Episode Summary
Executive Summary: This episode covers three main stories: Sam Bankman-Fried’s bail revocation and looming FTX trial, where prosecutors allege he violated restrictions by contacting witnesses, using a VPN, and leaking Caroline Ellison documents; a conversation with Wharton professor Ethan Mollick on how generative AI is upending education and how schools should adapt; and listener reaction plus fresh news in autonomous vehicles, including new regulatory approvals, a Cruise traffic pileup, and debate over whether AVs help or hurt car-free urban futures.
Main Topics: SBF bail revoked ahead of FTX trial: David Yaffe-Bellany explains why Sam Bankman-Fried was jailed after allegedly pushing bail limits through witness contact, VPN use, and media leaks, and how the judge viewed the conduct as the final straw. Caroline Ellison documents and witness intimidation concerns: The hosts and David discuss reports about Ellison’s writings, how prosecutors used them to argue SBF was influencing a key witness, and why the material could matter at trial even if not directly exculpatory or inculpatory. The evidence avalanche in the FTX case: Prosecutors filed a detailed pretrial evidentiary roadmap citing recordings, text messages, notes, Google Docs, and company records from multiple insiders, underscoring how much documentary evidence will shape the October trial. How generative AI is changing education: Ethan Mollick argues that AI has broken traditional homework and that schools need to move toward flipped classrooms, active learning, and clearer rules about what AI use is allowed. AI as a teaching tool, not just a cheating tool: Mollick describes mandatory AI use in his classes, including impossible assignments, prompt-based critique, and AI-generated tutoring/mentoring supports that push students to do more advanced work. Autonomous vehicles, safety, and city design: The hosts revisit listener criticism of last week’s AV segment, discuss new AV news in San Francisco and beyond, and wrestle with whether self-driving cars are a transitional good or a distraction from car-free urbanism. Listener backlash and nuance on AVs: A large volume of feedback challenged the hosts on transit politics, climate framing, safety evidence, and whether the real goal should be fewer cars rather than better cars.
Key Arguments: Prosecutors argue SBF violated bail by contacting a former FTX employee, using VPNs to evade monitoring, and leaking Caroline Ellison’s documents to the press. The judge revoked bail because the pattern of behavior suggested SBF was trying to evade restrictions and potentially interfere with witnesses. The upcoming FTX trial will hinge on extensive documentary and testimonial evidence from close insiders, including recordings and contemporaneous notes. Generative AI has effectively broken standard take-home homework because detection is unreliable and often biased. Schools should adapt by emphasizing flipped classrooms, active learning, in-class writing, oral work, and AI-assisted practice rather than banning AI outright. AI can help teachers create richer exercises, feedback loops, and personalized prompts, but it also risks flattening students’ original thinking if overused. Car-free cities are a compelling ideal, but many people still need cars for disability access, childcare, or work, so incremental safety improvements may be more realistic in the near term. Autonomous vehicles may reduce congestion if they improve utilization, but evidence is still limited and self-reported safety data should be treated cautiously.
Data Points: Trial start date: October 2 - Prosecution said the FTX criminal trial begins October 2 after pretrial wrangling. Bail website access limit: About 25 websites - SBF’s revised bail terms allowed access to only a limited list of websites. Time since initial collapse: Late last year / November 2022 - FTX collapsed in late 2022 and the episode frames the trial as nearing one year later. Close associates expected to testify: 3 - Prosecutors said three close associates of SBF, including Caroline Ellison, are expected to testify. Evidence volume: Millions of pages - The hosts note the case involves millions of pages of documents, including Google Drive contents, texts, Slack, and Signal chats. Self-driving vehicle safety claim: 54% fewer collisions - Cruise’s report claimed its AVs had fewer collisions than human rideshare drivers in the first one million miles. Primary-cause collision reduction: 92% fewer collisions - Cruise reported 92% fewer collisions where the AV was the primary contributor. Meaningful injury-risk collision reduction: 73% fewer collisions - Cruise said AVs had 73% fewer collisions with a meaningful risk of injury. Other-party fault share: 94% - Cruise reported that 94% of collisions it encountered were caused by the other party. AV pileup duration: About 15 minutes - The Cruise traffic blockage in North Beach lasted roughly 15 minutes before clearing. AV network expansion: 24/7 paid rides approved - CPUC voted to allow Cruise and Waymo to expand San Francisco service and charge fares at all hours. Cheating workforce estimate: 20,000 people - Mollick cited a study estimating roughly 20,000 people in Kenya write essays for others.
Pivotal Quotes: "The word I'd probably use would be chaos or apocalypse." — Ethan Mollick: Mollick describing the state of higher education as schools confront generative AI. "The old homework assignment is basically cracked by AI." — Ethan Mollick: He explains why traditional take-home assignments no longer reliably measure student work. "Nice diary you got there. Be a shame if something happened to it." — Casey Newton: A joking but pointed remark about prosecutors’ theory that SBF leaked Caroline Ellison’s writings to intimidate her.
Implications: The episode suggests courts, schools, and cities are all being forced to adapt to fast-moving technology. FTX shows how digital traces can dominate a case; education may need to redesign assessment; AVs may advance, but only amid major debate over safety and urban priorities.
From the Episode
Inside the world of higher education, I know you've been talking with a lot of faculty members and administrators at schools across the country who it's fair to say, I think are confused and disoriented about what to do about all of this new technology. So, I want to just start with kind of a vibe check. Like, can you just paint a picture for us of what is going on with generative AI at schools right now as we head into the fall semester? I think the word I'd probably use would be chaos or. Apocalypse. I think that people are just starting to dawn on them what this means. And I think when we talk about what this means, I think level one of what this means is what's dawning on them right now, which is, oh God, all my homework assignments don't work anymore. And people haven't started to think about the other implications fully yet. There's a lot of exceptions out there, but generally, that's the vibe I'm getting. Now, this technology has been available since last November. So, what has happened between November and now, where some of you know, some other folks in higher ed or maybe secondary still are.
Exams, there are other ways of checking, but the old homework assignment is basically cracked by AI. I wonder if you could just take us inside one of these kind of faculty meetings where professors and school administrators are trying to figure out how to adapt to the post-Chat GPT world. What are some of the common sort of things that you hear brought up in this world? And what are some of the sort of common objections to like, why shouldn't we be changing our policies and procedures? So, the first thing is the same problem everybody is suffering from. Generative AI, which is a no instruction manual. We're all figuring out as we go. There's literally papers coming out regularly about what kind of questions should you ask AI to get the best answers. We don't know whether it's really good at these tests or whether it's faking being good at these tests. So you come to a faculty meeting, and the first 20 minutes are debunking rumors and then supporting others about how it learns and what it knows, and is it stealing information and what's ethical? So there's a lot of that kind of discussion you'd expect among academics. Then there is sort of a discussion about how do you.
About Hard Fork
“Hard Fork” is a show about the future that’s already here. Each week, journalists Kevin Roose and Casey Newton explore and make sense of the latest in the rapidly changing world of tech. Unlock full access to New York Times podcasts and explore everything from politics to pop culture. Subscribe today at nytimes.com/podcasts or on Apple Podcasts and Spotify. Also, for more podcasts and narrated articles, download The New York Times app at nytimes.com/app.