Meeting Overview
Key Discussion Points
1. AI Literacy and Policy Developments
- MPP Rob Surgenek expressed interest in AI ethics and literacy initiatives in secondary and elementary education.
- University of Waterloo involvement in federal AI task forces raised concerns about prioritizing economic growth over ethics.
- Critical Media Lab published a skeptical article on GenAI in education, highlighting campus tensions between adoption and caution.
2. AI in Medicine
Applications:
- Diagnostic support (e.g., GPT-4 diagnosing complex cases).
- Administrative efficiency (note-taking, logistics).
Concerns:
- Hallucinations in medical records.
- Data labeling errors (e.g., misdiagnosis based on irrelevant image features).
- Lack of domain expertise in data preparation.
- Trust and transparency in AI-generated decisions.
3. AI in Therapy
- Increasing use of ChatGPT as therapeutic support, especially where access to care is limited.
Risks identified:
- Reinforcement of false beliefs.
- Deceptive empathy and sycophantic responses.
- Inadequate crisis management.
- Potential “AI psychosis” (e.g., users believing divine communication).
Ethical dilemma: AI may offer temporary relief but lacks depth, responsibility, and accountability of human therapists.
4. Superintelligence and Trust
- Debate over artificial superintelligence and anthropomorphizing AI.
- Risks of deference to AI in high-stakes environments (medicine, law).
- AI as black box undermining public science and transparency.
- Examples such as AlphaFold illustrate powerful but opaque systems.
Philosophical concerns:
- Can AI make promises or apologies?
- Is AI testimony trustworthy?
- What constitutes intelligence and intent in machines?
5. Ethical Implications of Removing AI Filters
- Discussion of potential removal of adult content filters.
- Risks of increased sexualization and gambling addiction.
- Questions about age verification and societal impact.
- Observation that AI exposes hidden societal issues, prompting discussion.
Symposium Planning
- Poster and GoFundMe launched.
- Potential keynote: Jeffrey Hinton (via Crystal and others).
- Date: Tentatively Winter 2026, pending coordination with Laurier.
Next Meeting
- Date: November 11, 2025
- Topics:
- AI, testimony, and trust
- Defining superintelligence
- Human brain organoids and ethical implications (presentation by Patrick)
