Path: csiph.com!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail
From: rbowman <bowman@montana.com>
Newsgroups: comp.os.linux.misc
Subject: Re: Warning - Serious 'sudo' Flaw Compromises Security
Date: 21 Oct 2025 19:13:45 GMT
Lines: 12
Message-ID: <mlq4b9FtkjiU7@mid.individual.net>
References: <t5qcnTPojIJuBHX1nZ2dnZfqn_udnZ2d@giganews.com> <1NucnW53DZrMOHH1nZ2dnZfqnPGdnZ2d@giganews.com> <10cihiq$23kb8$6@dont-email.me> <10cios3$25sim$2@dont-email.me> <10cj667$29n96$5@dont-email.me> <gRaHQ.184188$6_hb.85411@fx46.iad> <10cjf5m$2cnh5$2@dont-email.me> <SrbHQ.68586$RrE7.59909@fx45.iad> <RI2cnfyQDJK4fnD1nZ2dnZfqnPednZ2d@giganews.com> <ml69afFhs47U1@mid.individual.net> <wpKcnVN4_MydQ3P1nZ2dnZfqnPGdnZ2d@giganews.com> <10co09b$3jj2s$11@dont-email.me> <sE2dnQnUGeD6HW31nZ2dnZfqnPqdnZ2d@giganews.com> <10cqhde$a1mj$4@dont-email.me> <2smdnYMrA9D3Y2_1nZ2dnZfqnPoAAAAA@giganews.com> <mlgmtsFc4rqU2@mid.individual.net> <10cvnsr$1nij3$4@dont-email.me> <mli58vFjgrdU3@mid.individual.net> <10d2eq5$2dnk0$9@dont-email.me> <mlkrdrF2rt4U1@mid.individual.net> <unGdnYj3Caf-WWj1nZ2dnZfqn_idnZ2d@giganews.com> <10d4tqa$337uc$1@dont-email.me> <mlnetfFgob9U3@mid.individual.net> <10d644u$3evgm$1@dont-email.me> <mloaerFku2iU3@mid.individual.net> <IXqdnQIXcbkziGr1nZ2dnZfqn_WdnZ2d@giganews.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
X-Trace: individual.net R0RKedsFgEodmKvYRKBiWA84ltlLN0+fKzIdTKQqC7zi6llXnG
Cancel-Lock: sha1:+eAat8JNHCcLjFrxse1JmUlEKB8= sha256:FAkBhrD215lRWWbxkyYFfiIqVqyoaonEeyvB7pG8jkQ=
User-Agent: Pan/0.162 (Pokrosvk)
Xref: csiph.com comp.os.linux.misc:76473

On Tue, 21 Oct 2025 01:09:47 -0400, c186282 wrote:

>    LLMs and related are a sort of different way to emulate "IQ". They
>    are a sort of "fake", built- up emulations/rules based on seeing how
>    millions of humans think about such things.
> 
>    Note that if you fake something WELL ENOUGH it's not really "fake"
>    anymore ... we're almost approaching that point with the LLMs now.


As a thought experiment, how much human discourse is really group think 
constructed from social clues?