Post AYUZdahdodXipmcPuS by mykl@infosec.exchange
(DIR) More posts by mykl@infosec.exchange
(DIR) Post #AYUPEdTBOPOydNJkHo by clive@saturation.social
2023-08-07T13:57:10Z
0 likes, 0 repeats
A superb essay by Karawynn Long pointing out a central challenge in today’s language-focused AI — which is the assumption that fluency with language *is* intelligence: https://karawynn.substack.com/p/language-is-a-poor-heuristic-forAs Long notes, many folks with autism are exceptionally intelligent but don’t possess high fluency with spoken language — and they get regarded as unintelligent: https://karawynn.substack.com/p/language-is-a-poor-heuristic-for(via @emilymbender)1/x
(DIR) Post #AYUPEt61IINf6AbTP6 by clive@saturation.social
2023-08-07T13:57:11Z
0 likes, 0 repeats
The whole essay is crackling and worth reading in full; Long predicts that the penchant of large language model chatbots for screwing up basic facts will nudge users to stop associating language with intelligence …I hope she’s right, though the counterargument would be that humans have for aeons been perfectly happy to ascribe intelligence — and, indeed, absolute genius — to gladhanding bullshitters 2/x
(DIR) Post #AYUPF35g8Q3860WLGC by clive@saturation.social
2023-08-07T13:57:11Z
1 likes, 0 repeats
I really enjoyed this point Long made — which is that ChatGPT users seem particularly affected by the bot’s *apologies*, when it gets something wrong and is told “yeah no that’s wrong”They think the bot is learning But it’s not. That’s just a canned replyBut it has a powerful psychological effect on human users3/x
(DIR) Post #AYUPFAb8DsMtNaI3l2 by clive@saturation.social
2023-08-07T13:57:12Z
0 likes, 0 repeats
This is something I wrote about a few months back when that Google engineer was claiming that LaMBDA, Google’s large language model, was sentient When you looked at some of the records of his chats with the bot, what leapt out was that the chatbot’s replies evoked *vulnerability*Sherry Turkle has written a lot about how this is a classic trick of bot makers and robot creators — a bot that seems needy or fallible seems more human: https://clivethompson.medium.com/one-weird-trick-to-make-humans-think-an-ai-is-sentient-f77fb661e1274/4
(DIR) Post #AYUQ37sEcpETPT6wt6 by Ertain@mast.linuxgamecast.com
2023-08-07T14:06:28Z
0 likes, 0 repeats
@clive @emilymbender That's a nice observation.
(DIR) Post #AYUQIQMq9Gzz6uimCu by UP8@mastodon.social
2023-08-07T14:09:18Z
0 likes, 0 repeats
@clive a lot of people don’t get that the “Chat” in “ChatGPT” means that such a system is animated by a human user and takes advantage of that’s user psychology; chatbots have a power of seduction like the Emperor in that story of the Emperor’s new clothes — going out in public without any clothes isn’t something just anybody could pull off but requires a special competence (or at least social status) all of its own
(DIR) Post #AYURI1CFpgiL008kPw by clive@saturation.social
2023-08-07T14:20:22Z
0 likes, 0 repeats
@UP8 Yes!
(DIR) Post #AYURQ91ibXkpbH5Ry4 by clive@saturation.social
2023-08-07T14:20:58Z
0 likes, 0 repeats
@emilymbender @Ertain It’s good eh?
(DIR) Post #AYURdfKTocgdY8JGYS by parsley@sfba.social
2023-08-07T14:24:14Z
0 likes, 0 repeats
@cliveSo people repeat the question until they receive an answer they believe it's correct? Using this to argue for the intelligence of the answerer?That reminds me of Clever Hans... https://en.m.wikipedia.org/wiki/Clever_Hans
(DIR) Post #AYUSO7rM4aEHM5qBdo by StephanieMoore@mastodon.online
2023-08-07T14:28:39Z
0 likes, 0 repeats
@clive @emilymbender great article, thanks for sharing!
(DIR) Post #AYUSU3pzwHnqcS2IcK by EpiphanicSynchronicity@pkm.social
2023-08-07T14:26:40Z
0 likes, 0 repeats
@clive > humans have for aeons been perfectly happy to ascribe intelligence — and, indeed, absolute genius — to gladhanding bullshittersAnd even want them to be the leaders of their countries.
(DIR) Post #AYUSU4egtoLv9fkmdE by clive@saturation.social
2023-08-07T14:28:50Z
0 likes, 0 repeats
@EpiphanicSynchronicity Alas, yes
(DIR) Post #AYUSakrms6otoB45WS by clive@saturation.social
2023-08-07T14:29:32Z
0 likes, 0 repeats
@Retreival9096 I think I’m going to create a chatbot that does nothing *but* apologize
(DIR) Post #AYUTFzP9iwF4lqKpOa by edwiebe@mstdn.ca
2023-08-07T14:39:16Z
0 likes, 0 repeats
@clive That fellow has lots of degrees but none of them are in engineering.
(DIR) Post #AYUZOmdJB4cLUVWzho by RnDanger@infosec.exchange
2023-08-07T15:50:47Z
0 likes, 0 repeats
@clive That behavior drove me crazy for hours before I realized this is fancy autocomplete and nothing more. I feel much better after accepting that it won't learn anything from me and has no capacity to do better.
(DIR) Post #AYUZdahdodXipmcPuS by mykl@infosec.exchange
2023-08-07T15:49:51Z
0 likes, 0 repeats
@clive @briankrebs “... shaking the Magic 8 Ball over again…”
(DIR) Post #AYUZyLuE1fK11VP8hk by Ertain@mast.linuxgamecast.com
2023-08-07T15:57:42Z
0 likes, 0 repeats
@clive Yes, about the autism one.
(DIR) Post #AYUlvsNa25kWJqikFc by clive@saturation.social
2023-08-07T18:11:41Z
0 likes, 0 repeats
@mykl @briankrebs Heh yeah I loved that line
(DIR) Post #AYVKuXvG6NgYynZhYW by jobsboils@mstdn.social
2023-08-08T00:43:35Z
0 likes, 0 repeats
@clive I've also noticed this about many of the financial analysts I work with.Very smart people when it comes to mathematics and statistical theory, bur completely lost when it comes to expressing their thoughta in the English language,I often wondered if there's connection between that and the frequency of financial crises,
(DIR) Post #AYVuRkFqlkIXB9lwCu by miladiir@mastodon.social
2023-08-08T07:21:52Z
0 likes, 0 repeats
@clive I believe this is why they build it in. It‘s just too good to be a coincidence, in my opinion.
(DIR) Post #AYWZy7BJ9xjBfkCYBk by clive@saturation.social
2023-08-08T15:06:16Z
0 likes, 0 repeats
@miladiir wouldn't be surprising, yeah