Search for a command to run...

Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
In this episode of Hard Fork, Kevin and Casey tackle the intersection of technology and political violence in Minneapolis, where ICE operations and protester confrontations are being shaped by tech platforms, AI-generated content, and federal media strategies. (03:03) The hosts examine how tech CEOs like Sam Altman, Dario Amodei, and Tim Cook are carefully navigating political responses while facing pressure from employees to speak out against federal overreach.
The discussion reveals a troubling new dynamic where smartphones have become weapons in a "phone versus phone" battle between protesters documenting violence and federal agents using influencers and AI-manipulated images to control the narrative. (17:56) Meanwhile, Casey experiments with MoltBot (formerly ClaudeBot), an open-source AI assistant that can integrate with email, calendar, and messaging apps - but comes with significant security risks. (25:49)
Kevin Roose is a tech columnist for The New York Times and co-host of Hard Fork. He covers artificial intelligence, social media, and technology's impact on society, bringing deep expertise in how emerging technologies intersect with politics and culture.
Casey Newton is the founder of Platformer, a newsletter covering the intersection of technology and democracy. He previously worked as a senior editor at The Verge and is known for his deep reporting on social media platforms, content moderation, and tech policy.
The Minneapolis events demonstrate that smartphones have become essential tools for accountability in confrontations with law enforcement. (18:22) Governor Tim Walz explicitly urged Minnesotans to "carry your phone with you at all times in order to help us create a database of the atrocities against Minnesotans." Videos of the Alex Preti shooting, captured from multiple angles by different witnesses, became undeniable evidence that shaped public opinion and even prompted criticism from the NRA. This represents a fundamental shift where citizen journalism through smartphones can counteract official narratives, even when federal agents are actively trying to discourage filming by claiming it constitutes "doxing."
AI-manipulated images are now being weaponized by government officials at the highest levels, creating what Casey calls the "liar's dividend" - where people question all evidence because they know it can be fabricated. (13:45) The White House released an AI-altered image of civil rights attorney Nikima Levy Armstrong to make her appear to be crying when she wasn't, which was then retweeted by Vice President JD Vance. When questioned about sharing doctored content, a White House spokesman simply said "the memes will continue." This represents a dangerous evolution where state power combines with AI tools to systematically blur the line between truth and fiction.
Personal AI agents like MoltBot offer compelling glimpses of the future but come with severe security risks that most users shouldn't accept. (28:01) Casey's experiment with MoltBot, which he connected to his email and calendar to create personalized daily briefings, demonstrates both the potential and the pitfalls. While the tool can theoretically handle complex tasks like making restaurant reservations via phone calls, it also creates multiple attack surfaces - from prompt injection attacks to potential remote access through messaging apps. The key insight is that these tools work best in sandboxed environments with limited access to sensitive data, not as full-access personal assistants.
A significant divide is emerging between early AI adopters who are experimenting with cutting-edge tools and mainstream users who are still navigating basic enterprise AI policies. (48:10) As Kevin observed, "People in San Francisco are putting multi agent clod swarms in charge of their lives, consulting chatbots before every decision" while others "are still trying to get approval to use Copilot in Teams." This gap matters because legendary programmer Andre Karpathy noted that AI coding tools represent "easily the biggest change to my basic work coding workflow in two decades of programming." If these tools continue to improve, the productivity differential could become a significant competitive advantage.
Government agencies like ICE have developed sophisticated content creation teams that use the same techniques as brands and celebrities to control narratives on social media. (11:11) According to Washington Post reporting based on internal documents, ICE now employs teams of producers, editors, and video makers who use paid social media tools to steer public perception of their operations. This represents a fundamental shift where "winning on social media has become almost the entire point" rather than just a byproduct of policy enforcement. The Trump administration has taken this further by bringing conservative influencers along on operations to create favorable content.