The Vermont Senator Keeps Platforming Boosterism

By: Decca Muldowney

Bernie Sanders published an op-ed in The Guardian this week, warning of the existential dangers posed to civilization by AI. “It will bring unimaginable changes to our economy, our politics, warfare, our emotional wellbeing, our environment, and how we educate and raise our children,” he writes. Equally, Bernie seems to be deep in the AGI sauce, writing that “there is a very real fear that, in the not-so-distant future, a super-intelligent AI could replace humans in controlling the planet,” and citing his recent discussions with Geoffrey Hinton.

The piece comes on the heels of a report published in October from Bernie’s Health, Education, Labor and Pensions (HELP) Senate committee that predicted “artificial intelligence and automation could replace nearly 100 million jobs over the next ten years.” The problem with this report? It used ChatGPT for its analysis instead of using any real data. Not to mention, they didn’t speak instead of speaking to any workers facing displacement threats directly. As another slap in the face, Not to mention the report uncritically cites job loss estimates from Anthropic CEO Dario Amodei, Elon Musk, and McKinsey.

While we applaud Bernie’s concern about the impacts of “AI” on working people, we wish he would focus on the actual effects of automation on workers, didn’t use chatbots to make predictions, and stopped taking cues from AI Doomers like Hinton and others with vested industry interests. 

As we discussed in detail in the latest episode of Mystery AI Hype Theater 3000 with researcher and organizer Sophie Song, the HELP committee report relied on ChatGPT’s “analysis” of job automation, and didn’t talk to any actual workers about their experience, instead quoting Dario Amodei, Elon Musk and McKinsey to make its argument. As Alex put it: “They do this bullshit analysis, but then the people that they quote approvingly are certified enemies of the working class. Why are you taking these people at their word, when you shouldn't?”

While the report makes some good recommendations that we support, such as the introduction of a 32-hour work week, profit-sharing with workers, a “robot tax”, doubling union membership and bringing back pensions, we wish this intervention had been based on a proper, robust analysis of the available data on automation, as well as conversations with workers and trade unions about the impacts they are already seeing. 

“One of the core demands that Labor is making around technology is that they want workers' voices, workers’ say, worker's agency and decision making power, to be a core part of how technology shapes their work,” Sophie said on the podcast. “And this report, that is essentially about technology and workers, refuses to include that worker’s voice."

Interested in learning more about our thoughts on labor and “AI”? Here are some MAIHT3K episodes to get you started:

• Episode 67: You Talked to Workers for This Labor Research... Right? Our latest episode features guest Sophie Song, a researcher, organizer, and advocate working at the intersection of tech and social justice. We unpack the issues when self-professed worker advocates use chatbots for "research."  [Podcast, Transcript, Livestream]

• Episode 25: An LLM Says LLMs Can Do Your Job. Alex and Emily unpack two hype-tastic papers that make implausible claims about the number of workforce tasks LLMs might make cheaper, faster or easier. Alas, this is the same methodology in the HELP report. [Podcast, Transcript, Livestream]

• ‘Don't Be Fooled By Trump's Plan To 'Upskill' Workers To Prepare For AI,’ Tech Policy Press. Alex, Sophie, and DAIR Research Associate Tina Park take on the claims about “AI” and productivity in response to Donald Trump’s AI Action Plan.