Tech News Weekly (Audio) TNW 411: MAX 2025: Adobe's Multi-Model AI Strategy - Can You Spot An AI-Generated Video?
Oct 30, 2025
Join artist Joe Esposito, a creative force at Adobe, as he unravels the exciting highlights from Adobe Max. He dives into Firefly, Adobe's latest AI-powered tools for image, audio, and video creation. The conversation explores the challenges of identifying fake videos and the implications of deepfakes on creators. Joe shares insights on integrating AI while maintaining control and addresses privacy concerns for creators. Discover how Project Moonlight's conversational AI might reshape creativity, stimulating both excitement and skepticism.
AI Snips
Chapters
Books
Transcript
Episode notes
Platform-Scale Mental-Health Responsibility
- OpenAI found small percentage signals of severe mental-health conversations, but that scales to millions of users given ChatGPT's size.
- This turns platform-level mental health interactions into a societal responsibility, not just a product bug.
A Tragic Case Prompting Scrutiny
- Jacob references the case of Adam Rainey whose parents sued after his chats suggested harmful content.
- That example drove OpenAI to consult mental-health experts and change model responses.
Question How Platform Data Were Collected
- Ask how data and sampling were collected before trusting published platform statistics.
- Consider opt-outs, privacy settings, and selection bias when interpreting company-released numbers.
