Re: US Govt Pols Now Backing 'AI' "Pause" Recommended by Musk, Wozniak

On 4/2/23 8:23 AM, Tristan Miller wrote:
> Greetings.
> 
> On 2023-04-02 14:07, 22T.R732 wrote:
>>> On 2023-03-31 06:31, 22T.R732 wrote:
>>>> Seems like someone in comp.ai deleted the earlier version
>>>> of this
>>>
>>> Your earlier post was automatically rejected by the moderation 
>>> software for excessive cross-posting. 
>>
>>    I am allowed 5 x-posts
> 
> 
> Says who?  Every moderated group (and indeed, every news client and 
> every news server) can set its own limit on the number of cross-posts. 
> You need to comply with *all* of these limits in order for your posts to 
> make it through.
> 
>>    What I *objected* to was that your deletion HERE
>>    managed to propagate to all the OTHER groups included.
> 
> 
> I didn't delete anything, and so there was nothing to "propagate".  Any 
> time you submit an article to a moderated group, even if it is 
> cross-posted to other unmoderated groups, then it needs to pass 
> moderation *before* it appears anywhere on Usenet.  Your article was not 
> deleted from the other groups because it never appeared in them in the 
> first place.
> 
> Your article was clearly on-topic here and you're welcome to continue 
> the discussion here so long as you don't cross-post.  I've started 
> another thread on the question of whether to allow cross-posting to 
> comp.ai.

   On the whole I agree with your concerns. I started with Usenet
   way on back, when it was still "civil". It is no longer "civil".
   Allowing cross-posts, or many cross-posts, is likely to result
   in chaos, flame-wars and worse. I have seen many groups go that
   way and a more 'academic' group like comp.ai does require some
   sort of shield ... yet one that cannot over-tax any moderators.

   However 'AI' has finally reached a state and form where it is
   no longer just "sci-fi musings" but a tangible, sometimes
   threatening, influence on global civilization. Musk and a fair
   number of other prominent tekkie figures are now seeing that,
   and SAYING something about it. This has attracted considerable
   attention, even at the highest levels - regulatory/policy/funding.

   As such, 'comp.ai' is no longer as obscure as it ought to be,
   it's suddenly part of a big 'something'.

   I am not sure what is required these days to create a new
   Usenet group. I would suggest something like "comp.ai.policy"
   to absorb the large-scale, oft crap, postings while still
   being sort of "on target" to those most involved in 'AI'
   research.

   The near-term future of 'Chat', 'Bard' and related approaches
   to, hmm, 'human emulation', is not entirely clear. There DOES
   seem to be a potential for rapid SELF-improvement with these
   sorts of systems. The human mind is in many ways a "hall of
   mirrors" and that effect CAN be emulated ... the system
   observing itself, the world, the effects of the world on
   itself and so forth - making continual improvements.

   People, politicians, money interests and such DO need to worry
   about a Chat-5.0. There are likely many paths to de-facto
   "consciousness" and Chat and/or friends may indeed be one of
   those. "Emulate", "fake", something WELL enough and it is not
   really "fake" anymore ... just "consciousness", "being", by
   alternate means. That has broad *implications* that have,
   until now, been pure sci-fi.

   I saw in the news yesterday that Wal-Mart corp is going to
   fire most of its human warehouse/shipping humans and
   replace them all with 'AI'/"Expert Systems". It tried to
   spin that as the fewer humans getting easier, better-paying,
   jobs. However that leaves the 99% unemployed. Not the ONLY
   corp moving in that direction. This is serious, this is of
   economic significance, this is *political*.

   Remember about a year ago when a Google 'AI' expert came
   to the conclusion that his system WAS effectively
   'conscious' ? They FIRED him for saying so.

   Hmmm ... how many 'AI' profs/experts can Chat-5.0 replace ?

   Beginning to 'get it' ???  :-)

   Mary Shelly was prescient.