Where does "self" reside in the brain?

[RonO]

Press release:
https://scopeblog.stanford.edu/2023/06/14/where-in-the-brain-is-my-sense-of-self/

Original article:
https://www.sciencedirect.com/science/article/pii/S0896627323003860?via%3Dihub

Unless you have library access you have to pay to read the original
article. The press release claims that they are trying to determine
where the sense of self resides in the brain. The article decribes a
series of observations when participants do tasks or think about doing
things that need to be related to their self existence. They find brain
activity in a region that they call the PMC (Posteromedial cortex).

Ron Okimoto

[Kalkidas]

The self does not reside in the brain.

[Bob Casanova]

On Sat, 24 Jun 2023 13:24:03 -0700, the following appeared
in talk.origins, posted by Kalkidas <e...@joes.pub>:

>The self does not reside in the brain.
>

So where does it, in your opinion, reside?
>
--

Bob C.

"The most exciting phrase to hear in science,
the one that heralds new discoveries, is not
'Eureka!' but 'That's funny...'"

- Isaac Asimov

[Bob Casanova]

On Sat, 24 Jun 2023 14:02:34 -0700, the following appeared
in talk.origins, posted by Bob Casanova <nos...@buzz.off>:

>On Sat, 24 Jun 2023 13:24:03 -0700, the following appeared
>in talk.origins, posted by Kalkidas <e...@joes.pub>:
>
>>On 6/24/2023 8:22 AM, RonO wrote:
>>> Press release:
>>> https://scopeblog.stanford.edu/2023/06/14/where-in-the-brain-is-my-sense-of-self/
>>>
>>> Original article:
>>> https://www.sciencedirect.com/science/article/pii/S0896627323003860?via%3Dihub
>>>
>>> Unless you have library access you have to pay to read the original
>>> article.  The press release claims that they are trying to determine
>>> where the sense of self resides in the brain.  The article decribes a
>>> series of observations when participants do tasks or think about doing
>>> things that need to be related to their self existence.  They find brain
>>> activity in a region that they call the PMC (Posteromedial cortex).
>>
>>The self does not reside in the brain.
>>
>So where does it, in your opinion, reside?
>>

No idea, beyond "It's not in the brain"? OK.

[JTEM is my hero]

Bob Casanova wrote:

[No idea]

You being clueless, you've put zero thought into any of this,
and probably don't even know how you might. But...

You SEE with your eyes. That's why you assume that
consciousness has to be inside your skull.

Consciousness. Is it unique to us humans? Is it common to
all living things? What?

If consciousness is very old, evolutionarily speaking, there's
no reason why it would reside in the brain. In fact, intuition
told all of humanity that it resided within the heart. The
ancient all believed that the heart was important, not the
brain. And that's actually not a bad guess.

Look. If consciousness is old, evolutionarily speaking, then
put it inside the rib cage, or at least along side all the other
important organs. Whittle our brains to it's oldest, most
basic parts, there isn't much if anything to argue that
consciousness would have to reside there. But, our heart?
Which is right there with our lungs, pancreas & liver? Right
above our kidneys... adrenal gland... gallbladder... you get
the point.

Well. Not YOU, you worthless rotating sock puppet. But
somebody normal can probably see where I'm going with
this: If consciousness is base, or fairly base, then why
wouldn't it reside where nearly all our important stuff is
located, instead of where the primitive, tiny, excessively
limited brain was?




-- --

https://jtem.tumblr.com/post/721153190279643136

[*Hemidactylus*]

That’s correct. There is no self.

[Bob Casanova]

On Mon, 26 Jun 2023 12:03:02 +0000, the following appeared
in talk.origins, posted by *Hemidactylus*
<ecph...@allspamis.invalid>:

I believe that's taking it a bit too far. The "self" (IMHO,
a loose synonym for "consciousness") exists, but not as a
"thing"; as a totality of experiences, and influenced by
heredity.

But I'm sure Kalki, who apparently has me killfiled for
embarrassing him on multiple occasions, would disagree,
since he seems to equate "self" with "soul", which,
according to various beliefs, *is* an independent "thing".

[*Hemidactylus*]

Strict deterministic philosophy and Buddhism both negate the self as
illusion.

[Pandora]

On Mon, 26 Jun 2023 08:45:18 -0700, Bob Casanova <nos...@buzz.off>
wrote:

We should make a distinction between the self and the sense of self.
Every autopoietic system by virtue of its boundary establishes an
inside and an outside, a self and a non-self (the environment)
This is the most basic, first-order, instantiation of a self. The
unicellular organism bounded by its cellmembrane is the archetype.
See chapters 5 and 6 in:
https://www.hup.harvard.edu/catalog.php?isbn=9780674057517

The sense of self is a psychological phenomenon that emerges in
certain second-order autopoietic systems (Metazoa) with highly
developed central nervous and sensory systems. Here too the self
resides within a body, bounded by a 'membrane' (the skin) that
establishes an inside and an outside, a self and a non-self.
The sense of self is the subjective side of this kind of organization
and appears to require a brain.

[Bob Casanova]

On Mon, 26 Jun 2023 16:09:36 +0000, the following appeared

in talk.origins, posted by *Hemidactylus*
<ecph...@allspamis.invalid>:

>Bob Casanova <nos...@buzz.off> wrote:
>> On Mon, 26 Jun 2023 12:03:02 +0000, the following appeared
>> in talk.origins, posted by *Hemidactylus*
>> <ecph...@allspamis.invalid>:
>>
>>> Kalkidas <e...@joes.pub> wrote:
>>>> On 6/24/2023 8:22 AM, RonO wrote:
>>>>> Press release:
>>>>> https://scopeblog.stanford.edu/2023/06/14/where-in-the-brain-is-my-sense-of-self/
>>>>>
>>>>> Original article:
>>>>> https://www.sciencedirect.com/science/article/pii/S0896627323003860?via%3Dihub
>>>>>
>>>>> Unless you have library access you have to pay to read the original
>>>>> article.  The press release claims that they are trying to determine
>>>>> where the sense of self resides in the brain.  The article decribes a
>>>>> series of observations when participants do tasks or think about doing
>>>>> things that need to be related to their self existence.  They find brain
>>>>> activity in a region that they call the PMC (Posteromedial cortex).
>>>>>
>>>>> Ron Okimoto
>>>>>
>>>>
>>>> The self does not reside in the brain.
>>>>
>>>>

>>> That?s correct. There is no self.

>>>
>> I believe that's taking it a bit too far. The "self" (IMHO,
>> a loose synonym for "consciousness") exists, but not as a
>> "thing"; as a totality of experiences, and influenced by
>> heredity.
>>
>> But I'm sure Kalki, who apparently has me killfiled for
>> embarrassing him on multiple occasions, would disagree,
>> since he seems to equate "self" with "soul", which,
>> according to various beliefs, *is* an independent "thing".
>>
>Strict deterministic philosophy and Buddhism both negate the self as
>illusion.
>

OK. I disagree (see comment re: "consciousness"), but that's
my prerogative.

[Bob Casanova]

On Mon, 26 Jun 2023 19:36:04 +0200, the following appeared
in talk.origins, posted by Pandora <pan...@knoware.nl>:

Pretty much agree. So when you (whatever that may mean to
you) typed the above, what exactly did the thinking which
resulted in that typing? And for that matter, what
controlled the fingers doing the typing? Please be specific,
with objective descriptions, regarding the difference
between "self" and "sense of self". How it happens
(emergence from complexity) is not in dispute: what it
actually is, is.

And FWIW, I agree completely with "requires a brain".
although Kalki seems to disagree.

[Martin Harran]

This is a good example of genuine and worthwhile research being overly
hyped beyond any justification. Neuro research is extremely valuable
and may lead to improvements treatment of mental issues but
identifying where activity takes place in the brain tells us nothing
at all about 'self '.

An electronics engineer with little knowledge of computing could
examine my PC and make a pretty good effort at identifying where
various electronic processes are taking place. With a bit of further
training, he could probably work out how different applications are
using the various bits of hardware. None of that, however, would tell
him anything about the thought processes that I am employing to use my
PC to write this argument.

It is the same with the brain - understanding the neural activities
tells us nothing about what 'self' is as a concept, let alone where it
comes from.

[*Hemidactylus*]

I’m probably mangling Dennett badly, but when you perceive the desktop
environment of your graphical user interface, what actually is *that*?

[broger...@gmail.com]

I think Dennett would say that looking for where the self resides within the brain means that you are accepting the idea of a Cartesian theater and looking for the spectator. I'm not sure Dennett would say this, but I'd say the self is the body.

[Burkhard]

and I'd add to that (parts of) the environment with which the body interacts (Andy Clark's "extended mind" and enactivism)

[*Hemidactylus*]

Andy Clark’s notions are intriguing. Dennett may have referred to extended
mind (or a similar idea) as “off-loading”. Notes scattered on my smart
phone are a part of “me”. I have Clark’s book, but haven’t been doing much
reading in cognitive philosophy lately.

I’m tempted to reread Dennett’s self as center of narrative gravity.

[Martin Harran]

You and others having to turn to philosophy highlights the fact that
*science* doesn't really have much to offer in this area.

[broger...@gmail.com]

I agree; science does not have much to offer, since the question is more about how to use words than anything else.

[Martin Harran]

On Tue, 27 Jun 2023 10:36:14 -0700 (PDT), "broger...@gmail.com"
<broger...@gmail.com> wrote:

I think that whatever "self" is, it's a lot more than just words but
anyway, I'll take what you said as some sort of acknowledgement that
the cited article about where "self" resides is pure hyperbole.

[broger...@gmail.com]

Well, obviously the self is more than words; i did not suggest otherwise. But the question about where the self resides is mostly a question about what it is that you want to call the self.

[Öö Tiib]

ChatGPT is not sentient and does not possess consciousness or
self-awareness. Yet its mental abilities are impressive. It can be
matter of years when that self-awareness is also simulated.

[Bob Casanova]

On Mon, 26 Jun 2023 17:01:38 -0700, the following appeared
in talk.origins, posted by Bob Casanova <nos...@buzz.off>:

Well?

[Bob Casanova]

On Tue, 27 Jun 2023 12:20:39 -0700 (PDT), the following
appeared in talk.origins, posted by Öö Tiib <oot...@hot.ee>:

From what I've seen, ChatGPT is what we would have called an
"expert system" when I was working as an engineer before I
retired; it returns exactly what its programmers have
designed it to do. It's far from AI, and has zero "mental
abilities".

[*Hemidactylus*]

Martin Harran <martin...@gmail.com> wrote:
>
[brevity snip]

>
> I think that whatever "self" is, it's a lot more than just words but
> anyway, I'll take what you said as some sort of acknowledgement that
> the cited article about where "self" resides is pure hyperbole.
>
>

Going somewhat on what Dennett gets at here with his three alternative
scenarios based on the now dated Libet experiment:

https://dl.tufts.edu/pdfviewer/1831cw57f/br86bf394

I had kept thinking every time I read it of Mel Gibson’s character in the
final scene of Gallipoli as a homunculus and quoting myself from a post
here back in 2017 when this stuff was fresher in my mind:
“…in the movie Gallipoli, the runner played by Mel Gibson is going back
and forth between important modules just as Dennett argues against the
popular notions of the Libet wrist flick timing experiments. Gibson runs
from front trench to command head quarters and back. And he rushes the
countermanding order analogous to Libet's veto or "free won't" but it was
too late. Yet instead of focusing on Gibson in the Cartesian theater, it
would be better to take the situation as a whole as how human choice works.

It's all there, asynchronous time smearing, delays, different speeds of
transmission, vision versus action, command centers...”

And in another post from 2016: “Dennett gave free will worth having his
best effort which for me came down to communication in the neural trenches
in a very similar manner to what Mel Gibson's character dealt with in the
movie "Gallipoli" where he was an ANZAC runner in the trenches who
ironically could not outrun a suddenly operational faster but lower
ranking and no longer relevant decision communicated over a fixed
telephone line.”

Based roughly on that with the analogy of the Gallipoli trenches being your
brain, where do “you” actually fit in?

The Wikipedia on the film:
“The Light Horse are to attack in three waves across a stretch of ground
defended by Turkish machine gunners. The first wave is to go at 4:30 AM,
after an artillery bombardment. Unfortunately, the commanders' watches are
unsynchronized and the bombardment ends too early. The brigade's commander,
Colonel Robinson, insists the ANZAC attack proceed; the first wave is cut
down by the Turks within seconds. The second wave goes over, to a similar
fate. Major Barton wants to halt the attack to end the carnage, but the
Colonel says that somebody told him ANZAC marker flags were seen in the
Turkish trenches, indicating that the attack was partially successful. The
phone line goes dead, and Barton gives Frank a message to carry to Brigade
HQ, but when he arrives, the Colonel insists the attack continue.

Lieutenant Gray, Major Barton's second-in-command, admits to Barton that he
was the soldier who said that he saw marker flags, though he did not
remember who told him. Frank suggests to the Major that he go over the
Colonel's head to General Gardner. Frank hurries to Gardner's headquarters
down on the beach. The General is informed that, at Suvla, the British
landing party is brewing tea on the beach. He tells Frank that he is
reconsidering the attack. Frank sprints back to convey this news, but the
phone lines are repaired and Colonel Robinson orders the attack to
continue. Barton joins his men in the attack, climbs out of the trench
pistol in hand, and signals his men to charge. Archy joins the last wave
and goes over the top. Frank arrives seconds too late and lets out a scream
of anguish and despair. As Archy's companions are cut down by gunfire, he
drops his rifle and runs as hard as he can. The final frame freezes on
Archy being hit by bullets across his chest, head flung back, as if
breaking the tape at the finish of a 100-yard sprint, and falling
backwards.”

https://en.m.wikipedia.org/wiki/Gallipoli_(1981_film)

From Ron’s OP link I got:
“Our results resonate with the extant literature and our prior findings,
and in addition, we highlight a specific subregion of the PMC in its
anterior precuneus (aPCu) subregion. We suggest that the aPCu is a key node
with access to neural processes that are causally relevant to forming the
subjective sense of a physical and spatial self, although this region is
distinct from the other PMC sites that are part of the DMN,22 which is
well-known for its high relevance to the autobiographical dimension of the
self.”

https://www.sciencedirect.com/science/article/abs/pii/S0896627323003860?via%3Dihub

Note the wording that the “anterior precuneus” being “a key node” and not
the key node. I had long thought that executive functions stemmed from
areas of the dorsolateral prefrontal cortex. The anterior precuneus seems
to be located in the parietal lobe instead. The danger in neuroscience is
succumbing to the ever present temptation of phrenology. It hasn’t gone
away though no longer implies skull bumps.

But whatever the scientific findings which don’t rule out other neural
regions (Gallipoli components) one must do the ontological heavy work of
establishing an actual coherent “self” versus what decomposes to a bunch of
subconscious behind the scenes workings, kinda like what happens at lower
software and hardware levels to give you the perception of an actual
desktop environment. The self may be no more than a useful fiction
construed as a working narrative. It is largely a religious vestige though
rightly dismissed by the Buddhists.

I haven’t given any of this self consciousness related stuff much deep
thought for quite some time so the above is merely a hot take.
“My”interests lie elsewhere at the moment.

[Öö Tiib]

It is more like "just a chat bot" based on trained neural network (not
"programmed"). It can be already used as expert system on
literally any topic. It has trained that all by just reading Internet,
without any actual experience so obviously its reasoning contains
flaws and wild guesses. But if to point out and reason about flaws
in whatever it suggests then it briefly corrects those. So it trains
further despite being already impressive.

[*Hemidactylus*]

Yep: “Our brains are modestly larger than the brains of our nearest
relatives (although not larger than the brains of some dolphins and
whales), but this is almost certainly not the source of our greater
intelligence. The primary source, I want to suggest, is our habit of
offloading as much as possible of our cognitive tasks into the environment
itself—extruding our minds (that is, our mental projects and activities)
into the surrounding world, where a host of peripheral devices we construct
can store, process, and re-represent our meanings, streamlining, enhancing,
and protecting the processes of transformation that are our thinking. This
widespread practice of off-loading releases us from the limitations of our
animal brains.” From _Kinds Of Minds_ by Daniel Dennett

> I’m tempted to reread Dennett’s self as center of narrative gravity.
>

In which he ends with the ultimate skeptical decomposer of “self”: “And
remember that even a center of gravity has a fairly robust presence, once
we start playing around with it. But no one has ever seen or ever will see
a center of gravity. As David Hume noted, no one has ever seen a self,
either.

"For my part, when I enter most intimately into what I call myself, I
always stumble on some particular perception or other, of heat or cold,
light or shade, love or hatred, pain or pleasure. I never can catch myself
at any time without a perception, and never can observe anything but the
perception.... If anyone, upon serious and unprejudiced reflection, thinks
he has a different notion of himself, I must confess I can reason no longer
with him. All I can allow him is, that he may be in the right as well as I,
and that we are essentially different in this particular. He may, perhaps,
perceive something simple and continued, which he calls himself; though I
am certain there is no such principle in me." (Treatise on Human Nature, I,
IV, sec. 6.)”

I couldn’t find a good link to this article in which Dennett also says: “A
center of gravity is just an abstractum. It's just a fictional object. But
when I say it's a fictional object, I do not mean to disparage it; it's a
wonderful fictional object, and it has a perfectly legitimate place within
serious, sober, echt physical science.

A self is also an abstract object, a theorist's fiction.”

[*Hemidactylus*]

In “The Self as a Center of Narrative Gravity” Dennett says: “"Where is the
self?" a materialist philosopher or neuroscientist might ask. It is a
category mistake to start looking around for the self in the brain. Unlike
centers of gravity, whose sole property is their spatio- temporal position,
selves have a spatio-temporal position that is only grossly defined…Brain
research may permit us to make some more fine-grained localizations, but
the capacity to achieve some fine-grained localization does not give one
grounds for supposing that the process of localization can continue
indefinitely and that the day will finally come when we can say, "That cell
there, right in the middle of hippocampus (or wherever)--that's the self!"”

[Bob Casanova]

On Tue, 27 Jun 2023 17:05:47 -0700 (PDT), the following

Is that why it returns fabricated data, including references
to nonexistent court cases which a rather incompetent law
firm used in a filing and got their heads handed to them by
a judge who actually looked up references and found they
didn't exist?

https://news.yahoo.com/lawyer-uses-ai-cites-fake-173634597.html

Thanks, I'll refrain from using it for anything beyond a
joke.

[*Hemidactylus*]

I asked ChatGPT about a book I was reading called The Daughter of Doctor
Moreau by Silvia Moreno-Garcia. It replied with answers from The Madman's
Daughter by Megan Shepherd also based on the HG Wells original. I pushed
back. It apologized profusely (weird) and after much back and forth where
it was cocksure about itself it finally figured out what book I was talking
about. Not a good start but an OK finish. I got into a brutal fist fight
with ChatGPT about the BS label “Social Darwinism”. It didn’t seem to want
to back down over that.

Reading Candice Millard now about Teddy Roosevelt suffering Brazil’s River
of Doubt. Wonder how ChatGPT might handle that. TR’s own book Through the
Brazilian Wilderness is in the public domain. They should converge somewhat
given historical factuality.

[Öö Tiib]

Yes, if something it says feels odd then it is worth to verify if he did
understand what was asked and ask references from where it
gathered that "knowledge". Can be garbage site that some
saboteur has put to some semi-credible site like Wikipedia
as reference for "lulz". But it is trainable and learning and it is
actually surprising that it has managed to collect such data
from net. One will typically get far worse info by googling
themselves. Also its ability to translate is rather decent.

[Bob Casanova]

On Wed, 28 Jun 2023 05:01:37 +0000, the following appeared
in talk.origins, posted by *Hemidactylus*
<ecph...@allspamis.invalid>:

>>>>>>>>> I?m probably mangling Dennett badly, but when you perceive the desktop

The problem is to determine which of its
claims/assertions/"facts" are accurate, which is *not*
trivial (and why use it if you have to verify everything it
comes up with?). Until it's exhibited reliability for an
extended period in multiple subjects I'll continue to regard
it as nothing much better that Eliza, since "making up shit
out of whole cloth is *not* a trait I want in a reference
helper.

And if you're uncomfortable with something originating on
Fox as a reference, here are more about the "Lazy Lawyer":

https://www.legaldive.com/news/chatgpt-fake-legal-cases-generative-ai-hallucinations/651557/
https://www.entrepreneur.com/business-news/lawyer-regrets-using-chatgpt-in-legal-brief-cites-fake/453149

[Bob Casanova]

On Tue, 27 Jun 2023 23:03:13 -0700 (PDT), the following

The point is that it apparently did *not* get it from online
sources; it created it ex nihilo, using it's "intelligence"
(actually sophisticated algorithms; no intrinsic
intelligence involved). *Not* a Good Thing (TM) at any time,
but especially for a legal brief.

>
> One will typically get far worse info by googling
>themselves. Also its ability to translate is rather decent.
>

Here are two more cites; read them to get the info regarding
how (and possibly why*) it generated the fake data:

https://www.legaldive.com/news/chatgpt-fake-legal-cases-generative-ai-hallucinations/651557/
https://www.entrepreneur.com/business-news/lawyer-regrets-using-chatgpt-in-legal-brief-cites-fake/453149

* I've seen the "why" conjectured as a result of the
algorithm structure, with "satisfy the questioner with
*something*" given perhaps more weight than was wise.

[broger...@gmail.com]

It's true that Chat GPT's output is not collected directly from the net. But all the data it is trained on comes from the net. All it is doing is training itself to predict the most likely word to follow an existing string of words, based on a huge number of such correlations in huge amounts of text on the internet. Words are just treated as individual objects that have a greater or lesser chance to be found adjacent to other words. And because our language follows pretty reliable patterns, it can generate very human sounding content. But there's no guarantee that anything it says is true - it does not, for example, recognize a string of its own output as an assertion and then check it against reliable sources - it just predicts what words are likely to follow one another. Because the texts it is trained on often contain true assertions, ChatGPT will often generate texts that contain true assertions, but there's no guarantee. It would not know how to fact check, and it certainly has no common sense.

[Burkhard]

"stochastic parrots" in the words of Emily Bender, Timnit Gebru, Angelina McMillan-Major, and Margaret Mitchell

[Bob Casanova]

On Wed, 28 Jun 2023 15:08:38 -0700 (PDT), the following
appeared in talk.origins, posted by Burkhard
<b.sc...@ed.ac.uk>:

>On Wednesday, June 28, 2023 at 5:20:21?PM UTC+1, broger...@gmail.com wrote:

Good term. Yep.

[Bob Casanova]

On Wed, 28 Jun 2023 09:18:17 -0700 (PDT), the following
appeared in talk.origins, posted by "broger...@gmail.com"
<broger...@gmail.com>:

And apparently has no requirement to use reality for its
answers, as the legal case shows quite well.

My point was that, regardless of where the basic concepts
(briefs, rulings, etc.) came from, it used those concepts to
generate completely spurious citations to nonexistent cases.
So "AI" may be an accurate description after all, since
humans sometimes do the same. And sometimes get caught, as
ChatGPT did.

[Ernest Major]

On 28/06/2023 23:08, Burkhard wrote:
>>> The point is that it apparently did*not* get it from online

>>> sources; it created it ex nihilo, using it's "intelligence"
>>> (actually sophisticated algorithms; no intrinsic

>>> intelligence involved).*Not* a Good Thing (TM) at any time,

>>> but especially for a legal brief.
>> It's true that Chat GPT's output is not collected directly from the net. But all the data it is trained on comes from the net. All it is doing is training itself to predict the most likely word to follow an existing string of words, based on a huge number of such correlations in huge amounts of text on the internet. Words are just treated as individual objects that have a greater or lesser chance to be found adjacent to other words. And because our language follows pretty reliable patterns, it can generate very human sounding content. But there's no guarantee that anything it says is true - it does not, for example, recognize a string of its own output as an assertion and then check it against reliable sources - it just predicts what words are likely to follow one another. Because the texts it is trained on often contain true assertions, ChatGPT will often generate texts that contain true assertions, but there's no guarantee. It would not know how to fact check, and it certainly has no common sense.
>
> "stochastic parrots" in the words of Emily Bender, Timnit Gebru, Angelina McMillan-Major, and Margaret Mitchell
>

Also described as a "blurry JPEG of the web".

https://www.newyorker.com/tech/annals-of-technology/chatgpt-is-a-blurry-jpeg-of-the-web

--
alias Ernest Major

[G]

>>with ChatGPT about the BS label ?Social Darwinism?. It didn?t seem to want

>>to back down over that.
>>

>>Reading Candice Millard now about Teddy Roosevelt suffering Brazil?s River
>>of Doubt. Wonder how ChatGPT might handle that. TR?s own book Through the

>>Brazilian Wilderness is in the public domain. They should converge somewhat
>>given historical factuality.
>>
> The problem is to determine which of its
> claims/assertions/"facts" are accurate, which is *not*
> trivial (and why use it if you have to verify everything it
> comes up with?). Until it's exhibited reliability for an
> extended period in multiple subjects I'll continue to regard
> it as nothing much better that Eliza, since "making up shit
> out of whole cloth is *not* a trait I want in a reference
> helper.
>
> And if you're uncomfortable with something originating on
> Fox as a reference, here are more about the "Lazy Lawyer":
>
> https://www.legaldive.com/news/chatgpt-fake-legal-cases-generative-ai-hallucinations/651557/
> https://www.entrepreneur.com/business-news/lawyer-regrets-using-chatgpt-in-legal-brief-cites-fake/453149
>>

A friend of mine used AI to generate a proposal for a grant, but he knew that
the scientific references where not good (he said AI warns about it), he was
interested in the structure of the proposal and the wording. Only a few
adjustment where needed and with the proper citation he made it in record
time. If you want it to do all your work for you it will be a disaster, but if
you use it as a helping hand and check it properly it can be really useful.

The problem will be when everyone does the same and every proposal or paper or
whatever will be the same.....

G

[Burkhard]

The next big problem will be the next iteration - when LLMs are trained on webdata that in ever increasing amounts have been created by other generative AIs

[Martin Harran]

Sorry, all a bit deep for me.

To be honest, I haven't read much of Dennett; what I did read didn't
excite me enough to read more.

[Bob Casanova]

On Thu, 29 Jun 2023 02:22:42 -0700 (PDT), the following
appeared in talk.origins, posted by Burkhard
<b.sc...@ed.ac.uk>:

>The next big problem will be the next iteration - when LLMs are trained on webdata that in ever increasing amounts have been created by other generative AIs
>

"It's bogus AI data all the way down."

[Bob Casanova]

On 29 Jun 2023 08:15:25 GMT, the following appeared in
talk.origins, posted by G <g...@nowhere.invalid>:

>Bob Casanova <nos...@buzz.off> wrote:
>> On Wed, 28 Jun 2023 05:01:37 +0000, the following appeared
>> in talk.origins, posted by *Hemidactylus*
>> <ecph...@allspamis.invalid>:
>>
>>>Bob Casanova <nos...@buzz.off> wrote:
>>>> On Tue, 27 Jun 2023 17:05:47 -0700 (PDT), the following

>>>> appeared in talk.origins, posted by 嘱 Tiib <oot...@hot.ee>:

>>>>
>>>>> On Wednesday, 28 June 2023 at 02:45:21 UTC+3, Bob Casanova wrote:
>>>>>> On Tue, 27 Jun 2023 12:20:39 -0700 (PDT), the following

>>>>>> appeared in talk.origins, posted by 嘱 Tiib <oot...@hot.ee>:

Apparently he was more intelligent and far less lazy than
the cited lawyer.

>
>The problem will be when everyone does the same and every proposal or paper or
>whatever will be the same.....
>

...and *nothing* can be trusted to be accurate, or even
close. Yep.

[Burkhard]

sort of relevant for this thread, or at least the direction it developed. I'm not a great fan of his, but the best part in his latest book, from Bacteria to Bach and Back, is for me when he takes on Bostrom and the other singularity crap-peddlers - the danger of AI is not that it gets human-like intelligence, but that some idiots think some brittle and bug ridden statistics programs has human-level intelligence and give them tasks they are not suited or.

[Martin Harran]

On Tue, 27 Jun 2023 17:05:47 -0700 (PDT), Öö Tiib <oot...@hot.ee>
wrote:

Sounds quite like some posters here except for the bit about
correcting flaws that are pointed out!

[JTEM is my hero]

Bob Casanova wrote:

> Apparently he

You're a worthless, super disordered fraud cowering behind sock
puppets. I'd worry more about yourself, if I were you, than anyone
else.




-- --

https://jtem.tumblr.com/post/721666107169292289