Big Ideas The AI Con — unpacking the artificial intelligence hype machine
39 snips
Sep 2, 2025 
In this engaging discussion, linguistics professor Emily M Bender, digital ethics lawyer Kobi Leins, and journalist Tracey Spicer dive into the AI hype machine and its societal implications. They dissect the exaggerated narratives around AI, challenging misconceptions while emphasizing the environmental costs of automation. With humor and critical insights, the panel explores the ethical dilemmas of AI, the need for education in technology, and the importance of passion in studying these fields. They advocate for a future where responsible AI integration is prioritized.
AI Snips
Chapters
Books
Transcript
Episode notes
Ask Hard Questions Before Adoption
- If you can't refuse an AI rollout, ask probing questions about evaluation, metrics, and context-specific testing.
- Demand evidence that systems were tested with the actual user groups and conditions where they'll be used.
Harms Start In Model Production
- Harm arises both from production (data theft, exploitative labor) and from use (misapplications, bias).
- Massive-scale scraping treats creative work and personal posts as a dehumanized "sludge" of training data.
Hidden, Traumatic Data Labor
- Data-labeling work for AI is outsourced to precarious gig workers who face psychological harms and low pay.
- Reporters found workers in Kenya paid about $2 an hour with no mental-health support doing traumatic content moderation.
