MESSAGE
| DATE | 2022-04-18 |
| FROM | Ruben Safir
|
| SUBJECT | Re: [Hangout - NYLXS] AI screwing Free Software - OpenAI
|
On Mon, Apr 18, 2022 at 01:52:43PM -0700, Rick Moen wrote:
> Quoting Ruben Safir (ruben-at-mrbrklyn.com):
>
> > https://www.nytimes.com/2022/04/15/magazine/ai-language.html
>
> I've heard about this, and already seen discussion of the implication
> that this breakthrough in machine-only composition is going to be
> deployed very widely -- with the effect that, soon if not already,
> the someone-you-don't-know who's pushing a point of view from a GMaail
> (etc.) account may no longer be _just_ an Eastern European troll spewing
> partly programatic propaganda (as was that laughably obvious one on
> Hangout in 2016 who claimed dishonestly to be a woman in Germany),
> but may be a fully automated software drivel factory wasting your time
> and doing paid shilling.
The biggest giveaway that this is a NY Times magazine article is that it
is all over the place without a coherent thread. That is classic for
the NY Times magazine, going back to the 1960s. What I found most
important was the slice of Free Software wars that it discussed as to
the openAI status as free software after MS entered into the project and
they argued they can't release the code because it is too dangerous to
do so without being sure the software is not tethered to the correct
ethics.
This is not much different that clearviewai which refuses to stop using
your images until you upload a clear picture of yourself with ID...
REALLY, you can't write stuff like this into a movie script because it
would be seemingly unbelievable. It is the climax of newspeak.
These dangerous corperations are never off message and NEVER lose a
moment to advance their goals... much like Putin and Pottering..
The use of publically available data from individuals for extrapolation
or study should be illegal UPFRONT without written permision by
indivdiuals and should never be allowed to be sold by 3rd parties
without prior specific knolwedge of each individual transaction.
>
> I've already started assuming possible fraud of either of those types
> from Internet entities I don't know.
>
> Just today, there was a subscription onto the SVLUG Web Team mailing
> (which is public, but is a working list for the active volunteers) by an
> unfamiliar GMail address with not even a pretense of an attached name.
> The old version of Mailman is vulnerable to forged subscriptions via
> a Web page authentication flaw, so I normally just quietly unsubscrbe
> _any_ new subscriptions to _any_ of the SVLUG lists, pending a software
> rebuild that may not happen. Today, I unsubbed the suspicious person
> (without notice to the subscriber), but also wrote a pleasant note
> asking if the possible-human wished to help out with SVLUG's Web team,
> and if the subscription was deliberate and desired. The possible-human
> wrote back only "Yes, please.". I replied, "Thanks for that, but, I
> hopen you don't mind saying who and where you are?" The possible-human
> wrote back only "Cody, Illinois."
>
> I'm left with: Probably human, probably severely Aspie if human
> (because the nitwit didn't pick up my implication of "Wait, are you
> someone we know? Why are you acting like you're a Web Team volunteer,
> when your address isn't even on any of the other SVLUG mailing lists?
> Are you offering help, and if so, how come that arrives out of the blue
> from an apparent total stranger? Are you even in Silicon Valley? Why
> the heck would we give you shell on our servers when you can't even
> bother to say who you are?"
>
> Odds are, it's just some net.random in Illinois, some dweeb who thinks
> its cute to be pseudonymous. Anyway, at that point, I just considered
> the matter done and moved on. Just another day of random idiocy on the
> Internet.
>
>
> I don't believe the article's claim that "A few months after GPT-3 went
> online, the OpenAI team discovered that the neural net had developed
> surprisingly effective skills at writing computer software" without
> strong evidence, because I've heard that bushwah before, and
> machine-written programming code has a hilariously bad history.
>
> I also would not believe this bit without a lawyer saying whether the
> alleged licensing agreements or leases were _competent_: "GPT-3 can
> already generate sophisticated legal documents, like licensing
> agreements or leases."
>
> The reporter ducks responsibility for looking into whether the claims
> are bullshit or not with a disappointing waffle about "Some skeptics
> argue...", "Other critics believe...", "Wherever you land in this
> debate" -- but this is _New York Times Magazine_, so I suppose we
> shouldn't expect authors to give a damn about truth or falsity.
>
>
>
> > --
> > So many immigrant groups have swept through our town
> > that Brooklyn, like Atlantis, reaches mythological
> > proportions in the mind of the world - RI Safir 1998
> > http://www.mrbrklyn.com
> > DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
> >
> > http://www.nylxs.com - Leadership Development in Free Software
> > http://www.brooklyn-living.com
> >
> > Being so tracked is for FARM ANIMALS and extermination camps,
> > but incompatible with living as a free human being. -RI Safir 2013
--
So many immigrant groups have swept through our town
that Brooklyn, like Atlantis, reaches mythological
proportions in the mind of the world - RI Safir 1998
http://www.mrbrklyn.com
DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
http://www.nylxs.com - Leadership Development in Free Software
http://www2.mrbrklyn.com/resources - Unpublished Archive
http://www.coinhangout.com - coins!
http://www.brooklyn-living.com
Being so tracked is for FARM ANIMALS and extermination camps,
but incompatible with living as a free human being. -RI Safir 2013
_______________________________________________
Hangout mailing list
Hangout-at-nylxs.com
http://lists.mrbrklyn.com/mailman/listinfo/hangout
|
|
