wherever you get your podcasts (links here) or visit IPWatchdog released on Buzzsprout. We also published a full transcriptand you can also watch the video on YouTube.

During our conversation, we talk about the monetization of various investable revenue streams across media, sports, and intellectual property, and we also make an important pivot to later discuss artificial intelligence and what should be considered those looking to invest in AI. We also discuss the viability of fair use defenses currently being advanced in ongoing lawsuits filed by copyright owners against AI developers and the likely future market for monetization of training data and AI results.

As you will hear during our conversation, Harlan is considering the possibility of awarding fairly significant damages on a one-off basis to content creators and their copyright lawsuits against large AI developers , but he doesn’t see this as an existential threat to the AI ​​industry. itself, which it sees as generating significant amounts of revenue with large players who will have the ability to pay for content, both on the input and output side of the equation. And without providing specific details, Harlan confirms that he is currently exploring opportunities related to training data aggregation, which he believes could be very interesting from an investment perspective.

Monetize asset-based revenue streams

“When I started working with ICM, the film and literary agency…I went into it expecting the business to be quite volatile from year to year,” Harlan explained. “But when we went under the hood, what was actually the case was that they had… accumulated all these final royalties or stakes in movies and TV shows. I was working with them in the mid-2000s, and they were still cashing a check every year from the first Star Wars, and they hadn’t represented George Lucas since the ’70s. They were cashing a check from Pretty Woman every year, and they weren’t representing Julia anymore. Roberts since the ’80s. So it was really striking how these types of income streams could be long-term, very predictable, very annuity-like. . And that impressed me a lot and had a lot to do with the development of the strategy that I pursued in my own business.

Many lawyers might understand monetization of an income stream as what happens when you settle a case for your client and instead of collecting payment which may be annual payments for several years, the client sells the stream to an investor. at a reduced price and receives a lump sum.

According to Harlan, the situation is in some ways similar, but not entirely comparable to apples. In case you have a certain guaranteed income stream, it is easy to determine the appropriate discount rate based on the value of the long-term income stream and calculate a discounted present value that can be paid immediately . And while this type of deal can be done with intellectual property-backed revenue streams, it is also common for it to invest in asset-backed revenue streams that are less certain, or even speculative. For example, when you acquire distribution rights, you then actively configure them on different platforms and see how they perform on those platforms. This presents a different type of risk profile, as a security can underperform, which must be evaluated from a risk-reward perspective.

“We see these things a little more holistically than other people,” Harlan says. There are royalty funds dedicated to medicines. They will only invest in pharmaceutical royalties. There are funds dedicated to music royalties. They will only invest in music royalties. We like to look at it a little more holistically and look at a whole range of opportunities related to media, sports and intellectual property. In fact, we have focused a lot in recent years on media, sports and intellectual property. Historically, we have also made investments related to litigation finance, insurance-related investments and several types of hard assets in which we have invested. But about 80% of the capital we’ve deployed over the last two years has been media related. , sport and intellectual property. We see a lot of transactions there.

Focusing too much on AI technology

“I think people spend too much time thinking about technology and not enough time thinking about the legal environment around technology, particularly intellectual property issues and intellectual property issues related to AI . on the production side,” Harlan explained. “It reminds me to some extent… of the battles over piracy in the entertainment industry in the ’80s and ’90s.”

Harlan explained that with music, record companies release unencrypted CDs and face a huge privacy challenge. Learning from this mistake, the film industry went to Congress and won protection from the Digital Millennium Copyright Act, which makes it illegal to circumvent technological measures put in place to protect copyrighted works . So even if the protections put in place on DVDs were weak and could be broken, “no reputable hardware or software manufacturer would make it easy for you to copy a DVD,” Harlan explained. “At one point, technology didn’t really stop anyone from copying DVDs, but the law made it much more difficult. So, in my opinion, this is a lesson for the entire AI economy, in that law tends to trump technology.

Harlan sees the possibility of “some pretty significant damages on a one-off basis” for content creators in their copyright suits against big AI developers, but doesn’t view these legal issues as a threat existential for the AI ​​industry.

“There will be some sort of market clearing price for all of these things to be used as training data, whether it’s text images, videos, geospatial data, etc.,” Harlan predicts. “There is no doubt that what AI brings to the world will be of tremendous value. So there will likely be companies that generate a lot of profit, meaning they will have the financial capacity to pay large sums of money to license the data. But it suggests to me, if that’s the case, that there will be some pretty interesting markets for aggregating and distributing this kind of data.

And without providing specific details, Harlan confirmed that he is currently investigating opportunities related to training data aggregation, which he believes could be very interesting.