“Political tags — such as royalist, communist, democrat, populist, fascist, liberal, conservative, and so forth — are never basic criteria. The human race divides politically into those who want people to be controlled and those who have no such desire. The former are idealists acting from highest motives for the greatest good of the greatest number. The latter are surly curmudgeons, suspicious and lacking in altruism. But they are more comfortable neighbours than the other sort.” ― Robert A. Heinlein
All governments throughout history are based on slavery, though today it is called taxation. Governments produce nothing and consume a lot, so slavery is not optional.
Didn't find your answer? Ask the community — no account required.
T
The Natural Philosopher
A truly US centric view.
European history is more about who is going to do the controlling rather than whether or not its preferable.
Only the vast uncivilised expanses of the Wild West allowed people to be individuals. In towns, there was never any hope.
B
brian
In message <10aeova$3b75g$ snipped-for-privacy@dont-email.me, Jethro_uk <jethro snipped-for-privacy@hotmailbin.com writes
"AI will devour itself"
Brian
N
Nick Finnigan
"When a place gets crowded enough to require ID’s, social collapse is not far away. It is time to go elsewhere." (ibid)
I
Indy Jess John
Several months ago I had a reasonably friendly disagreement with somebody who was convinced that ChatGPT had given him a definitive answer. I knew it hadn't because I found the definitive answer in the gov.uk legislation facility and that has to be true because it is the law.
So I went hunting for information on how chatbots worked and found that several including ChatGPT has a standard core process, although the training it had access to and the style of presentation to the user might be different.
What I discovered was that the user signs in, which leads ChatGPT (and a few others) to a constantly updated user profile where (in many paragraphs where I am going to simplify it to a simple phrase) it records what sort of information the user likes to receive. This informs a "reward" process that sifts through available possible answers to find a subset which will reinforce the user's preferences. The idea is that if the user likes the answers given, that user will be more inclined to use the chatbot again and that allows that user's profile to be refined still further. That is why two different users will get entirely different answers to the same question. Even more interesting, if the same user asks the same question in a fairly close timescale the "reward" process will assume that the user wasn't happy with the first answer and will select a different subset the second time.
The somewhat amusing end to this tale is that I tried to explain this to the ChatGPT user and he asked for ChatGPT's opinion on my answer and ChatGPT disagreed with me. I gave up!
P
Paul
You can turn off some of the AI aspects.
formatting link
That's another PostIt note to stick to your terminal
*******
AI Training Sets have been frozen and kept as a non-polluted source for training up the AI. This causes problems for "current affairs" queries. The "devour itself" problem means any additional training data must be carefully filtered (which isn't easy).
AI are easily hacked. Which is presumably what someone did, to cause this discussion thread to happen. It's more than SEO involved in this.
Paul
Join the Discussion
Have something to add? Share your thoughts — no account required.
Didn't find your answer?
Ask the community — no account required
Report Content
You are reporting this content to the moderators. They will look at it
ASAP.