BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Palaner//palaner Single Event//EN
METHOD:PUBLISH
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:meetup-856aa77e4b@palaner.app
DTSTAMP:20260802T155659Z
DTSTART:20260817T080000Z
DTEND:20260817T093000Z
SUMMARY:AI and the Limits of Human Judgement
LOCATION:The Stolberg Hotel
DESCRIPTION:Saved in Palaner as: saved\n\nPalaner match: 29% · why: lots of
  people going\n\nSkeptics Café will be held in the Function Room at The St
 olberg Hotel\, 197 Plenty Road Preston. The 86 tram route is close\, and B
 ell train station – situated on the Epping line – is a short walk away. It
  will also be a virtual event accessible via Zoom\, as per the details her
 e.\nStart Time: 6:00pm for dinner and a chat. Talk and Zoom session starts
  at 7.30pm.\nWe aim to wrap up after 9:00pm.\n\nMore Moral Than Us?\nWhy t
 he question matters for alignment\, moral progress\, and long-term flouris
 hing\n\nTL\;DR: On some measurable moral tasks\, today's AI already looks 
 competitive with humans\, or better\, and the alignment community is oddly
  quiet about it. But moral capability is not the same as moral motivation:
  a system can compute the right answer without being moved by it. Pulling 
 apart knowledge\, reasoning\, judgement and motivation is what makes the q
 uestion tractable\, and what makes the honest answer uncomfortable. Caveat
 s attached\, in quantity.\n\nAlmost nobody working on AI alignment wants t
 o say this out loud\, at least not about current systems. So let's say it:
  AI might already be more moral than us\, in some measurable respects.\nTw
 o questions fall straight out of that:\n\nWhy does it feel like a dangerou
 s thing to claim?\nAnd is human moral reasoning really a standard worth br
 agging about as an alignment target?\n\nThe provocation isn't idle. In a m
 odified Moral Turing Test\, people rated a large language model's moral re
 asoning as better than other humans' on almost all measured dimensions - s
 ee paper Attributions toward artificial agents in a modified Moral Turing 
 Test + associated interview and talk with lead author Eyal Aharoni.[1] Fin
 dings like that are easy to over-read and easy to wave away\, which is mos
 t of the problem.\n\nProvocative questions? Yes\, and they are backed by e
 mpirical research. In a modified Moral Turing Test\, people rated GPT-4's 
 moral evaluations as better than other humans' on almost every dimension\,
  though they could still tell which was the machine (Aharoni et al. 2024\;
  I interviewed and hosted a talk by lead author Eyal Aharoni).[1] The resu
 lt has since been contested and defended\, which is exactly the kind of th
 ing this evening is for. Findings like that are easy to over-read and easy
  to wave away\, which is a problem that may obscure appropriate interpreta
 tions.\n\nThe claim gets tabooed for reasons that aren't stupid. Premises 
 do get smuggled in unnoticed. Morality is bound up with identity and tribe
 \, so "AI is more moral" can land as an attack rather than a hypothesis. A
 nd it sits close to AI-worship\, or to motivated reasoning looking for an 
 excuse to defer to the machine\, and deferring to AI on moral questions is
  genuinely epistemically risky. So the caution is understandable. The trou
 ble is that it has hardened into a taboo\, and the taboo now costs more th
 an it protects.\n\nMost of the bad arguments\, on both sides\, come from r
 unning four different things together:\n\nMoral knowledge: knowing the fac
 ts that bear on a decision.\nMoral reasoning: drawing sound inferences fro
 m values to conclusions.\nMoral judgement: applying principles well in a m
 essy particular case.\nMoral motivation: actually being moved by moral con
 siderations\, not merely computing them.\n\nGroup the first three as epist
 emic competence: marshalling the relevant knowledge\, drawing sound infere
 nces\, judging the particular case. Now grant\, purely for argument\, that
  AI already does all three better than we do. Notice what happens to the a
 lignment problem. It doesn't dissolve. It relocates\, wholesale\, onto the
  fourth thing: whether the system is actually moved by any of it. Concede 
 every epistemic point and the hard question is untouched. That gap\, betwe
 en knowing and caring\, is where alignment actually lives.\n\nA system cou
 ld plausibly beat humans on the first three while having nothing resemblin
 g the fourth. Keep them apart and the conversation becomes tractable. Coll
 apse them and you get overclaiming and underclaiming in the same breath\, 
 which is roughly what the current debate produces.\n\nThere's also a real 
 risk in how the framing gets used. Point it one way and it justifies handi
 ng AI authority over human decisions. Point it the other and it's ammuniti
 on for painting alignment researchers as unhinged techno-utopians. Both fe
 ed motivated reasoning. Worth naming out loud\, not a reason to stop.\n\nN
 one of this settles the answer\, and the evening won't pretend to. The nar
 rower claim is enough: once a serious community stops itself from asking w
 hether AI could have better-grounded moral reasoning than humans\, the tab
 oo has become the expensive thing. It's a question worth putting on the ta
 ble\, with every caveat attached. Come and help pull it apart.\n\nAlso see
 :\n- More Moral than Us\n- Why Are We Afraid to Ask Whether AI Could Be Mo
 re Moral Than Humans?\n\n[1] AI didn't pass as human\, it was rated as bet
 ter than human. The same study found that when participants were then aske
 d to identify which evaluation came from a machine\, people performed sign
 ificantly above chance levels. So GPT-4 didn't fool anyone. The authors' r
 eading is that people could pick the AI out not because it reasoned worse 
 but because it seemed to reason better\, i.e. its perceived superiority wa
 s the tell.
URL:https://www.meetup.com/science-technology-and-the-future/events/3153822
 61/
CATEGORIES:palaner,saved
STATUS:TENTATIVE
END:VEVENT
END:VCALENDAR