Deven R. Desai

The Centralization Paradox or How To Be A Good Counsel

Lawyers, at their best, are counsel. When technology changes or disrupts the core economic and social structure of fields and industries, it also challenges a counsel’s ability to do her work well. On one hand, a counsel may see her job as simply aiding a client’s drive to be the winner in the field, in part by pushing the edge of the law or arguing for deregulation. On the other hand, just because a technology is cutting edge does not mean it is gospel.[1] Society can reject even good science, if the social cost is high. Good counsel thus must understand its client’s business sector, the technology at hand, and the law that does, or could, govern the business and technology. Even then her work is not done. Good counsel looks into the future and asks not whether something can be done, but what the risks are. Ultimately, good counsel presents clients with insights about whether it should be done. These realities mean good lawyers must not shy away from STEM. At the same time, they must not embrace STEM to the exclusion of economics, sociology, history, and ethics, lest they stop being counsel and become cheerleaders.[2]

Although these ideas apply for any counsel, they become acute when a practice becomes decentralized and democratized.[3] Lawyers are trained in a system that assumes large, centralized players. In simplified terms, law tends to work on a business-to-business model. Despite all the celebration of deregulation and decentralization, when society moves from relatively few firms to many players, the way the law works shifts. In some cases, the economics of production and invention mean that large-scale entities like recording companies, hotel chains, taxi companies, and even automakers may be less necessary. But just because technology, especially digitization and networking technology, alters a field does not mean that a new centralized player or players will not emerge. The emergence of new, centralized players is obvious: think of AirBnB, Amazon, Apple, Facebook, Google, Lyft, Netflix, YouTube, or Uber.

The dynamic of democratized technology enabling decentralized production and regulation evasion leads to two questions a good counsel must answer: When does decentralization recentralize? What can be done when a powerful practice remains stubbornly decentralized?[4]

With recentralization, an industry that once was the darling of innovation and quasi-rebellion will be asked to fill the gaps it created. Issues of trust, safety, and sometimes the enforcement of legal and social norms do not go away after a new player upsets a possibly obsolete business model. They remain, waiting for competitors and society to demand that the new winners address the issues. Providing trust systems becomes a powerful way to assure the public that regulation is not needed—and to make money. Offering malware-free music, verified design files, background checks on drivers, dispute resolution, and safety assurance for rooms and goods has allowed Amazon, Apple, Uber, and AirBnB, respectively, to provide value despite the transaction costs. With millions or billions of transactions, they can absorb the costs, provide a marketplace, and in essence charge for being a clearinghouse service. But not all self-regulation is self-motivated. Pressure from the music industry, and perhaps a desire to enter that sector too, forced YouTube to find a way to police copyright. eBay has ramped up its anti-counterfeiting efforts. Regardless of why self-regulation occurs, once these actors attain scale and show that they can regulate, society may ask for more regulation.

With networked, monitored industries, the capacity for public or private control increases, and the ability to make money changes. For example, autonomous driving systems should reduce accidents and also ensure that cars almost always obey laws. Thus “driving while black” should vanish, because neither the driver nor the police officer will have discretion about whether an infraction occurred. If all cars obey traffic laws, almost all revenue from enforcing traffic laws will go away. Law enforcement could focus on more serious crime, but local police force budgets reliant on traffic violations for revenue would face gaps. Insurance rates should plummet, which is good for consumers, but the related industry would face revenue shortfalls. All these possibilities require firms and their counsel to understand the technology that drives a given change, as well as the social, moral, and economic issues that go with the change. Failing to understand all these aspects of change can mean that a firm finds its assumptions about how the sector operates are incorrect, and so the firm’s business model is defunct.

Good counsel will have different perspectives on what to do depending on the client, but all counsel involved in these outcomes should try to see and explain the implications to clients as they plan what to do as a response to the changes. Thus, if the state demands that AirBnB, Lyft, and Uber meet safety, employment, and other regulations, they may object, but good counsel should also see that the companies are well-placed to use their scale to come up with solutions. As one example, after claims that a driver raped a passenger in India, Uber created a panic button for riders.[5] Yet, Uber did not deploy the panic button in the United States.[6] Uber claimed that the 911 call system is the panic button and “it would be ‘a stretch’ to try and do better than formal infrastructure.”[7] This moment might be seen as one in a long line of Uber’s missteps, but it reveals the idea of what good counsel can do. Rather than avoid responsibility or save costs, Uber could have added the panic button in the U.S. Indeed, had such a button been available to passengers of Jason Brian Dalton, who was charged with a spree of shootings and seems to have taken passengers for rides on the same day, perhaps more people would have reported claims of erratic driving, and he would have been investigated before his more violent acts. None of which is to say that the technology would have necessarily stopped Mr. Dalton.[8] The point here is the absurdist statement that Uber does not think it can do better than formal infrastructure, when that claim is a key part of its overall claim to existence. Good counsel must thus see that disruption may untether a client from current regulations, but that embracing responsibility and filling gaps created by disruption can help a company rather than hurt it. And because the disruptors who have taken over a large sector are already incumbents, they would be able to keep newcomers and future disruptors at a disadvantage. New players would have to comply with regulations and accompanying technological solutions, whereas previous disruptors did not.

One can debate what regulation is proper, but that does not change the fact that new centralization means new regulation is possible. Nonetheless some practices that are now decentralized and democratized might stay that way. Those areas create a different problem for law, society, and the counsel who work in those areas. Two technologies, additive manufacturing (3D printing) and CRISPR gene editing, are powerful decentralizing and democratizing forces that can have large effects on society and do not necessarily lead to recentralization, thus showing the limits of regulation’s reach.

Additive manufacturing uses files from a range of sources, but nothing requires the technology to be networked. Although the digitization of things has allowed almost anyone to have a personal or small business factory, the process can be done without the central platforms that drive Lyft, YouTube, and other platform players.[9] Indeed, 3D printer hardware and software are often open-source. Users can build and improve without a central, coordinating force. If one tried to lock down files, peer-to-peer and other strategies that fuel copyright file-sharing would take hold. Unlike music or film, where most users may prefer an authorized and fairly-priced file, those who wish to use 3D printers for weapons or pharmaceuticals will have a higher risk threshold. Issues around decentralized and democratized biological engineering tools are similar but pose larger threats. They also show how old models fail.

Responsible actors in genetic engineering saw the problems coming, yet drew on the centralized playbook to solve them. One group fully grasped that the ideals and methods of self-regulation that were born in 1975 rested on scientists and good faith.[10] This work understood that it is not only a few centralized players that matter, but that “a practitioner community proliferating globally”[11] is important due to the “increased ease of reading and writing genetic information.”[12] Yet even if one has centralized players and a practitioner community agreeing on best practices and ethics, with decentralized and widespread creators, “securing materials in a handful of established labs is not feasible.”[13] In short, control over inputs or “materials” works with a small number of actors but is not going to work today. Nonetheless, the authors called for regulation and a new, independent agency to address the new “scope and scale” of new biological technologies.[14] Although seeking an agency that knows the area and has power over it is not foolish, and other sectors have used the approach before,[15] it will not solve the problem that genetic engineering knowledge and tools are now low-cost. The difficulty of using the technology for possibly dangerous biological engineering is also low.

Good counsel must understand that these technologies operate outside of market discipline. The users or consumers of the end products care less about safety and reliability. Those who are near death or cannot afford healthcare are vulnerable and desperate in the deep sense of being without hope. Promises of gene therapies or low-cost pharmaceuticals would be tempting. Yet those outcomes affect the consumer rather than creating a large-scale societal change. In contrast, anyone who wanted to use CRISPR to control pests or plants in the yard or on their farm might change an entire species and related ecosystem.[16] The protocols, agreements, and regulations of an industry group, a country, or a consortium of countries would not bind such an actor.

As counsel, simply backing calls for innovation and markets in these situations will not carry the day in the long run. Understanding the underlying technology and how it affects society allows good counsel to educate the public about the changes. It also allows good counsel to provide good advice; that is to be a true counselor. A better understanding of technology permits one to say “yes, with some small changes” rather than “no, absolutely not.” Knowing when other companies have faced scrutiny, lawsuits, and obstacles or product cancellation regardless of whether the law permits the activity due to technology that was not well-understood or poorly designed to address social concerns[17] allows counsel to speak with authority on law, society, and technology. It is that combination that makes a lawyer into a good counsel who should be valued and trusted as she helps guide a company to lasting, rather than, transitory success.


Deven Desai is an Associate Professor at Georgia Tech’s Scheller College of Business.  Prior to joining Scheller, Professor Desai was an associate professor of law at the Thomas Jefferson School of Law. He was also the first, and to date, only Academic Research Counsel at Google, Inc., and a Visiting Fellow at Princeton University’s Center for Information Technology Policy. Professor Desai’s scholarship examines how business interests, new technology, and economic theories shape privacy and intellectual property law and where those arguments explain productivity or where they fail to capture society’s interest in the free flow of information and development.

  1. See Steven Shapin & Simon Schaffer, Leviathan and the Air Pump: Hobbes, Boyle, and the Experimental Life at Kindle Loc. 555–56 and at 8093–94 (2011) (examining the way in which knowledge is constructed and offering “Solutions to the problem of knowledge are solutions to the problem of social order.”).
  2. Julie Cohen, Configuring the Networked Self: Law, Code, and the Play of Everyday Practice 3–4 (2012) (examining the link between political and economic interests behind arguments for technological constructs and by extension arguing that technological ordering is not a given to which we should defer).
  3. See Deven R. Desai, The New Steam: On Digitization, Decentralization, and Disruption, 65 Hastings L.J. 1469 (2014).
  4. For one view of this dynamic as a good one, see Orly Lobel, The Law of the Platform, 101 Minn. L. Rev. 87, 92 (2016) (“[L]egal disruption by the platform economy should be viewed as a feature rather than a bug of regulatory limits”).
  5. See Davey Alba, Uber’s New Panic Button Beams Real-Time Alerts to Police, (April 30, 2015) [].
  6. Andrea Peterson & William Wan, Uber Has a Panic Button in India, But Don’t Expect It to Come to U.S., Wash. Post, The Switch, (Feb. 22, 2016) [].
  7. Id.
  8. Id.
  9. On the implications of the digitization of things, see Deven R. Desai & Gerard Magliocca, Patents Meet Napster: 3D Printing and the Digitization of Things, 102 Geo. L.J. 1691 (2014).
  10. Megan Palmer, Francis Fukuyama, & David A. Relman, A More Systematic Approach to Biological Risk, 350 Science 1471, 1471 (2015).
  11. Id.
  12. Id.
  13. Id.
  14. Id.
  15. Id. at 1472.
  16. See e.g., Brad Plumer, “Gene Drive.” Learn the Term. Because It Could One Day Transform the World, Vox, Energy & Environment, (June 12, 2016) []; see also Janet Fang, A World Without Mosquitoes, 466 Nature 432–33 (July 2010) (noting eradication of mosquitoes could affect Arctic caribou).
  17. Products involving privacy provide examples of launches followed by quick shut downs at losses that might be in the millions.