🤖 Whose 'AI'?
PLUS: A podcast from the archives that will contextualize the downfall of SBF & FTX
Hi, and welcome back to Untangled, a newsletter and podcast about technology, people, and power. In this issue, I untangle a few news stories and then open up the archives to share a podcast episode that will help contextualize the downfall of SBF, FTX, and the dynamics that make crypto more susceptible to collapse.
I’ve been on the road the last few weeks, attending:
The Trust & Safety Research Conference, hosted by the Stanford Internet Observatory;
The “Blueprint for an AI Bill of Rights: One Year Later” in Washington D.C.
Here’s a fun li’l picture from the plane on my way back to LA.
Anyway, I’ll be in Seattle next week at the University of Washington to honor Dr. Francesca Tripodi for winning this year’s CIP Award for Impact & Excellence. I had the privilege of serving on the selection committee this year (the team I led at Data & Society won the award last year) and I can’t wait to celebrate Francesca and her work!
But enough about me and my travels! Let’s untangle the news, shall we?
🤖 Whose ‘AI’?
Last month, Senate Leader Chuck Schumer’s AI Insight Forum brought together tech CEOs, civil society representatives, and researchers. This week, one of the researchers in attendance, Inioluwa Deborah Raji, wrote an incredible reflection on the forum, and offered this nugget of truth:
"The truth is, “AI” does not exist. The technology may be real, but the term itself is air. More specifically, it’s the heated breath of anyone with a seat across from the people with the authority to set the rules. AI can be the enthused pitch of a marketing executive. Or it can be the exhausted sigh of someone tired and perhaps confused about how minute engineering decisions could upend their entire life. As lawmakers finally start to make moves on AI, we all have a choice about whom we listen to.”
I couldn’t agree more! In Primer theme number three, I wrote that “Who we select to help us understand the future is ultimately about power in the present — it’s about whose experience, expertise, and epistemology gets to decide what problems matter and how we understand them.” Well, I for one hope we listen to more people like Inioluwa. If you want to take a wonky dive into Inioluwa’s scholarship after reading her piece in The Atlantic, two of my favorites are “The Fallacy of AI Functionality” and “AI and the Everything in the Whole Wide World Benchmark.” Enjoy!
🔮 ‘Predictive’ algorithms chase the past under the guise of progress
Just last week I wrote an essay about building alternative futures. I drew on the example of predictive policing algorithms to make the broader point that predictive tools “represent a future that chases the past under the guise of ‘data-driven’ tools.” Well, this week, a new analysis from The Markup showed that these tools work (wait for it!) less than 1 percent of the time. Maybe now police departments will immediately stop using these tools?
Okay, probably not, but in the meantime we can remember Primer theme number four: data and technologies often say more about organizations and companies — their organizational structures, interests, and actions — than they do about us or the technology. In this context, it helps us understand that the data training so-called predictive policing algorithms aren’t predictive’ at all —they’re descriptive and diagnostic of the practices and behaviors of police departments.
🎙️SBF is on trial - but what about the dynamics that make crypto more susceptible to collapse?
One news story I couldn’t avoid (I promise, I tried) is the trial of Sam Bankman-Fried (SBF), which kicked off this week. Unsurprisingly, Michael Lewis also chose this week to launch his new book on SBF and the collapse of FTX, Going Infinite. The point is, SBF will be in the news a lot in the coming months. I don’t intend to write about the trial. (I highly recommend the coverage from
.) But I do want to make sure that as we hold individuals like SBF accountable for their crimes, we don’t lose sight of the broader system that treats the dynamic interaction of leverage and rigidity as features to be celebrated, rather than bugs to be avoided… or, ya know, regulated.That’s why I’m resharing a podcast interview I conducted with law professor Hilary Allen in July 2022:
In the conversation, Hilary and I discussed:
How decentralized finance or ‘DeFi’ — despite having ‘decentralization right there in the name — isn’t actually decentralized. (For early Untangled readers and listeners — and here, I’m mostly referring to my dad and my sister — this won’t be surprising.)
How transparent on-chain data isn’t terribly useful when systems are complex. Ah, yes, remember the limits of transparency?
How Hilary would regulate DeFi if given a blank sheet of paper.
Okay, that’s it for now.
Charley
It’s a great point, “Whose Ai?”. Whose AI is this? And AI has been around for over a century, employed by each generation, iteratively. Always broad industrial impacts to displace workers and social dissolution so we all reach for less human, more isolated and tech dependent nature to self-determine.