Friday, August 09, 2019

IPSC: Closing Plenary


Stephanie Plamondon Bair, Brigham Young University J. Reuben Clark Law School
Innovation’s Paradox

Innovation begets innovation in a virtuous cycle … at least sometimes. Not all innovations are productive, which is fine; it’s trial and error. But some innovations may inhibit our potential for future innovation: how?  Neuropsychology: certain innovations may change the way we think in ways that make it more difficult to think creatively.  Attention span and filtering out unwanted information: focus on a task and selective filtering are both important to innovation, and heavy digital media use seems to interfere with our ability to do that.  Processing information differs too: our brain processes digital information differently from hard copy. People are more focused on concrete details than to think abstractly for digital relative to analog, and abstraction is a necessary component of creative thinking.  Screen time may impact empathic accuracy, which is also important to creativity.  Creative tools: photoshop etc. Are they unmitigated wins for creativity? Downsides: default settings on programs may channel creativity in particular directions—users of one music editing software program are very likely to produce 120 bpm tempo songs b/c that’s the default while users of another are very likely to produce 140 bpm tempo songs b/c that’s the default.  Also concerns about whether people can produce crap: “just b/c tech  has made it easier to create and express yourself does not mean you’re any good at it”: Hegarty on Advertising. 

2012 analysis of creativity scores show declining scores since 1990 while intelligence scores have gone up. Over last 20 years, “children have become less emotionally expressive, less energetic, less talkative and verbally expressive, less humorous, less imaginative, less unconventional, less lively and passionate, less perceptive, less apt to connect seemingly irrelevant things, less synthesizing, and less likely to see things from a different angle”: Kyung Hee Kim, The Creativity Crisis. Not causation argument, but correlates w/rise of tech. [And with massive explosion of inequality/precarity?]

Tech/innovation: we tend to think about them as passive tools that we can use or not as we see fit. But innovation can be acting on us, in ways that might impact our ability to continue to innovate.  Not new idea to philosophy literature.  Even if you’re practically minded, it’s an idea worth grappling w/ b/c of potential for negative effects on innovation going forward.  Also connected to literature on quality of innovation.  IP rights may incentivize low-value innovations more than high-value innovations; this dynamic she identifies here may be another social cost to grapple w/.

Has focused on costs and not benefits, which are undoubtedly present. Also it’s not necessarily the existence of these innovations, but the way they’re used/overused/misused.  California required all textbooks to be available digitally; Florida followed suit; state educational budgets have been devoted to getting tablets into public schools. We should have a conversation about the downside of these policy decisions.  Mo. Sen. Josh Hawley’s bill targeting social media companies; he wants to require them to get rid of features that might lead to overuse of their products like infinite scroll, autoplay.  It’s an intriguing idea in this context.

RT: inequality/precarity; also want to know how much of this research relates to WEIRD subjects versus others.

A: bound by what the existing research says; we might want different policies for developing countries. Just wants to be clear that she is vigorously extrapolating.

McKenna: you’re talking about two different kinds of studies. One tests microeffects of particular technologies. One looks at macro effects on children over extended periods, and causation is hard. Our education system has shifted to testing drill & kill over time and it’s no surprise kids are less creative in response.

A: meta-analysis on creativity shores up the results, but most literature she’s relying on is testing for particular technologies.

Charles Duan: look at historical examples: people said similar things about television, the printing press, etc.  [I’m very sympathetic to this criticism and I very much want it to be true but on the other hand people have predicted climate apocalypse for hundreds of years and this time they seem to be right (and also we have a lot more empirical information than we used to about both of these topics).]

Q: There is some work on autistic people in the patent system. 

Q: innovations you point at are general purpose or platform technologies, or both.  There are literatures about how these technologies diffuse and affect development of downstream tech.

Rosenblatt: how do you view the relationship b/t progress and path dependence? 

A: interesting things in philosophical literature on tech as monolith that moves without our direction.

Justin Hughes, Loyola Marymount University Loyola Law School
The Law of Digital Avatars (or Replicas)

His definition: digital manifestation of person.  Convincingly lifelike representations of people are currently 2D but will be 3D soon.  They’ll make banking and travel more tolerable; they may provide companionship to elderly, infirm, or lonely.  May also convincingly replicate people we know.  Pandora’s Box: the person represented may never have said or done the things we will see.  Stunning range of potentially relevant legal doctrines, from ROP to fraud.

Kim Kardashian and Barack Obama have been convincingly replicated; could be used to deter voting among African-Americans by telling them voting hours have changed, etc.  Deepfake porn, though need not be porn (e.g., fake crime). The question of whether some sort of lower bound of authenticity is required to care is key.  E.g., 2018 NY proposal started by covering realistic depiction and then that was amended to remove a realism requirement.  Grandparent/military romance scams will proliferate.  Hologram actors will replace human actors. Xinhua news agency launched a digital news anchor, which will say whatever you want it to say.  DeepNude: worked on a single picture of a woman (it only worked on women) to show “her” nude because it was trained on images of naked women.  Single pictures will soon be enough for a lot of things.

Is there a possible common approach that would be sensible, helpful, and constitutional? It is generally thought that false designation/defamation won’t work on deepfake porn because that mostly has disclaimers.  Revenge porn laws used against digital replica porn?  Depends on the wording of the statute. Connecticut law clearly requires actual photo, film, or other recorded image; Penn. law says visual depiction defined as representation by film, including but not limited to photo, video, film or computer image.

Digital replica porn proposal in California—predicts that something like this will pass soon—carefully includes digitization. Creates liability for intentional creation w/knowledge of lack of consent; victim has to prove some harm. Despite those limits, seems to cover deepfakes of celebrities, digital replica revenge porn, and unconsented enhanced sex scenes in feature films where actor didn’t agree to what was shown on film.

NY proposal on digital replicas isn’t focused on revenge but on employment: provides a right to control digital replicas if they show person performing activity for which s/he is known in the form of a fictional character: the idea is to permit unconsented depiction of digital avatar of Hugh Grant being arrested but not unconsented depiction of digital avatar of Hugh Grant playing Wizard of Oz.

Federal proposal from Sen. Ben Sasse: define deepfakes & create offense of distributing it with intent to facilitate criminal or tortious conduct. Tries to address 1A issues including US v. Alvarez by carving out anything protected by 1A; may have to rethink Alvarez to address the dystopic future.

Rothman: (1) you talked about 1A but not about copyright conflicts. (2) Should you really think of all these issues as amenable to common solution?

A: grappling with (2). Simplest global issue: a ban on unauthorized digital replicas, period.  That’s not a viewpoint restriction, probably a content based restriction (though he says you could conceive of it as a time place and manner restriction, which sounds wrong to me).  But Sasse’s approach seems close to a globalized solution.  [That seems wrong, at least w/r/t revenge porn—the Sasse proposal is targeted at fraud, but the revenge porn issue is that there isn’t particularly a cause of action for many instances of revenge porn unless we add a new law.]

Mike Carroll: disclaimer requirement?

A: convincing deep fakes with disclaimers may still fool people; disclaimers often fail. TM law doesn’t favor disclaimers [but First Amendment law does].

Said: actual v. constructive knowledge in Sasse proposal—wondering what the thinking is.

A: probably not much thinking as yet. 

Betsy Rosenblatt, UC Davis School of Law
IP and the Question of Who Creates

What’s the difference b/t an inventor and a technician (e.g. calculators in Hidden Figures)? What’s the difference b/t an inventor and a mechanic?  Is a scrapbooker an artist?  Who is a filmmaker?  What biases could affect our answers to these questions?  Law is a particularly powerful voice in this discourse b/c it speaks w/the power of the state.

IP systems create a discourse of value, assigning it to particular creations by giving their creators exclusive access to markets. Promotes (only) creation of things w/predictable market value, covered by IP. Lots of other things get made, but IP only promotes some (negative space literature). Influences what kinds of things get made—Kate Darling and Sprigman/Raustiala on adult entertainment; Fumi Arewa on Nollywood. Low-IP spaces seem to feature fast turnover and live performances. Higher protection results in more investment in polished products.

But that’s all about the what. She’s interested in the who. Critical theory/critical race theory: the IP system isn’t neutral in application and effects, structural benefits to wealthy/powerful.  Everyone is a maker of sorts but some of that making is special, recognized by law/society as meaningful. Foucault described the author-function as a discursive mechanism by which we grant respect/authority to some people and not others. Distinct from authorship in © law—under © everyone’s an author—but not all authors benefit from © discourse.  We can extend author-function to inventor-function, brand-function (legitimizes brand owners over consumers even though both contribute to meaning).  Foucault envisioned recognition of dialogic relation as erasing the author-function, but that’s not our world.  IP law picks some practical winners & losers as well as winners & losers in discourse by elevating certain voices.  Reflects & reinforces hierarchies of race, class, and gender. She’s not the first to observe this.  Wants to add that IP law may also concentrate cultural respect in certain hands.

Norms work well w/in communities and not very well in regulating interactions b/t communities.  In these situations we need to be especially mindful of opportunity for exploitation of less powerful by more powerful who disrupt norms that are working. Not just where norms govern subcultures (street artists need protection against developers, as with 5Pointz/VARA claim), but also where things like norms of agent submission protect the film industry from random people who send stuff in and then sue.  That last norm exists because of law. Lots of specific doctrines favor those who control the means of production.  Product patents being more powerful than process patents: but who comes up with new uses for existing things? Those who have to make do what they have.  The access/similarity test favors the famous. Fair use—see Andrew Gilden—favors the famous. Ownership/authorship standards giving mastermind the ownership do similar things, as does WFH.  Effort of users redounds to the benefit of TM owners. 

What’s new here? The way this feeds our discourse about who is doing things of value. We end up revering visionaries who may not be the people who should be revered. Edison and Musk hired a lot of innovative workers whose names we will never know. The garage inventor is more likely to be a white man because being a garage inventor requires having a garage and we live in a world with redlining/massive wealth extraction from African-Americans. Those who live in a world of precarity are considered amateurs and considered bad.

But doesn’t IP allow those with humble beginnings to rise and thrive?  In many ways IP makes it harder for those voices to rise, making it harder to get out of the amateur box. At the beginning, when you don’t have many resources, IP stands in your way.

What is to be done? Has some, including small doctrinal fixes, some of which might fall into the category of more rights: attribution, termination. There may also be places where less makes sense. But also wants to change discourse about who matters as a creator.

Maggie Chon: good to hear about justice, power, representation.

RT: The concept of respect explains why your proposal isn’t to decide all disputes in favor of the less powerful person before the court. But what if it’s really all about labor law and tax law? [Lemley adds antitrust law, but I’m not sure we’d need antitrust as badly if FB and Alphabet and Amazon didn’t have so many billions sloshing around to be used to crush competition where necessary.]  FB could buy out innovative competitors b/c the tax rate is too low.  What if you only change the superstructure by changing the structure?

A: Catherine Fisk’s work on labor is really instructive. Labor law has a role to play here. But we should also be thinking about innovation law specifically.

Zahr Said: distinguish discursive harm from legal harm.  Some people will distinguish the two.  What about remedies?  Even if on the merits a court finds for a party, the better-resourced P gets so much more out of the process (or the D gets to deduct a bunch of expenses).

A: Costs of acquiring/enforcing rights is a huge piece of this.  Defenders say that marginal value of money means a small recovery for a small P is fine, but she’s not convinced.

No comments:

Post a Comment