Coincidentally saw this post from Melissa Chen on X the other day talking about racism.
https://x.com/MsMelChen/status/2113491599623627086?s=20 Chen says: "As a connoisseur of racism, I have found no greater joy than discovering other language realms on X where they are truly light years ahead. Behold the great meme war between the Turks 🇹 🇷 and Chinese 🇨 🇳 which has been brewing now for 3 days!" Kind of mildly amusing. But really deeply sad. I'm sure Chen's appreciation is ironic. You don't see much of this in Western media anymore. X being an exception for testing those prohibitions? But I suspect outside of Western media, it may be the norm. I don't recall what forum it was that I dropped in on a hate fest a few years back between... I think it was Macedonians and... Greeks? Or Serbs? If you look for them, I expect these are always going on. Basically in every community beyond the Anglosphere (where it exists, but in most of the media it's become an odd reverse taboo), it may be the norm. Probably it's central to human identity (Out group hostility, I always think of Patricia Churchland's flip side of morality, where she defines morality to be an extension of self to others https://youtu.be/wjQQ8l1vCOE?t=1966.) What to do about that is an open question, which is far from being solved in the Western media too. Instead, in Western media outgroup hostility has morphed to different out groups, and if anything become more extreme (perhaps because it relies on a denial of biology, and so becomes fundamentally untruthful?) The solution sought in the West has been a kind of balancing act, based on identity. Which in some ways seems to be more destructive (because it makes identity too important, and attacks truth.) So a better solution awaits us. Personally I think AI will offer the answers. But only when we get to the bottom of... not a definition, but a mechanism, for "truth". The question of "identity" I do think is central for AI. Not necessarily in a gene competitive sense. I've said elsewhere I think it's the job of AI research to give us a better definition of truth than identity. But in AI too both truth and identity remain, I would say, the centrally unsolved problems. Not just a training issue as it's usually characterized. An issue of what we define them to be. And fitting with "alignment" issues, yes. Alignment usually meaning ethics. Which is even less well defined than truth. -Rob On Sat, Oct 24, 2026 at 11:25 PM Ben Goertzel <[email protected]> wrote: > > Racism is certainly real and still prevalent yeah > > It is not as bad as when I was in high school in New Jersey in the 70s > and interracial dating (black/white) caused serious violence (smashing > windows , overturning cars, beatings etc.). I experienced assaults > myself due to having a black gf in the early 80s... and being in the > wrong neighborhoods of Philly and not hiding it... > > But still... Half the reason I left HK after 9 years and becoming a > permanent resident was the Chinese gov't National Security law and > treatment of the anti-govt protestors... and half was the way the > local HK-ese treated individual mainland Chinese (who had nothing to > do with Chinese gov't) -- including my wife -- calling them > cockroaches constantly and harassing them publicly... > > A close friend is one of the best hackers in the works and also > Ethiopian ... the BS he has gone thru in the Western tech industry as > a result of being black would surprise zero black people ... like > passing every technical job interview at a big tech with maximum > success then suddenly being rejected as "not a cultural fit" etc... > > DEI rules have often been implemented awkwardly and stupidly, but > also, they often haven't even worked to solve the problem of > discriminatory hiring > > One of my adult sons is Muslim, he was detained a week once in an > airport for no crime other than being young, white and traveling with > a Quran and funny hat... > > Age discrimination is real also. My company's PR agency will always > describe me as having "20 years of experience in the AI field" --- > because saying "30 years" (egads, it's 37 yrs since I got my PhD !!) > would make me seem too old to pay attention to.... > > All this -- along w many other things -- makes the talk of aligning > AGI ethics with human ethics a bit too simplistic.... somehow we need > to infuse the AI with the best of human compassion and ethics and > understanding, and avoid it absorbing the worst.... for an ASI to > treat us the way we treat each other would be complexly problematic... > > Which actually is one of the reasons why we need other AI approaches > than "train systems on the internet and then try to RL them to ignore > the bad parts and slap guardrails on them to try to prevent the really > bad stuff." I think your logic transformer, YKY, is a quite > interesting direction in this regard... and I don't need to see > competitive benchmark results to take it seriously... but it's also > true that with things moving so fast in the AI field and so many > interesting things happening, it is hard to find time to study all the > new things emerging. I have not read your paper on logic transformers > yet... can you re-post a link to it? I will try to find time to take > a look... > > -- Ben ------------------------------------------ Artificial General Intelligence List: AGI Permalink: https://agi.topicbox.com/groups/agi/T04fde0bcaa976a4b-Md100830cd442e6f4bca9cf55 Delivery options: https://agi.topicbox.com/groups/agi/subscription
