Philosopher's Walk - Royal Botanic Gardens

Weather looks reasonable on Sunday. Participants are invited to join us for a walk and talk about science, philosophy, AI and emerging technology.

Schedule

• 10.30 am – Meet outside the entrance to the Royal Botanic Gardens Visitor Centre (near the Observatory House, and a walk away from the Shrine of Remembrance)
• 11.00 am – Venture into the gardens and walk the outer path anticlockwise, spiralling inwards towards the lake
• ~1.00 pm – (optional) Lunch by the lake, (BYO or there are Terrace Café options)

Getting there: If coming by tram, get off at Stop 19 - Shrine of Remembrance/St Kilda Rd
Parking is free on Anderson St far side of the Botanic Gardens (near eastern gate B and Tecoma gate C) subject to time limits in some stretches (mostly 3 hrs) - which is the only free parking area near the gardens that I can find. Check signage on the day. See this PDF for more detail.

Possible discussion topics

• Latest AI models - Anthropic's Mythos and Fable blocked after US bans foreign use
• AI Pause / Stop - will it work? (Regulatory catch-up, xrisk, global-coordination & geopolitical arms races, enforcement & logistics, existing capability overhang)
• Technical AI safety - Mechanistic interpretability, linear probes, RLHF, behavioural evals
• General AI alignment approaches: control vs motivation:
  - control (External Constraints) focuses on building walls around the AI to restrict what it can do, trip wires etc.
  - motivation (Internal Alignment) focuses on changing the AI’s inner desires so it wants to be ethical/help humans/act in accordance with the well-being of all sentience
• AI ethics: how is it related to AI safety? What should be the target for AI alignment? Should the target be aggregate human preferences, religious doctrine, moral naturalism or something else?
• Metaethics and ethics
• Moral realism: a stance-independent moral ontology, useful for capturing non-arbitrary moral choice
• Moral anti-realism: error theory, non-cognitivism, moral relativism etc
• Can AI be fault-tolerantly aligned to moral realism or anti-realism?
• Is sentience required for moral reasoning or moral deliberation? Can it be a hindrance?
• Moral progress and indirect normativity
• AI safety, automated persuasion, the orthogonality thesis
• How epistemics (including science) makes progress
• Health, wellness, and how to live a good life
• Latest trends in AI: ChatGPT, Gemini, Claude
• Economic impacts of AGI, automation, and unbounded inequality
• UBI and universal wealth redistribution
• The likelihood of future doom or utopia
• Human augmentation and hacking human potential
• Plato's cave
• Fun theory and value theory

"All truly great thoughts are conceived while walking." - Nietzsche, Twilight of the Idols
Philosophy discussions can feel quite sedentary - hence our experiment in cross-pollinating exercise with engaging discussion. Exerting mind and body together may produce a fascinating dynamic. A sedentary lifestyle (and, by association, philosophy) has been linked to depression, anxiety and chronic stress.
Ease the anxieties of the armchair philosopher engaging in mental gymnastics with minimal muscle movement - join our raggle-taggle crew of aspiring philosophers walking the talk. Healthy body, healthy mind.
Bring elemental protection (umbrellas, sunscreen, etc).