Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > alt.comp.os.windows-10 > #183384

Re: Scary AI

Path csiph.com!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!eternal-september.org!.POSTED!not-for-mail
From Chris <ithinkiam@gmail.com>
Newsgroups alt.comp.os.windows-10
Subject Re: Scary AI
Date Tue, 8 Apr 2025 20:57:05 -0000 (UTC)
Organization A noiseless patient Spider
Lines 78
Message-ID <vt42j1$326a7$1@dont-email.me> (permalink)
References <vt3f7u$2eh7c$3@dont-email.me> <vt3jee$2krgp$1@dont-email.me> <vt3oli$2omc2$1@dont-email.me> <vt3qqt$2rfml$1@dont-email.me>
MIME-Version 1.0
Content-Type text/plain; charset=UTF-8
Content-Transfer-Encoding 8bit
Injection-Date Tue, 08 Apr 2025 22:57:06 +0200 (CEST)
Injection-Info dont-email.me; posting-host="6435cac40f1cb6a7941eae4709d3d704"; logging-data="3217735"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX1+UKw4v6I40GygLtYw3NiQKZvCPHLPJ7ds="
User-Agent NewsTap/5.5 (iPhone/iPod Touch)
Cancel-Lock sha1:Kc2SjSJRRKis0+ncZPLGZVj0wuE= sha1:W/0w+z+ZNqCAD6rcR1m9IqWYpwU=
Xref csiph.com alt.comp.os.windows-10:183384

Show key headers only | View raw


Newyana2 <newyana@invalid.nospam> wrote:
> On 4/8/2025 2:07 PM, Ed Cryer wrote:
> 
>> How does the program do that?
>> I used to earn my daily bread programming commercial mainframes. But 
>> this AI "in the style of" beats me. I've done Walt Whitman, Louis 
>> Carroll, William Wordsworth and others; and all win my praise.
>> And that's only one facet of the AI phenomenon.
>> 
>> I'd have needed a vast database of examples set up just for Shakespeare 
>> alone.
>> 
>> I'll not rest until I grab the working essentials of how this thing 
>> works; call it a program, call it an algorithm, call it what the hell 
>> you want, but I want to know how it produces its stuff.
>> The sort of things I've read so far are; "It has all the Internet at its 
>> call, and summons up vast resources". But that gives nothing away. If 
>> you've ever had to write a computer program, you'll know that is true.
>> 
>> So, how does it do its stuff?
>> 
> 
>   As a programmer you know that it's all math. It's all
> calculations on an abacus. It just gets increasingly complex.
> The marketers say they've lost track of how it's working
> and that must be intelligence and blah, blah, blah. But it's
> still all just binary data and math calculations.

Although true, it's very reductionist. It's like saying reading is just
photons bouncing off a page of a book. 

>   So how do they make these things up? They copy billions
> of lines of data and look for patterns. 

It's more then copied; it is transformed to better inform the model. 

> Notice that your poem
> includes lots of cliches strung together. They mostly don't
> make a lot of sense. But they come close and we read
> into it.
> 
>   So how COULD they do it? There's only one way: To gather
> data and compare it in various ways, then synthesize new
> combinations. That's math, just like any other computer
> operation. You might download porn of a naked woman with
> an amazing ass and go into a long fantasy about having a
> relationship with her. Would it matter whether that woman
> exists or whether it's an AI production? What you're enjoying
> is a bitmap recording of RGB pixel values, projected on a
> screen. There's neither a real woman nor an AI picture there.
> There's a data stream of numeric values interpreted as
> RGB pixel colors. The rest is in your mind.
> 
>   I think that's the interesting thing about people who talk
> about tests for consciousness. We'll know a computer is
> conscious if it can make a human think it's conscious. But
> that assumes the human is conscious. What is conscious?

That's what is really interesting about Artificial Intelligence. It
challenges us to think about what is it that makes "special". We're not, of
course, but we keep thinking we are since at least the ancient greek
philosophers. 

> If I always stop at Dunkin Donuts for a chocolate donut
> when I pass by, is that conscious? If I give my life savings
> to a woman who says she loves me, is that conscious? How
> about if I decide an AI is conscious and empathetic because
> it seems to like me?
> 
>   So, no, AI is not conscious. The more intriguing question
> for me is whether it's possible for humans to be conscious. :)

You can't argue than an AI is not conscious while at the same time asking
whether humans are conscious. 



Back to alt.comp.os.windows-10 | Previous | NextPrevious in thread | Next in thread | Find similar | Unroll thread


Thread

Scary AI Ed Cryer <ed@somewhere.in.the.uk> - 2025-04-08 16:26 +0100
  Re: Scary AI "Alan K." <alan@invalid.com> - 2025-04-08 11:45 -0400
  Re: Scary AI Newyana2 <newyana@invalid.nospam> - 2025-04-08 12:39 -0400
    Re: Scary AI Ed Cryer <ed@somewhere.in.the.uk> - 2025-04-08 19:07 +0100
      Re: Scary AI Newyana2 <newyana@invalid.nospam> - 2025-04-08 14:45 -0400
        Re: Scary AI Ed Cryer <ed@somewhere.in.the.uk> - 2025-04-08 20:35 +0100
          Re: Scary AI jetjock <jetjock@unkown.com> - 2025-04-09 14:01 -0500
        Re: Scary AI Chris <ithinkiam@gmail.com> - 2025-04-08 20:57 +0000
          Re: Scary AI Newyana2 <newyana@invalid.nospam> - 2025-04-08 17:16 -0400
            Re: Scary AI Chris <ithinkiam@gmail.com> - 2025-04-09 19:53 +0000
              Re: Scary AI Newyana2 <newyana@invalid.nospam> - 2025-04-09 16:47 -0400
                Re: Scary AI Chris <ithinkiam@gmail.com> - 2025-04-11 12:23 +0000
                Re: Scary AI Newyana2 <newyana@invalid.nospam> - 2025-04-11 13:31 -0400
                Re: Scary AI Chris <ithinkiam@gmail.com> - 2025-04-13 02:30 +0000
                Re: Scary AI Newyana2 <newyana@invalid.nospam> - 2025-04-13 07:49 -0400
                Re: Scary AI Chris <ithinkiam@gmail.com> - 2025-04-13 13:08 +0000
                Re: Scary AI Chris <ithinkiam@gmail.com> - 2025-04-13 02:38 +0000
            Re: Scary AI Anton Shepelev <anton.txt@gmail.moc> - 2025-04-18 01:40 +0300
              Re: Scary AI Newyana2 <newyana@invalid.nospam> - 2025-04-17 20:22 -0400
                Re: Scary AI Borax Man <rotflol2@hotmail.com> - 2025-04-18 02:12 +0000
                Re: Scary AI Chris <ithinkiam@gmail.com> - 2025-04-18 06:44 +0000
                Re: Scary AI Anton Shepelev <anton.txt@g{oogle}mail.com> - 2025-04-18 14:34 +0300
      Re: Scary AI Chris <ithinkiam@gmail.com> - 2025-04-08 20:46 +0000
        Re: Scary AI Ed Cryer <ed@somewhere.in.the.uk> - 2025-04-08 22:48 +0100
          Re: Scary AI Chris <ithinkiam@gmail.com> - 2025-04-09 09:50 +0000
    Re: Scary AI occam <occam@nowhere.nix> - 2025-04-18 16:45 +0200
  Re: Scary AI Steve Hayes <hayesstw@telkomsa.net> - 2025-04-09 05:29 +0200
  Re: Scary AI Anton Shepelev <anton.txt@gmail.moc> - 2025-04-18 01:02 +0300

csiph-web