GeoLegal Weekly #44 - AI, the Future and Tech Rule-Breaking with Jae Um

Jae Um and I go deep on the impact of AI on the law and society, which inspires me to think of tech rule-breaking as similar to "California stops."

Elon Musk’s last minute burst onto the US election scene got me thinking about technology company attitudes toward regulation. The world’s richest man could have stopped at donating wads of money to Donald Trump but instead he also used the big tech playbook of network effects and gamification. He harnessed the first technique by offering referral fees as direct cash payments to citizens who got others to sign his petitions. He used the second by offering up a sweepstakes prize of $1 million for one signer per day, surely knowing that gamification promotes even more uptake.

But, in the process, he demonstrated a third strategy - one which has much greater implications for society: He proved that he could move faster than regulators. His tactics may or may not have amounted to legal activity. But it doesn’t really matter: The election is decided and he has money to pay any resulting fines (which are even less likely since his team now controls the government).

Buckle up because the age of AI will be characterized by this approach to disregarding rules at a pace we can barely fathom. The US, one of the leading AI superpowers, will be headed by a CEO president, with a venture capitalist Vice-President and Musk somewhere in the wings. The implication is not just one of light touch regulation - it’s one that will ultimately reward bare-knuckle market participants with power because they have market share. Remember, all of Musk’s ventures, from Paypal to Tesla and SpaceX to his approach to Twitter, have been successful expositions of outpacing regulators. Those regulators ultimately did have opinions on financial flows, autonomous driving, or private spacecraft but often too late to make a difference.

Rapid Speed of Diffusion

To better understand this, I spoke with Jae Um, Founder and Executive Director of Six Parsecs. Jae is one of the top minds in the world on the pace of technology change and its implications for the law. We couldn’t help ourselves but spend 30 minutes on a whole range of topics including the legal sector and its discontents, access to justice and innovation, all of which you can watch here or below.

As Jae tells me, legislators and regulators fell far behind with the onset of the internet, which enabled new business models and unfathomable scale. This becomes a much bigger issue with the advent of AI. After all, many jobs at OpenAI net around a million dollars per year and top government executives earn a fifth of that. It’s hard to see how the market for talent reaches an equilibrium in favor of smart regulation.

While most of us think about the challenges posed by AI as having to do with unexplainability or with the fact that government doesn’t understand these emerging technologies, there is actually a less well-appreciated risk vector at play.

Jae raises the insightful point that one of the biggest challenges we are seeing is that technology incumbents are using their sizeable war chests and existing go to market channels to get new AI products into the market at an incredible speed. While there will be low-end disruption of existing business models by smaller, more nimble upstarts, the general thrust of AI is that the biggest companies in the world are falling over each other to bring complex, disruptive technology to the market through the world’s most established sales and product channels.

As Jae notes, the speed of innovation means we are facing a lot of unknown unknowns as technology businesses are largely unconstrained and in a risk-taking position. When billions of people start talking to AI without understanding the risks - as is already happening - it will be tremendously challenging to put in place safeguards that undo or limit harm. Look at social media and the challenge of misinformation. The challenges are now obvious. But doing anything retrospectively to stop them without trampling on free speech and consumer choice is less so.

Jevon’s Paradox

Jae and I also spent time discussing Jevon’s paradox, the concept that as the efficiency of an input goes up, its usage also goes up. When Uber disrupted the transportation market and brought trip prices down, my personal transportation spend went up because efficient car service was a better alternative to other forms of transit. This created a lot more jobs driving people in cars than the taxi cartels of days past. This is a reason to cheer, right?

Not from a political perspective as AI devours traditional jobs. As Jae points out, while we may see an aggregate increase in the amount of usage of, say, legal services, the abrupt change will leave many of the former practitioners unemployed. Try telling a London cabbie whose take home pay dropped due to competition with Uber that they should cheer innovations that have boosted the aggregate money spent on transportation. You’ll have an uncomfortable journey.

The upshot of this is that a bunch of people currently employed in high-status, well paying professional jobs like legal services are going to find their jobs at risk as AI devours them. Sure, in economic theory textbooks, winners can compensate losers. But good luck finding a politician campaigning on universal basic incomes to support the holiday villas of out of work junior partners.

Fighting Back

So, if technology is outpacing regulation, being distributed through high-velocity go to market channels AND a lot of existing professionals are going be knocked out of the work force by this change, there will be a dangerous brew of societal discontent in coming years. Rather than dive into that today, what you really should be asking is how you can prepare your company and your employees not to fall victim.

Technology companies don’t drive untrammelled through regulatory stop signs; that’s a good way to crash. Instead, they do a series of “California stops” - slowing down as they approach the stop sign, stopping only if danger or authority is obvious, and proceeding without a full stop if not. Sure, when they get caught there may be a ticket to pay if the officer doesn’t accept their excuse (see below). But they realize they are in a winner-take-all race and that type of strategy is optimized for beating risk-averse competitors who are stopping and looking both ways at every regulatory sign.

What you probably don’t realize is that even the most analog businesses are becoming tech companies whether they like it or not. Just like the distinction between online and offline companies makes no sense in today’s world, the distinction between AI and non-AI companies will fail to be helpful soon. If you are approaching AI tools with total caution or advocating for regulators to hold back tech-driven competitors, you will be beaten by competitors who adopt a technology-first mindset today and grab market share that they then weaponize against you.

You will be like London cabbies trying to close ranks to throttle Uber rather than asking how they can use new technologies to win in the first instance. Or you may be like Democrats playing low-risk, low-reward strategies against no-holds-barred rivals.

In Other News

If you want to see me in the hot seat for a change, check out my turn on Tyler Finn’s Abstract podcast for SpotDraft here and below. We go deep on how political forecasting works, how Hence is working to link geopolitics to legal, and why my biggest work pet peeve is business meetings without decaf.

That’s it for this week.

-SW