The Artificial Intelligence Show #164: New MIT Study Says 95% of AI Pilots Fail, AI and Consciousness, Another Meta AI Reorg, Otter.ai Lawsuit & Sam Altman Talks Up GPT-6
233 snips
Aug 26, 2025 A new MIT study reveals a staggering 95% failure rate for generative AI pilots in businesses. The ethical implications of seemingly conscious AI are discussed, highlighting potential societal risks. There's a major reorg at Meta, aiming to enhance AI innovation. Otter.ai faces a lawsuit over privacy violations concerning meeting recordings. Meanwhile, Sam Altman teases GPT-6’s capabilities, focusing on personalization and privacy concerns, alongside insights into next-gen tech from Google and Apple. The environmental impact of AI is also a pressing issue.
AI Snips
Chapters
Transcript
Episode notes
Scrutinize Viral Research Claims
- Bold headlines (like "95% of pilots fail") require scrutiny of methods before sharing.
- Read methodology and limits; small interview samples can’t justify sweeping claims.
Run Pilots As Change Management
- Do treat AI pilots as change management projects, not just tech rollouts.
- Provide use-case training, measured success metrics, and short guided tasks for day-one value.
Company Goes All-In And Rebuilds Staff
- Ignite Tech enforced an all-in AI shift and replaced 80% of staff within a year.
- The CEO later admitted the approach was extreme but financially successful.