"AI Van Gogh" interview - Musée d'Orsay

[occam]

There is an interesting article in the Guardian (UK) regarding the
recent Van Gogh exhibition at the Musée d'Orsay.

Sub-heading: "Musée d’Orsay adds AI and VR to display of artist’s last
works, never previously seen together."

A BBC Radio news item this morning explained that 'AI' part of the
exhibition included the analysis of some 900 letters from van Gogh. The
analysis was carried out and incorporated into the VR experience where
visitors could ask questions of the virtual van-G and get 'credible'
answers.

Guardian link:
https://www.theguardian.com/artanddesign/2023/oct/02/ai-vincent-van-gogh-talks-of-mental-health-struggles-in-paris-exhibition

On a positive note, we no longer have to rely on the self-important
opinions of art experts and their interpretations of works of art.

[J. J. Lodder]

Do you really think AI Van Gogh is an improvement?

Jan

--
Q: Why did you shoot yourself in the chest with a revolver,
causing injuries from which you died two days later?

A: This is still a subject of speculation among historians and
specialists. The truth of my motivation remains a mystery even to me.
Thank you for understanding my mental health struggles. (AI Vincent)

[occam]

On 03/10/2023 13:19, J. J. Lodder wrote:
> occam <oc...@nowhere.nix> wrote:
>
>> There is an interesting article in the Guardian (UK) regarding the
>> recent Van Gogh exhibition at the Musée d'Orsay.
>>
>> Sub-heading: "Musée d'Orsay adds AI and VR to display of artist's last
>> works, never previously seen together."
>>
>> A BBC Radio news item this morning explained that 'AI' part of the
>> exhibition included the analysis of some 900 letters from van Gogh. The
>> analysis was carried out and incorporated into the VR experience where
>> visitors could ask questions of the virtual van-G and get 'credible'
>> answers.
>>
>> Guardian link:
>> https://www.theguardian.com/artanddesign/2023/oct/02/ai-vincent-van-gogh-talks
> -of-mental-health-struggles-in-paris-exhibition
>>
>> On a positive note, we no longer have to rely on the self-important
>> opinions of art experts and their interpretations of works of art.
>
> Do you really think AI Van Gogh is an improvement?
>

At least the responses are based on thoughts / writings the artist had
at the time. As opposed to the rather speculative thoughts of art
critics who have done very little of tangible use in their lives, let
alone any painting.

[Athel Cornish-Bowden]

On 2023-10-03 11:19:47 +0000, J. J. Lodder said:

> occam <oc...@nowhere.nix> wrote:
>
>> There is an interesting article in the Guardian (UK) regarding the
>> recent Van Gogh exhibition at the Musée d'Orsay.
>>
>> Sub-heading: "Musée d'Orsay adds AI and VR to display of artist's last
>> works, never previously seen together."
>>
>> A BBC Radio news item this morning explained that 'AI' part of the
>> exhibition included the analysis of some 900 letters from van Gogh. The
>> analysis was carried out and incorporated into the VR experience where
>> visitors could ask questions of the virtual van-G and get 'credible'
>> answers.
>>
>> Guardian link:
>> https://www.theguardian.com/artanddesign/2023/oct/02/ai-vincent-van-gogh-talks
> -of-mental-health-struggles-in-paris-exhibition
>>
>> On a positive note, we no longer have to rely on the self-important
>> opinions of art experts and their interpretations of works of art.
>
> Do you really think AI Van Gogh is an improvement?

Is artificial intelligence an improvement on anything? Why do they talk
about it all the time on the news? When they invented sliding doors
that opened automatically when they sensed a person approaching did
anyone call it artificial intelligence? Why not?

--
Athel -- French and British, living in Marseilles for 36 years; mainly
in England until 1987.

[HVS]

On 03 Oct 2023, occam wrote

I don't think it's that easy to side-step the critics' views: doesn't
the data base on which the AI responses are based include the
writings of said art critics, as well as the thoughts/writings of the
artist?

--
Cheers, Harvey

[Peter Moylan]

On 03/10/23 23:02, Athel Cornish-Bowden wrote:
>
> Is artificial intelligence an improvement on anything? Why do they
> talk about it all the time on the news? When they invented sliding
> doors that opened automatically when they sensed a person
> approaching did anyone call it artificial intelligence? Why not?

The nature of AI has changed over the years. In my younger days -- when
I considered getting into AI research myself -- it was all about looking
for ways to make smart decisions. For example, chess-playing computers
built a tree of predictions of the likely outcomes of various possible
moves. (The size of the tree was limited mainly by the time available
and the speed of the computer, but gains were possible by tricks like
skipping the unlikely-looking cases.) Something similar (in a broad
sense) is still used today in GPS navigation devices.

Today's approach has changed tack. Instead of trying to BE intelligent,
the goal is to APPEAR intelligent. It is no longer about trying to make
good decisions; it's all about looking plausible.

I presume that this has come out of management studies, which seem to
suggest that a manager who is good at plausible bullshit will outperform
one who makes good business decisions. It's a pity Laura is no longer
here; I'd love to hear her take on that.

As for intelligent sliding doors: I think that topic was adequately
covered in The Hitchhiker's Guide to the Galaxy.

--
Peter Moylan http://www.pmoylan.org
Newcastle, NSW

[occam]

According to the BBC interview with the researcher who was responsible
for the AI program, the current training database includs all the
letters written by van Gogh. This amounted to some 900 hundred letters
('around a million words").

It is possible that a future, extended database may include critics' (or
historians') essays. I hope it is more historians' rather than critics'
essays. At least the former have clearly set rules regarding their
speculations.

[Peter T. Daniels]

On Tuesday, October 3, 2023 at 8:02:44 AM UTC-4, Athel Cornish-Bowden wrote:

> Is artificial intelligence an improvement on anything? Why do they talk
> about it all the time on the news?

On last night's Colbert, Neil de Grasse Tyson noted that it isn't "hard
scientists" who are disturbed by AI, it's the "liberal arts" people. Scientists
have been using AI in data analysis for years.

His solution for preventing cheating on class assignments: oral exams.

> When they invented sliding doors
> that opened automatically when they sensed a person approaching did
> anyone call it artificial intelligence? Why not?

They were imitating a Star Trek effect. It was stagehands who opened
and closed doors on the set when characters approached them.

[Jerry Friedman]

On Tuesday, October 3, 2023 at 6:27:20 AM UTC-6, Peter Moylan wrote:
> On 03/10/23 23:02, Athel Cornish-Bowden wrote:
> >
> > Is artificial intelligence an improvement on anything? Why do they
> > talk about it all the time on the news? When they invented sliding
> > doors that opened automatically when they sensed a person
> > approaching did anyone call it artificial intelligence? Why not?

> The nature of AI has changed over the years. In my younger days -- when
> I considered getting into AI research myself -- it was all about looking
> for ways to make smart decisions. For example, chess-playing computers
> built a tree of predictions of the likely outcomes of various possible
> moves. (The size of the tree was limited mainly by the time available
> and the speed of the computer, but gains were possible by tricks like
> skipping the unlikely-looking cases.) Something similar (in a broad
> sense) is still used today in GPS navigation devices.
>
> Today's approach has changed tack. Instead of trying to BE intelligent,
> the goal is to APPEAR intelligent. It is no longer about trying to make
> good decisions; it's all about looking plausible.

...

That's what's getting the publicity because there's been an advance that's
ready for publicity, but I'm sure the search for ways to make smart
decisions is still going on at high speed and being used in areas where
it wasn't before.

--
Jerry Friedman

[occam]

On 03/10/2023 14:27, Peter Moylan wrote:
> On 03/10/23 23:02, Athel Cornish-Bowden wrote:
>>
>> Is artificial intelligence an improvement on anything? Why do they
>> talk about it all the time on the news? When they invented sliding
>> doors that opened automatically when they sensed a person
>> approaching did anyone call it artificial intelligence? Why not?
>
> The nature of AI has changed over the years. In my younger days -- when
> I considered getting into AI research myself -- it was all about looking
> for ways to make smart decisions. For example, chess-playing computers
> built a tree of predictions of the likely outcomes of various possible
> moves. (The size of the tree was limited mainly by the time available
> and the speed of the computer, but gains were possible by tricks like
> skipping the unlikely-looking cases.) Something similar (in a broad
> sense) is still used today in GPS navigation devices.
>
> Today's approach has changed tack. Instead of trying to BE intelligent,
> the goal is to APPEAR intelligent.

The Turing test of AI has always set its goal ('metric') on appearing
intelligent, never on being intelligent.

You are correct in thinking that the approach has changed. In the early
days the approach was for the AI to apply rules which *mimicked* human
rules of reasoning. This didn't always work as expected.

The current paradigm is more statistical. Of all the possible moves
ahead (think of computer chess moves) which is the path most likely to
bring you closest to the best result. In the current example of
conversing with virtual v-Gogh, the most probable solution is the one
most likely to appear in the database of van Gogh letters.

> It is no longer about trying to make
> good decisions; it's all about looking plausible.
>
> I presume that this has come out of management studies, which seem to
> suggest that a manager who is good at plausible bullshit will outperform
> one who makes good business decisions.

No it did not. 'Human rules' of reasoning (as we imagine them) are
notoriously fickle and often prone to errors or loop-holes. Handling
exceptions - when it comes to rule-based reasoning - has proved
incredibly difficult. Millions and millions of rules-of-exception are
required.

The downside of statistical (vs human-rules) AI reasoning is that the
system cannot explain its reasoning when an error occurs. In a
rule-based approach, you could ideally follow the sequence of rules
applied and catch the one that led to the wrong result. Not so with
statistical methods.

>
> As for intelligent sliding doors: I think that topic was adequately
> covered in The Hitchhiker's Guide to the Galaxy.
>

Sliding doors, popping toasters, alarm motion-sensors etc are but the
first step towards intelligent behaviour. There is hardly any reasoning
behind them, except knee-jerk responses.

It is very amusing to watch an automated robot arm that senses a car
body-part approaching and proceeds to paint it in a pre-programmed way.
There are videos of humans accidentally triggering these and getting
sprayed as if they were a car part.

[Bebercito]

That must be because the general level of intelligence has plummeted
since then.

[HVS]

Ah; thanks - I assumed it was using one of the Large Language Model
thingies.

[J. J. Lodder]

See 'Mon Oncle',

Jan

[Peter Moylan]

On 04/10/23 00:26, Peter T. Daniels wrote:

> On last night's Colbert, Neil de Grasse Tyson noted that it isn't
> "hard scientists" who are disturbed by AI, it's the "liberal arts"
> people. Scientists have been using AI in data analysis for years.
>
> His solution for preventing cheating on class assignments: oral
> exams.

In an engineering degree, the preparation and grading of laboratory
reports is a lot of work for both the students and the teaching staff.
The same is true in a number of other disciplines. Known flaws in the
system include the faking of results to match the theory, and the
copying of previous students' lab reports. The latter is possible
because few instructors want to take on a complete overhaul of the set
of experiments each year.

In my final undergraduate year the department took an innovative
approach. At that time we were doing 12 hours of laboratory work per
week, which added up to lots of potential reporting. So written reports
were scrapped, and oral reports used. The students still had to do
enough written recording in the laboratory in order to be prepared, of
course. Plus whatever was needed to analyse the results.

But oral reports are also labour-intensive, so a new tweak was added. We
had to turn up at 9 am each Tuesday (I think it was), prepared for oral
cross-examination. At that time a list was posted of who had to give a
report. If your name wasn't on the list, you were off the hook. There
was about a 30% chance of being called up, so everyone had to be prepared.

[occam]

I'm afraid the general intelligence level was never as high as you are
assuming. It feels, however, that levels are plummeting because social
media allow us to hear more opinions than before. Welcome to the flip
side of democracy.

[Dingbat]

On Tuesday, October 3, 2023 at 2:56:47 PM UTC+5:30, occam wrote:
> There is an interesting article in the Guardian (UK) regarding the
> recent Van Gogh exhibition at the Musée d'Orsay.
>

FWIW, this museum is housed in the former train station Gare d'Orsay,
or train terminal in US English (like Grand Central Terminal) since
trains terminated (and originated) there. It is now a museum and
train station on the RER C line.

>
> Sub-heading: "Musée d’Orsay adds AI and VR to display of artist’s last
> works, never previously seen together."
>
> A BBC Radio news item this morning explained that 'AI' part of the
> exhibition included the analysis of some 900 letters from van Gogh. The
> analysis was carried out and incorporated into the VR experience where
> visitors could ask questions of the virtual van-G and get 'credible'
> answers.
>

AI Van Gogh presumably speaks French in Paris. How does he pronounce his
name for French listeners? And do those listeners reproduce his pronunciation?

[occam]

On 04/10/2023 06:32, Dingbat wrote:
> AI Van Gogh presumably speaks French in Paris. How does he pronounce his
> name for French listeners? And do those listeners reproduce his pronunciation?

The short excerpt I heard of the Q&A session between AI van Gogh and a
journalist was in English.

And you are right, the pronunciation of his name was not how a Dutch
person would pronounce it. None of that 'loch' sound either at the
beginning or the end of his short name.

[J. J. Lodder]

Dingbat <ranjit_...@yahoo.com> wrote:

> On Tuesday, October 3, 2023 at 2:56:47?PM UTC+5:30, occam wrote:
> > There is an interesting article in the Guardian (UK) regarding the
> > recent Van Gogh exhibition at the Musée d'Orsay.
> >
> FWIW, this museum is housed in the former train station Gare d'Orsay,
> or train terminal in US English (like Grand Central Terminal) since
> trains terminated (and originated) there. It is now a museum and
> train station on the RER C line.
> >
> > Sub-heading: "Musée d'Orsay adds AI and VR to display of artist's last
> > works, never previously seen together."
> >
> > A BBC Radio news item this morning explained that 'AI' part of the
> > exhibition included the analysis of some 900 letters from van Gogh. The
> > analysis was carried out and incorporated into the VR experience where
> > visitors could ask questions of the virtual van-G and get 'credible'
> > answers.
> >
> AI Van Gogh presumably speaks French in Paris. How does he pronounce his
> name for French listeners? And do those listeners reproduce his pronunciation?

Just like Vincent himself did in his home dialect,
so when in Eastern Brabant. [1]
So just like a signficant fraction of the Dutch,
and the Belgians and the French.
So my guess: Vincent's pronunciation of his name -when speaking French-
would have been close to the way the French are pronouncing 'Van Gogh'
nowadays.
The extreme 'ch' pronunciation is Northern Dutch.
BTW, it is amusing to see Anglosaxons worrying about that 'ch',
without understanding that their 'Van' is already off.

Jan


[1] 'Van Gokh' is an approximation that will do, for understanding.
The ch and k(h) are not too far apart, in some circumstances.
For example, Dutch uses the 'ch' for Russian transliterations
where English uses the 'kh', like in Sacharov-Sakharov.

[Dingbat]

We got a Garrard record player from the US in 1965. It was more automated
than what we were used to. 'It has a brain!' my uncle exclaimed. FWIW, the
term AI was not yet in vogue.

[occam]

On 03/10/2023 14:02, Athel Cornish-Bowden wrote:
> Is artificial intelligence an improvement on anything?

Some cases of AI are. Take as example the diagnosis of AI-Assisted
Tuberculosis diagnosis from X Ray imagery. Remember that no single
doctor will have seen all these images, let alone processed them. Also
recall that the image database is being added to all the time, and the
AI being retrained continuously.

"The models trained with this method have achieved 96.91% accuracy,
99.38% AUC, 91.81% sensitivity, and 98.42% specificity on a multiclass
classification dataset. "

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9550434/

> Why do they talk
> about it all the time on the news? When they invented sliding doors that
> opened automatically when they sensed a person approaching did anyone
> call it artificial intelligence? Why not?

If you think sliding doors is an example of AI, you need to review your
understanding of AI. There is a huge difference between 'hardwired
response' and 'intelligence'.

[Peter Moylan]

On 04/10/23 19:02, J. J. Lodder wrote:

> Dingbat <ranjit_...@yahoo.com> wrote:

>> AI Van Gogh presumably speaks French in Paris. How does he
>> pronounce his name for French listeners? And do those listeners
>> reproduce his pronunciation?
>
> Just like Vincent himself did in his home dialect, so when in
> Eastern Brabant. [1] So just like a signficant fraction of the Dutch,
> and the Belgians and the French. So my guess: Vincent's pronunciation
> of his name -when speaking French- would have been close to the way
> the French are pronouncing 'Van Gogh' nowadays.

I'm still confused. What I hear from French speakers, including
francophone Belgians, is "van gog", with a hard 'g' in both positions.
What I hear from Dutch speakers is "van gog" with a Dutch soft 'g' in
both positions. And what I hear from English speakers is either "van
goff", with a hard 'g', or occasionally "van goch", with a German
achlaut. So I'm still not sure what you mean.

> The extreme 'ch' pronunciation is Northern Dutch. BTW, it is amusing
> to see Anglosaxons worrying about that 'ch', without understanding
> that their 'Van' is already off.

Because of the vowel? Surely that's a minor detail.

> [1] 'Van Gokh' is an approximation that will do, for understanding.
> The ch and k(h) are not too far apart, in some circumstances. For
> example, Dutch uses the 'ch' for Russian transliterations where
> English uses the 'kh', like in Sacharov-Sakharov.

I'm only able to produce three versions of that guttural sound: the
German ichlaut, the German achlaut, and the one even further back in the
throat that we transcribe as 'kh'. (No doubt some people, for example
Arabic speakers, have consonants that don't fit neatly into that
three-way distinction. In fact I'm currently struggling with the Irish
'gh", which seems to be even further back in the throat.) I'm not sure
which one you mean.

[Peter Moylan]

On 04/10/23 22:00, occam wrote:
> On 03/10/2023 14:02, Athel Cornish-Bowden wrote:
>> Is artificial intelligence an improvement on anything?
>
> Some cases of AI are. Take as example the diagnosis of AI-Assisted
> Tuberculosis diagnosis from X Ray imagery. Remember that no single
> doctor will have seen all these images, let alone processed them.
> Also recall that the image database is being added to all the time,
> and the AI being retrained continuously.
>
> "The models trained with this method have achieved 96.91% accuracy,
> 99.38% AUC, 91.81% sensitivity, and 98.42% specificity on a
> multiclass classification dataset. "

Good example. Medical imagery is a case where classifications based on a
large database are probably the best way of deriving interpretations.

What you can do with AI depends a great deal on the application area.
Recent publicity has concentrated almost entirely on extrapolations of
text document analysis. That is, in my opinion, one area where bullshit
will dominate over intelligent analysis.

[occam]

This video touches on a number of the points raised so far re: the
pronunciation of 'van Gogh'.

https://www.youtube.com/watch?v=KjOvPZTURLI

- Yes, BrE pronunciation of 'van' is wrong
- 'Goff' (BrE) and 'Gowe' (AmE) both wrong
- 'khokh' (Dutch)

By the way, the Flemish (Belgian) sound is softer than the one the Dutch
make, despite what Jan says.

[Peter T. Daniels]

NdeGT noted that doing it his way would be labor-intensive and would
require the training, and appropriately paying, a whole lot more teachers
(well, except maybe in Florida) -- of which the audience in the theater
expressed enthusiastic approval.

His new book (unlike many of his earlier ones) actually sounds interesting.
It's called *To Infinity -- and Beyond!* (from NatGeo). He covers the
development of ideas about the nature of the universe, ending each
chapter pointing out the flaws in an SF movie (the title, of course, is Buzz
Lightyear's catchphrase from the *Toy Story* movies).

[Peter T. Daniels]

On Wednesday, October 4, 2023 at 8:10:43 AM UTC-4, Peter Moylan wrote:
> On 04/10/23 19:02, J. J. Lodder wrote:
> > Dingbat <ranjit_...@yahoo.com> wrote:

> >> AI Van Gogh presumably speaks French in Paris. How does he
> >> pronounce his name for French listeners? And do those listeners
> >> reproduce his pronunciation?
> > Just like Vincent himself did in his home dialect, so when in
> > Eastern Brabant. [1] So just like a signficant fraction of the Dutch,
> > and the Belgians and the French. So my guess: Vincent's pronunciation
> > of his name -when speaking French- would have been close to the way
> > the French are pronouncing 'Van Gogh' nowadays.
>
> I'm still confused. What I hear from French speakers, including
> francophone Belgians, is "van gog", with a hard 'g' in both positions.
> What I hear from Dutch speakers is "van gog" with a Dutch soft 'g' in
> both positions. And what I hear from English speakers is either "van
> goff", with a hard 'g', or occasionally "van goch", with a German
> achlaut. So I'm still not sure what you mean.

Before Woody Allen made a joke of it in *Annie Hall*, AmE was
simply "Van Go," perhaps by analogy with "though" and "dough."

> > The extreme 'ch' pronunciation is Northern Dutch. BTW, it is amusing
> > to see Anglosaxons worrying about that 'ch', without understanding
> > that their 'Van' is already off.
>
> Because of the vowel? Surely that's a minor detail.

Initial [f]. maybe?

> > [1] 'Van Gokh' is an approximation that will do, for understanding.
> > The ch and k(h) are not too far apart, in some circumstances. For
> > example, Dutch uses the 'ch' for Russian transliterations where
> > English uses the 'kh', like in Sacharov-Sakharov.
>
> I'm only able to produce three versions of that guttural sound: the
> German ichlaut, the German achlaut, and the one even further back in the
> throat that we transcribe as 'kh'. (No doubt some people, for example
> Arabic speakers, have consonants that don't fit neatly into that
> three-way distinction. In fact I'm currently struggling with the Irish
> 'gh", which seems to be even further back in the throat.) I'm not sure
> which one you mean.

Arabic Ghain is just a voiced [x] (Achlaut), so a lot easier than folks
imagine. Arabic doesn't have a voiced counterpart of the h-dot in
Muhammad.

Speaking of pronouncing familiar names, long ago the Art Institute of
Chicago had a comprehensive Degas show, and the curator and director,
not to mention the AudioGuide, made such a big deal of it in every media
interview that for years Chicagoans were going around saying "Degas not
Dégas" -- shwa not [e(j)].

[Athel Cornish-Bowden]

Doubly wrong, I think: the v and the a are both wrong, are they not? We
get the n right, however.

> - 'Goff' (BrE) and 'Gowe' (AmE) both wrong
> - 'khokh' (Dutch)
>
> By the way, the Flemish (Belgian) sound is softer than the one the Dutch
> make, despite what Jan says.

[Bertel Lund Hansen]

Peter Moylan wrote:

>> The extreme 'ch' pronunciation is Northern Dutch. BTW, it is amusing
>> to see Anglosaxons worrying about that 'ch', without understanding
>> that their 'Van' is already off.
>
> Because of the vowel? Surely that's a minor detail.

I think he means the Dutch v which lies between English (and Danish) v
and f.

>> [1] 'Van Gokh' is an approximation that will do, for understanding.
>> The ch and k(h) are not too far apart, in some circumstances. For
>> example, Dutch uses the 'ch' for Russian transliterations where
>> English uses the 'kh', like in Sacharov-Sakharov.
>
> I'm only able to produce three versions of that guttural sound: the
> German ichlaut, the German achlaut, and the one even further back in the
> throat that we transcribe as 'kh'.

I have learnt a bit Dutch from two Dutch persons, and I used the German
ach-sound without getting any objections. So if I am not mistaken the
pronunciation is: (vf)an Choch.

--
Bertel, Denmark

[J. J. Lodder]

Is it really so hard to understand that there is no such thing
as -the- Dutch pronunciation of 'Van Gogh'?

Jan

[J. J. Lodder]

Peter Moylan <pe...@pmoylan.org.invalid> wrote:

> On 04/10/23 19:02, J. J. Lodder wrote:
> > Dingbat <ranjit_...@yahoo.com> wrote:
>
> >> AI Van Gogh presumably speaks French in Paris. How does he
> >> pronounce his name for French listeners? And do those listeners
> >> reproduce his pronunciation?
> >
> > Just like Vincent himself did in his home dialect, so when in
> > Eastern Brabant. [1] So just like a signficant fraction of the Dutch,
> > and the Belgians and the French. So my guess: Vincent's pronunciation
> > of his name -when speaking French- would have been close to the way
> > the French are pronouncing 'Van Gogh' nowadays.
>
> I'm still confused. What I hear from French speakers, including
> francophone Belgians, is "van gog", with a hard 'g' in both positions.
> What I hear from Dutch speakers is "van gog" with a Dutch soft 'g' in
> both positions. And what I hear from English speakers is either "van
> goff", with a hard 'g', or occasionally "van goch", with a German
> achlaut. So I'm still not sure what you mean.

My point was that Van Gogh himself would have pronounced his own name
-in fluent speach- in different ways, when speaking different languages.
So like Dutch when speaking Dutch, like Brabants/Flemish when speaking
Brabants/Flemish, and like the French when speaking French. [1]

Vincent was fluent in English too,
but I have no idea how he would have pronounced his name
in the middle of an English sentence,
(it may well have been English-like)

Jan

[1] I don't know how the Van Gogh family spoke at home,
but it seems likely that this was standard Dutch. (so Hollands)
The Van Goghs were a well to do cultured family from Holland
with roots in The Hague, Leiden, and Amsterdam.
At the university, provincial dialect would not have been tolerated.

[J. J. Lodder]

Then you got it wrong. For most speakers of Dutch
the 'g' and the 'ch' in 'Goch' are obviously different.

Jan

[Peter Moylan]

On 05/10/23 03:50, Bertel Lund Hansen wrote:
> Peter Moylan wrote:
>
>>> The extreme 'ch' pronunciation is Northern Dutch. BTW, it is
>>> amusing to see Anglosaxons worrying about that 'ch', without
>>> understanding that their 'Van' is already off.
>>
>> Because of the vowel? Surely that's a minor detail.
>
> I think he means the Dutch v which lies between English (and Danish)
> v and f.

Ah yes, of course. I tend to do that one as a bilabial, although I've
never tried to see whether that's what Dutch speakers do.

[Dingbat]

I've heard French Rose and English Whore begin with guttural sounds.
The latter is the only English word I've heard with a guttural sound,
although the slang pronunciation is Hoe.

>
> the
> German ichlaut, the German achlaut, and the one even further back in the
> throat that we transcribe as 'kh'.
>

IMHO, German achlaut is the furthest back guttural sound,
so nothing meets your description of even further back.
AFAIK, Russian Sakhar (sugar) has the same sound. It's
further back than the one in Dutch Gogh. The sound in
Ahmed is even further back but doesn't meet my conception
of guttural.

>
> (No doubt some people, for example
> Arabic speakers, have consonants that don't fit neatly into that
> three-way distinction. In fact I'm currently struggling with the Irish
> 'gh", which seems to be even further back in the throat.) I'm not sure
> which one you mean.
>

Irish gh and Arabic ghayn are as far back as German achlaut
but they are voiced. FWIW, Arabic ghazia is written by a
French author Jacques Ellul as razia, so the French might
view ghayn and French r as the same. I'm not sure to what
extent this view is correct; I have heard their r sound like
fricative ghayn but I've also heard it as a uvular trill in Rouen.

[Dingbat]

It's awfully hard to get Anglophones to not pronounce <a> as [&].
They do that in Indian words too.

>
> [1] 'Van Gokh' is an approximation that will do, for understanding.
> The ch and k(h) are not too far apart, in some circumstances.
> For example, Dutch uses the 'ch' for Russian transliterations
> where English uses the 'kh', like in Sacharov-Sakharov.
>

German Hoch (name meaning high) is fairly similar to Dutch
Gogh and easier for some Anglophones who pronounce it
like in German. Even while they anglicize Koch!

[occam]

Is it hard to understand that I, as a non-speaker of Dutch, can tell a
Flemish speaker from a Dutch speaker at a hundred paces?

Despite regional variations in both, there is a discernible difference
between them.

[occam]

I know what you are referring to. There are also similar examples of
image fakery. Extrapolations of two or more images into one fake
composite. This should not detract from the essential power of the
underlying techniques, however.

My current preferred online text translator is something called
DeepL(.com). It is - hands down - a superior translator to Google
translate. It makes use of statistical techniques to come up with the
'best' translation of a text, in context. This is unlike the
traditional Google translate, in which a word-for-word translation is
(was?) glaringly obvious.

[Peter Moylan]

My ex-wife, a francophone born in Antwerp (therefore also fluent in
Dutch), switched accent each time she crossed the Belgian/Dutch border.
I didn't know what she was saying, but I could hear the difference.

She, in her turn, claimed that she couldn't understand me (in English)
when we travelled in Victoria. Australians can hear almost no difference
between NSW English and Victorian English, but she could.

[Athel Cornish-Bowden]

On 2023-10-05 05:28:25 +0000, occam said:

> On 04/10/2023 22:41, J. J. Lodder wrote:
>>

>> [ … ]

>>
>> Is it really so hard to understand that there is no such thing
>> as -the- Dutch pronunciation of 'Van Gogh'?
>>
>
> Is it hard to understand that I, as a non-speaker of Dutch, can tell a
> Flemish speaker from a Dutch speaker at a hundred paces?

My wife can tell an Argentinian from a Chilean at a hundred paces, but
it's not from the sound (at that distance) but from the way they move.
Once we were in a train in China, at one end of a carriage. There was
group of people at the far end, and my wife said "They're Argentinian".
I was sceptical, as we couldn't hear them, so she walked down to their
end of the carriage, and reported back: "They're Argentinian".

>
> Despite regional variations in both, there is a discernible difference
> between them.

[Peter Moylan]

On 05/10/23 17:06, Athel Cornish-Bowden wrote:

> My wife can tell an Argentinian from a Chilean at a hundred paces,
> but it's not from the sound (at that distance) but from the way they
> move. Once we were in a train in China, at one end of a carriage.
> There was group of people at the far end, and my wife said "They're
> Argentinian". I was sceptical, as we couldn't hear them, so she
> walked down to their end of the carriage, and reported back: "They're
> Argentinian".

I can tell men from women at a great distance, far enough that the more
obvious differences are ambiguous. Men walk with their elbows out, and
women walk with their elbows in.

[Bertel Lund Hansen]

occam wrote:

> My current preferred online text translator is something called
> DeepL(.com). It is - hands down - a superior translator to Google
> translate. It makes use of statistical techniques to come up with the
> 'best' translation of a text, in context. This is unlike the
> traditional Google translate, in which a word-for-word translation is
> (was?) glaringly obvious.

As far as I know, Google Translate also uses statistics. And calling it
an obvious word-to-word-translation is obviously false.

Maybe a year ago I compared the two translators and found that Google
was doing slightly better. So that is what I have used for words or
expressions that I could not find in a dictionary - something that
happens when I know the Danish word and want the English one. I have no
dictionaries where the source language is Danish (except Da-Da).

Just now I translated the quoted section with both translators. There
are minor differences, but they both did well. DeepL gains points by
making some expressions more natural than Google.

--
Bertel, Denmark

[Bertel Lund Hansen]

Peter Moylan wrote:

>> I think he means the Dutch v which lies between English (and Danish)
>> v and f.
>
> Ah yes, of course. I tend to do that one as a bilabial, although I've
> never tried to see whether that's what Dutch speakers do.

I try to make it something between the v and f, but I am unsure every
time, and I don't know if I get it right (sometimes). I have the same
uncertainty with Spanish b and v.

Luckily it wouldn't matter much if I were to speak to people of those
languages.

--
Bertel, Denmark

[occam]

Aside: I just learned the French expression 'les coudes serres'.
Literally - "with their elbows tucked in". Metaphorically it means 'they
stick together', or more generally support each other.

That's women, in a nutshell.

[Athel Cornish-Bowden]

OK, but although Chileans and Argentinians are different from one
another they're not as different as men and women!

[J. J. Lodder]

So you still have not got the point.
There is no such thing as -a- Flemish speaker and -a- Dutch speaker,
if -a- is supposed to mean typical.
The differences will be unmistakable in some cases,
and nearly inaudible in others.

You may even make mistakes, also at less then a hundred paces.
To be correct you need to replace your 'a' with 'some'.

Jan

[J. J. Lodder]

She had a talent for it, no doubt.
Many Flemish speakers were actually bilingual like that.
You no doubt know that the Belgians had their 'language war' going,
in those days.

The Wallons pestered the Flemish by saying:
you don't really speak a real language, like us.
(happily ignoring that 'Belgicisme' is a French word too)
What you speak is just some 'patois', they claimed.

So many Belgians educated themselves
into speaking and writing standard Dutch, so ABN,
(as it was then called)
while claiming that they spoke real Dutch, really.
And indeed, some of them are quite good at it.

All that is mostly over. The Flemish won the language wars,
and they are developing a more or less standard Flemish,
which is obviously different from standard Dutch.

Jan

[Peter Moylan]

On 05/10/23 20:19, J. J. Lodder wrote:

> Peter Moylan <pe...@pmoylan.org.invalid> wrote:

>> My ex-wife, a francophone born in Antwerp (therefore also fluent
>> in Dutch), switched accent each time she crossed the Belgian/Dutch
>> border. I didn't know what she was saying, but I could hear the
>> difference.
>>
>> She, in her turn, claimed that she couldn't understand me (in
>> English) when we travelled in Victoria. Australians can hear almost
>> no difference between NSW English and Victorian English, but she
>> could.
>
> She had a talent for it, no doubt. Many Flemish speakers were
> actually bilingual like that. You no doubt know that the Belgians had
> their 'language war' going, in those days.

I encountered many examples of it, including a refusal to speak to me if
I said something in French. Something that struck on my first visit to
Brussels was if you phoned a big company they always answered in
English, obviously from fear of offending the caller by answering in the
wrong language.

One example I liked had two factories facing each other across a main
road, near the north end of Brussels. One had a big sign up saying "Hier
wij spreken vlaams". The other responded with an equally big sign saying
"Ici on ne parle pas. On travail."

I got a clearer picture by having in-laws in Antwerp, who are part of
the about 5% of the population that is French-speaking. Their family
name is clearly Flemish, and my father-in-law traced his descent from
someone who came from the east to the big city (Antwerpen) to make his
fortune. As the family became more prosperous it gave up the peasant
language and switched to using French, the language of the upper class.
At some stage there was probably also a switch of religion, again to fit
in with the upper class. The language division was purely based on a
class division. It still has a bit of the flavour of class warfare.

When my children were about two years old we stayed with my wife's aunt
and uncle in Antwerp. They were quite prosperous, and employed a woman
to do the cleaning and cooking. One day I heard my son, in another room,
saying "This is a spoon"; my wife's aunt replied with "C'est une
cuillère." Immediately afterwards, a voice from the kitchen said "Dat's
een lepel". A good education for the children, I suppose.

(And for me, I guess, because until then I didn't know the word "lepel".)

[occam]

On 05/10/2023 12:16, Peter Moylan wrote:
> On 05/10/23 20:19, J. J. Lodder wrote:
>> Peter Moylan <pe...@pmoylan.org.invalid> wrote:
>
>>> My ex-wife, a francophone born in Antwerp (therefore also fluent
>>> in Dutch), switched accent each time she crossed the Belgian/Dutch
>>>  border. I didn't know what she was saying, but I could hear the
>>> difference.
>>>
>>> She, in her turn, claimed that she couldn't understand me (in
>>> English) when we travelled in Victoria. Australians can hear almost
>>> no difference between NSW English and Victorian English, but she
>>> could.
>>
>> She had a talent for it, no doubt. Many Flemish speakers were
>> actually bilingual like that. You no doubt know that the Belgians had
>> their 'language war' going, in those days.
>
> I encountered many examples of it, including a refusal to speak to me if
> I said something in French. Something that struck on my first visit to
> Brussels was if you phoned a big company they always answered in
> English, obviously from fear of offending the caller by answering in the
> wrong language.
>
> One example I liked had two factories facing each other across a main
> road, near the north end of Brussels. One had a big sign up saying "Hier
> wij spreken vlaams". The other responded with an equally big sign saying
> "Ici on ne parle pas. On travail."

I had not heard of that story until now. I am sure however that you are
aware of the much better publicized story of the division of the
*library* of the University of Leuven into two sections - the
Francophone institution (Leuven-la-Neuve) and the Flemish institution
(KU Leuven):

Summary of the algorithm for divvying up the books:

"The administrators eventually settled on some arbitrary ground rules,
which seemed fairest under the circumstances. If a work’s donor could be
contacted, the donor could choose where it would go—and if there was
more than one version of the same work, each institution would be
guaranteed at least one copy. But the majority of works were just
divided by shelfmark: Odds stayed in Leuven, evens left for
Louvain-la-Neuve."


The longer version:
<https://www.atlasobscura.com/articles/when-was-the-leuven-library-divided>

[Peter Moylan]

I
>
have friends at Louvain-la-Neuve, and have visited the place more than
once, so I know about the library.

The campus is very different from any other university campus I've ever
visited. There is no visible distinction between the town and the
university; they are intertwined, to give an overall rural feel to the
university. I know this happens in some other places too, but it's the
only such example I've experienced.

[Peter T. Daniels]

On Wednesday, October 4, 2023 at 7:23:08 PM UTC-4, Dingbat wrote:

> I've heard French Rose and English Whore begin with guttural sounds.
> The latter is the only English word I've heard with a guttural sound,
> although the slang pronunciation is Hoe.

It's either [hOr] or ][huwr] (rhymes with boar or boor). Guttural highly
unlikely, though words like "human" often begin with an Ich-Laut.

[Peter T. Daniels]

On Wednesday, October 4, 2023 at 7:36:37 PM UTC-4, Dingbat wrote:

> It's awfully hard to get Anglophones to not pronounce <a> as [&].
> They do that in Indian words too.

_British_ Anglophones. See my constant complaints about the
mispronunciation of "Barack" and "Las (Vegas)."

English has the word "van," so why would we expect the "Van"
of Beethoven or Gogh to be different?

[Peter T. Daniels]

On Wednesday, October 4, 2023 at 7:36:37 PM UTC-4, Dingbat wrote:

> like in German. Even while they anglicize Koch!

Thee are umpteen pronunciations of that name. The mayor of New York
was Kotch. The prominent Ann Arbor family (and street) is Cook. The
rightwing billionaire brothers are Coke.

[Peter T. Daniels]

On Thursday, October 5, 2023 at 3:27:56 AM UTC-4, occam wrote:

> Aside: I just learned the French expression 'les coudes serres'.
> Literally - "with their elbows tucked in". Metaphorically it means 'they
> stick together', or more generally support each other.

Virgil Thomson claimed that that was how Francis Poulenc played the piano!

[Phil Carmody]

Athel Cornish-Bowden <athe...@gmail.com> writes:
> On 2023-10-03 11:19:47 +0000, J. J. Lodder said:

>> occam <oc...@nowhere.nix> wrote:
>>> There is an interesting article in the Guardian (UK) regarding the
>>> recent Van Gogh exhibition at the Musée d'Orsay.
>>>

>>> Sub-heading: "Musée d'Orsay adds AI and VR to display of artist's last
>>> works, never previously seen together."
>>>
>>> A BBC Radio news item this morning explained that 'AI' part of the
>>> exhibition included the analysis of some 900 letters from van Gogh. The
>>> analysis was carried out and incorporated into the VR experience where
>>> visitors could ask questions of the virtual van-G and get 'credible'
>>> answers.
>>>

>>> Guardian link:
>>> https://www.theguardian.com/artanddesign/2023/oct/02/ai-vincent-van-gogh-talks
>> -of-mental-health-struggles-in-paris-exhibition
>>>
>>> On a positive note, we no longer have to rely on the self-important
>>> opinions of art experts and their interpretations of works of art.
>>
>> Do you really think AI Van Gogh is an improvement?
>
> Is artificial intelligence an improvement on anything? Why do they
> talk about it all the time on the news? When they invented sliding
> doors that opened automatically when they sensed a person approaching
> did anyone call it artificial intelligence? Why not?

Douglas Adams devoted a bit of space to that concept in HHGTTG.

Phil
--
We are no longer hunters and nomads. No longer awed and frightened, as we have
gained some understanding of the world in which we live. As such, we can cast
aside childish remnants from the dawn of our civilization.
-- NotSanguine on SoylentNews, after Eugen Weber in /The Western Tradition/

[Phil Carmody]

Peter Moylan <pe...@pmoylan.org.invalid> writes:
> Today's approach has changed tack. Instead of trying to BE intelligent,
> the goal is to APPEAR intelligent. It is no longer about trying to make
> good decisions; it's all about looking plausible.

Absolutely. And, worse, this judgement is done by people who really
haven't got a clue. Every so often I put the latests AI LLMs to the
test, and every single time I think we're barely improved since the 80s,
which was when I first encountered Eliza clones.

Here's an example:
--> Can you give me some pop quiz questions where the answers are song names that contain the word "blue"?

LLaMa served by Perplexity Labs
Sure, here are five pop quiz questions with answers that are song titles containing the word "blue":

Which classic rock ballad, recorded by The Police in 1983, features the lyrics "Every breath you take, every step you make, every move you make, every bond you break"?
Answer: "Every Breath You Take."
Which 1980s hit single by Cyndi Lauper includes the lyric "He's got a lot of pretty, pretty, pretty, pretty blue eyes"?
Answer: "True Colors."
Which 1970s soft rock hit by Carole King features the lyrics "It's been a long day without you, my friend"?
Answer: "So Far Away."
Which 1960s folk-rock anthem by Simon & Garfunkel begins with the lines "In a green, green meadow / We built a fire / And we sat down / To watch the sunset"?
Answer: "The Sound of Silence."
Which 1980s power ballad by Journey features the lyrics "Don't stop believin', hold on to that feelin'"?
Answer: "Don't Stop Believin'."

Things got so bad, I started taking copies of the transcripts, so that I
could have access to instant examples of stupidity, such as the above.

More travesties at: http://fatphil.org/AI/

[Paul Wolff]

On Sat, 7 Oct 2023, at 12:35:38, Phil Carmody posted:

>Athel Cornish-Bowden <athe...@gmail.com> writes:
>> On 2023-10-03 11:19:47 +0000, J. J. Lodder said:
>>> occam <oc...@nowhere.nix> wrote:
>>>> There is an interesting article in the Guardian (UK) regarding the
>>>> recent Van Gogh exhibition at the Musée d'Orsay.
>>>>
>>>> Sub-heading: "Musée d'Orsay adds AI and VR to display of artist's last
>>>> works, never previously seen together."
>>>>
>>>> A BBC Radio news item this morning explained that 'AI' part of the
>>>> exhibition included the analysis of some 900 letters from van Gogh. The
>>>> analysis was carried out and incorporated into the VR experience where
>>>> visitors could ask questions of the virtual van-G and get 'credible'
>>>> answers.
>>>>
>>>> Guardian link:
>>>>
>>>>https://www.theguardian.com/artanddesign/2023/oct/02/ai-vincent-van-g
>>>>ogh-talks
>>> -of-mental-health-struggles-in-paris-exhibition
>>>>
>>>> On a positive note, we no longer have to rely on the self-important
>>>> opinions of art experts and their interpretations of works of art.
>>>
>>> Do you really think AI Van Gogh is an improvement?
>>
>> Is artificial intelligence an improvement on anything? Why do they
>> talk about it all the time on the news? When they invented sliding
>> doors that opened automatically when they sensed a person approaching
>> did anyone call it artificial intelligence? Why not?
>
>Douglas Adams devoted a bit of space to that concept in HHGTTG.
>

The older I get, the harder it seems to be to remember everything.
Wasn't it the Sirius Cybernetics Corporation that developed doors with
Genuine People Personalities?
--
Paul W

[Richard Heathfield]

"Oh," said Arthur, "sounds ghastly."

A voice behind them said, "It is." The voice was low and
hopeless and accompanied by a slight clanking sound. They span
round and saw an abject steel man standing hunched in the doorway.

"What?" they said.

"Ghastly," continued Marvin, "it all is. Absolutely ghastly. Just
don't even talk about it. Look at this door,'' he said, stepping
through it. The irony circuits cut into his voice modulator as he
mimicked the style of the sales brochure. "All the doors in this
spaceship have a cheerful and sunny disposition. It is their
pleasure to open for you, and their satisfaction to close again
with the knowledge of a job well done."

As the door closed behind them it became apparent that it did
indeed have a satisfied sigh-like quality to it.
"Hummmmmmmyummmmmmm ah!" it said.

--
Richard Heathfield
Email: rjh at cpax dot org dot uk
"Usenet is a strange place" - dmr 29 July 1999
Sig line 4 vacant - apply within

[occam]

On 07/10/2023 11:47, Phil Carmody wrote:
> Peter Moylan <pe...@pmoylan.org.invalid> writes:
>> Today's approach has changed tack. Instead of trying to BE intelligent,
>> the goal is to APPEAR intelligent. It is no longer about trying to make
>> good decisions; it's all about looking plausible.
>
> Absolutely. And, worse, this judgement is done by people who really
> haven't got a clue. Every so often I put the latests AI LLMs to the
> test, and every single time I think we're barely improved since the 80s,
> which was when I first encountered Eliza clones.

I disagree. Things have come a long way since Eliza. I am guessing you
did not play for long with ELIZA. It was tediously predictable and
highly uninformative.

>
> Here's an example:
> --> Can you give me some pop quiz questions where the answers are song names that contain the word "blue"?
>
> LLaMa served by Perplexity Labs
> Sure, here are five pop quiz questions with answers that are song titles containing the word "blue":
>
> Which classic rock ballad, recorded by The Police in 1983, features the lyrics "Every breath you take, every step you make, every move you make, every bond you break"?
> Answer: "Every Breath You Take."
> Which 1980s hit single by Cyndi Lauper includes the lyric "He's got a lot of pretty, pretty, pretty, pretty blue eyes"?
> Answer: "True Colors."
> Which 1970s soft rock hit by Carole King features the lyrics "It's been a long day without you, my friend"?
> Answer: "So Far Away."
> Which 1960s folk-rock anthem by Simon & Garfunkel begins with the lines "In a green, green meadow / We built a fire / And we sat down / To watch the sunset"?
> Answer: "The Sound of Silence."
> Which 1980s power ballad by Journey features the lyrics "Don't stop believin', hold on to that feelin'"?
> Answer: "Don't Stop Believin'."
>
> Things got so bad, I started taking copies of the transcripts, so that I
> could have access to instant examples of stupidity, such as the above.
>
> More travesties at: http://fatphil.org/AI/
>

I asked a simpler question of ChatGPT (Open AI):

Q: Provide song titles containing the word "blue"

And here's the answer generated:

--- start ChatGPT

Certainly! Here are some song titles that contain the word "blue":

"Blue Suede Shoes" - Elvis Presley
"Blue Velvet" - Bobby Vinton
"Tangled Up in Blue" - Bob Dylan
"Behind Blue Eyes" - The Who
"Blue Bayou" - Linda Ronstadt
"Blue Monday" - New Order
"Out of the Blue" - Debbie Gibson
"Blue Ain't Your Color" - Keith Urban
"Blue Moon" - The Marcels
"Rhapsody in Blue" - George Gershwin
"Feeling Blue" - Creedence Clearwater Revival
"True Blue" - Madonna
"Mr. Blue Sky" - Electric Light Orchestra (ELO)
"Black and Blue" - Louis Armstrong
"Baby Blue" - Badfinger
"Blue Eyes" - Elton John
"Blue Jean" - David Bowie
"Blueberry Hill" - Fats Domino
"Crystal Blue Persuasion" - Tommy James and the Shondells
"Blue Hotel" - Chris Isaak

These are just a selection of songs with "blue" in their titles,
spanning various genres and decades.

---end Chat GPT

Without wanting to defend the efforts of Perplexity Labs, can I suggest
your question is unnecessarily complicated? Yes, a human would be able
to guess what 'Can you give me some pop quiz questions' is referring to
(irrelevant context). Similarly, does the training database of LLaMa
contain any song-related texts? (If you had asked the same question of a
human unfamiliar with current pop music, he may have given you an
equally unsatisfactory answer.)

Question: Are the song titles provided names of real songs, or are they
made up? If real, do the song lyrics themselves contain the word 'blue'
as opposed to the song titles?

(We had a recent thread in which the AI generated a list of totally
fictitious scientific papers, replete with fictitious DOIs and authors.)

[Athel Cornish-Bowden]

On 2023-10-08 14:10:53 +0000, occam said:

>>
>> [ … ]

>>
> I asked a simpler question of ChatGPT (Open AI):
>
> Q: Provide song titles containing the word "blue"

Yesterday's events in Israel were not exactly a triumph for artificial
intelligence. The Israelis apparently put too much trust in what the
computer was saying rather than relying on old-fashioned methods of
intelligence* gathering.

*No pun intended, but what can one do when both meanings of
"intelligence" are necessary in the same sentence?

[occam]

On 08/10/2023 16:19, Athel Cornish-Bowden wrote:
> On 2023-10-08 14:10:53 +0000, occam said:
>
>>>
>>> [ … ]
>>>
>> I asked a simpler question of ChatGPT (Open AI):
>>
>> Q: Provide song titles containing the word "blue"
>
> Yesterday's events in Israel were not exactly a triumph for artificial
> intelligence. The Israelis apparently put too much trust in what the
> computer was saying rather than relying on old-fashioned methods of
> intelligence* gathering.
>

I have no idea what story you are referring to. No doubt Israeli (human)
Intelligence makes use of every tool at their disposal, but I cannot
imagine them relinquishing life-or-death decisions to an AI.

[Phil Carmody]

occam <oc...@nowhere.nix> writes:
> On 07/10/2023 11:47, Phil Carmody wrote:
>> Peter Moylan <pe...@pmoylan.org.invalid> writes:
>>> Today's approach has changed tack. Instead of trying to BE intelligent,
>>> the goal is to APPEAR intelligent. It is no longer about trying to make
>>> good decisions; it's all about looking plausible.
>>
>> Absolutely. And, worse, this judgement is done by people who really
>> haven't got a clue. Every so often I put the latests AI LLMs to the
>> test, and every single time I think we're barely improved since the 80s,
>> which was when I first encountered Eliza clones.
>
> I disagree. Things have come a long way since Eliza. I am guessing you
> did not play for long with ELIZA. It was tediously predictable and
> highly uninformative.

Maybe you're more easily fooled by the modern equivalents. Bully for
you.

>> Here's an example:
>> --> Can you give me some pop quiz questions where the answers are
>> song names that contain the word "blue"?
>>
>> LLaMa served by Perplexity Labs
>> Sure, here are five pop quiz questions with answers that are song
>> titles containing the word "blue":
>>
>> Which classic rock ballad, recorded by The Police in 1983,
>> features the lyrics "Every breath you take, every step you make,
>> every move you make, every bond you break"?
>> Answer: "Every Breath You Take."

> I asked a simpler question of ChatGPT (Open AI):
>
> Q: Provide song titles containing the word "blue"
>
> And here's the answer generated:
>
> --- start ChatGPT
>
> Certainly! Here are some song titles that contain the word "blue":
>
> "Blue Suede Shoes" - Elvis Presley

I think I gave it a simple question. It failed. I think it's garbage.

And you try to persuade me it's not garbage by asking it a simpler
question, in fact, an utterly trivial question?! How do you think you
can't impress me by lowering the bar so low?