Snidely wrote:

> I've read an author who felt that else clauses meant that you got the
> logic wrong, but I've definitely had places where I couldn't figure out
> a nice or tidy way to avoid an else. In Fortran, of course, you used
> gotos for that.

The craziest thing I've experienced was a consultant whose job it was to
introduce computers and it-technology to school teachers. I was ahead of
him at the time, but my collegues knew nothing about computers. He had
gotten the idea that the only logic you needed when programming was
"while" and "if". He completely ignored the other structures (for, case
=switch, repeat) that Comal offered. If he had kept it to himself, I
wouldn't have cared, but he taught my collegues according to that
principle. The result was meterlong programs that could have been
written in a few lines.

--
Bertel
Kolt, Denmark

Hibou wrote:

> Perhaps self-taught AI marks the divide, the point at which humans can
> no longer unravel its inner workings.

Isn't that the case with the Go program that beat the world champion? I
read that the decisive move was one that a human never would have
thought of.

--
Bertel
Kolt, Denmark

occam wrote:

> If the statement is not true, the 'else' is implicit. You continue down
> the list of instructions.

I have thought a bit about this statement because I don't understand it.

if x<10:
print(date)
else:
print('No go!')

versus:

if x<10:
print(date)
print('No go!')

Those programs are not identical.

--
Bertel
Kolt, Denmark

On 19/04/2024 11:59, Peter Moylan wrote:
> On 19/04/24 19:20, Snidely wrote:
>> on 4/19/2024, Athel Cornish-Bowden supposed :
>>> On 2024-04-18 17:03:19 +0000, occam said:
>>>
>>>> On 18/04/2024 14:51, Athel Cornish-Bowden wrote:
>>>>
>>>> <snip>
>>>>
>>>>>>>>
>>>>
>>>>>>>> 'Conditional statements'
>>>>>>>> (e.g. IF ...THEN ...ELSE instructions) have been the mainstay of
>>>>>>>> programming languages forever.
>>>>>
>>>>> Not forever. When I learned Fortran IV in the 1960s it had
>>>>> arithmetic if
>>>>> and logical if, but neither corresponded to if...then...else.
>>>>
>>>>
>>>> Emm...actually they are.  Both logical and arithmetic 'if' statements
>>>> are of the form 'if...then...else'
>>>>
>>>> If the condition is true, 'then' follows (e.g. GO to 50)
>>>>
>>>> If the statement is not true, the 'else' is implicit. You continue down
>>>> the  list of instructions.
>>>
>>> Yes, if you go through some mental contortions you can interpret them
>>> that way, but I'd be surprised if all early users of Fortran IV did.
>>>
>>> Leaving the "else" implicit is a great way to confuse people.
>>
>> I've read an author who felt that else clauses meant that you got the
>> logic wrong, but I've definitely had places where I couldn't figure out
>> a nice or tidy way to avoid an else.  In Fortran, of course, you used
>> gotos for that.
>
> Unless you had a compiler that implemented COMEFROM.
>

[ASIDE] At the time I was interested in structured programming (a.k.a.
'NO to GOTO' programming) there was a very famous & accomplished
professor called E. GOTO.

https://en.wikipedia.org/wiki/Eiichi_Goto

That would be like having a professor of AI these days called Prof.
Truly Daft.

Bertel Lund Hansen hat am 19.04.2024 um 16:47 geschrieben:
> Hibou wrote:
>
>> Perhaps self-taught AI marks the divide, the point at which humans can
>> no longer unravel its inner workings.
>
> Isn't that the case with the Go program that beat the world champion? I
> read that the decisive move was one that a human never would have
> thought of.

Actually, there was no single decisive move. If I remember correctly,
the world champion lost a five-game series 4 to 1. Also, that program
(AlphaGo) did not revolutionise Go theory completely, but it played
quite a few moves* looking very strange to humans - and I do not mean
lousy amateurs like me, I mean humans who earn their living as
professional Go players.

BTW, if I remember correctly, some time later Alpha Go lost a series
against a more advanced program 100 to 0.

* As a learner of English, I want to express the concept "less than
several, but more than just two or three". Let's say, between five and
ten. Better suggestions are highly welcome.

I also thank Jerry for his "That should be "What else do you call it?".
But then I have to ask: why is "what else" correct and "how else" wrong
in questions like my "... else do you call it?" Just a matter of usage,
which I have to learn, or is there an explanation I can understand?

Silvano wrote:

> Bertel Lund Hansen hat am 19.04.2024 um 16:47 geschrieben:
>> Hibou wrote:
>>
>>> Perhaps self-taught AI marks the divide, the point at which humans can
>>> no longer unravel its inner workings.
>>
>> Isn't that the case with the Go program that beat the world champion? I
>> read that the decisive move was one that a human never would have
>> thought of.

> Actually, there was no single decisive move. If I remember correctly,
> the world champion lost a five-game series 4 to 1. Also, that program
> (AlphaGo) did not revolutionise Go theory completely, but it played
> quite a few moves* looking very strange to humans - and I do not mean
> lousy amateurs like me, I mean humans who earn their living as
> professional Go players.

> BTW, if I remember correctly, some time later Alpha Go lost a series
> against a more advanced program 100 to 0.

> * As a learner of English, I want to express the concept "less than
> several, but more than just two or three". Let's say, between five and
> ten. Better suggestions are highly welcome.

"Several" is fine there, in my opinion.

> I also thank Jerry for his "That should be "What else do you call it?".
> But then I have to ask: why is "what else" correct and "how else" wrong
> in questions like my "... else do you call it?" Just a matter of usage,
> which I have to learn, or is there an explanation I can understand?

It seems straightforward to me. You expect the answer to be a noun--
"I call it a backwards apostrophe"--so the question starts with "What".
What I find strange, but not hard to learn, is the use of words
corresponding to "How" in French and Spanish and, Google thinks,
Italian.

"How do you call" for "What do you call" is a very common error by
non-native English speakers. However, it can be used with this sense:

"How do you call your cat?"

"I say 'Heeeeeeere kittykittykittykittykitty.'"

--
Jerry Friedman

On 19/04/2024 16:50, Bertel Lund Hansen wrote:
> occam wrote:
>
>> If the statement is not true, the 'else' is implicit. You continue down
>> the list of instructions.
>
> I have thought a bit about this statement because I don't understand it.
>
> if x<10:
> print(date)
> else:
> print('No go!')
>
> versus:
>
> if x<10:
> print(date)
> print('No go!')
>
> Those programs are not identical.
>

...and you are a terrible programmer. You don't know when to HALT your
program.

occam wrote:

>> Those programs are not identical.
>>
>
> ..and you are a terrible programmer. You don't know when to HALT your
> program.

Please explain.

--
Bertel
Kolt, Denmark

On 19-Apr-24 8:15, Bertel Lund Hansen wrote:
> Peter Moylan wrote:
>
>> Like a former guitar teacher of mine, who insisted that rock and roll
>> was invented in the 1980s.
>
> I think that we should commemorate Rosetta Tharpe with a song from 1936,
> "This Train":
>
> https://www.youtube.com/watch?v=jOrhjgt-_Qc

I'm sure there are lots of technical reasons why that's rock & roll, but
I still hear it as (gospel) blues.

--
Sam Plusnet

On 2024-04-19 02:10, Athel Cornish-Bowden wrote:
> On 2024-04-18 17:03:19 +0000, occam said:
>
>> On 18/04/2024 14:51, Athel Cornish-Bowden wrote:
>>
>> <snip>
>>
>>>>>>
>>
>>>>>> 'Conditional statements'
>>>>>> (e.g. IF ...THEN ...ELSE instructions) have been the mainstay of
>>>>>> programming languages forever.
>>>
>>> Not forever. When I learned Fortran IV in the 1960s it had arithmetic if
>>> and logical if, but neither corresponded to if...then...else.
>>
>>
>> Emm...actually they are.  Both logical and arithmetic 'if' statements
>> are of the form 'if...then...else'
>>
>> If the condition is true, 'then' follows (e.g. GO to 50)
>>
>> If the statement is not true, the 'else' is implicit. You continue down
>> the  list of instructions.
>
> Yes, if you go through some mental contortions you can interpret them
> that way, but I'd be surprised if all early users of Fortran IV did.
>
> Leaving the "else" implicit is a great way to confuse people.

I don't understand the confusion.

An if without an else just says:

If whatever <do something>
or don't

--
This sentence no verb.

On 2024-04-19 13:39, occam wrote:
> On 19/04/2024 16:50, Bertel Lund Hansen wrote:
>> occam wrote:
>>
>>> If the statement is not true, the 'else' is implicit. You continue down
>>> the list of instructions.
>>
>> I have thought a bit about this statement because I don't understand it.
>>
>> if x<10:
>> print(date)
>> else:
>> print('No go!')
>>
>> versus:
>>
>> if x<10:
>> print(date)
>> print('No go!')
>>
>> Those programs are not identical.
>>
>
> ..and you are a terrible programmer. You don't know when to HALT your
> program.

Since we do not see what comes after "print('No go!')", we don't know if
the programmer wants the program to halt or not.

--
I before E... except when you run a feisty heist on a weird foreign
neighbour.

Sam Plusnet wrote:

>> https://www.youtube.com/watch?v=jOrhjgt-_Qc
>
> I'm sure there are lots of technical reasons why that's rock & roll, but
> I still hear it as (gospel) blues.

This is more clearly rock&roll, but it's later (1964):

https://www.youtube.com/watch?v=Y9a49oFalZE

--
Bertel
Kolt, Denmark

On 2024-04-19 06:56, Hibou wrote:
> Le 18/04/2024 à 06:47, Rich Ulrich a écrit :
>>
>> OBaue: What should we call examples of AI when the
>> achievements are too tiny to for the intelligence to baffle us?
>
> AM, artificial mediocrity, perhaps.
>
> I think bafflement may come at different points. We understand what a
> self-driving car is doing when it stays in lane or stops at lights,
> while likely being unable to understand how it does it.
>
> Then, at a higher level, if AI managed to stabilise all the world's
> economies or prevent disputes escalating into war, everything it does
> might be beyond our comprehension.
>
> Perhaps self-taught AI marks the divide, the point at which humans can
> no longer unravel its inner workings.

Instead of specifying only 'self-taught, I would include human-taught,
with the caveat that human-taught leaves the way open to abuse or stupidity.

I have two apps on my phone that I use fairly often. One is called
'LeafSnap' and one is called 'BirdNet'.

LeafSnap analyzes images, allowing me to specify foliage, flowers,
bark. It is pretty good, but it asks for feedback to tell it if it's
right or not.

BirdNet analyzes bird songs, with a similar feedback mechanism, which I
have used a few times when I knew the analysis was incorrect.

--
We are all born ignorant, but one must work hard to remain stupid.
—Anonymous

On 2024-04-19 06:56, Hibou wrote:
> Le 18/04/2024 à 06:47, Rich Ulrich a écrit :
>>
>> OBaue: What should we call examples of AI when the
>> achievements are too tiny to for the intelligence to baffle us?
>
> AM, artificial mediocrity, perhaps.
>
> I think bafflement may come at different points. We understand what a
> self-driving car is doing when it stays in lane or stops at lights,
> while likely being unable to understand how it does it.
>
> Then, at a higher level, if AI managed to stabilise all the world's
> economies or prevent disputes escalating into war, everything it does
> might be beyond our comprehension.
>
> Perhaps self-taught AI marks the divide, the point at which humans can
> no longer unravel its inner workings.
>

Instead of specifying only 'self-taught, I would include human-taught,
with the caveat that human-taught leaves the way open to abuse or stupidity.

I have two apps on my phone that I use fairly often. One is called
'LeafSnap' and one is called 'BirdNet'.

LeafSnap analyzes images, allowing me to specify foliage, flowers,
bark. It is pretty good, but it asks for feedback to tell it if it's
right or not.

BirdNet analyzes bird songs, with a similar feedback mechanism, which I
have used a few times when I knew the analysis was incorrect.

--
We are all born ignorant, but one must work hard to remain stupid.
—Anonymous
--
We are all born ignorant, but one must work hard to remain stupid.
—Anonymous

On 20/04/24 00:50, Bertel Lund Hansen wrote:
> occam wrote:
>
>> If the statement is not true, the 'else' is implicit. You continue down
>> the list of instructions.
>
> I have thought a bit about this statement because I don't understand it.
>
> if x<10:
> print(date)
> else:
> print('No go!')
>
> versus:
>
> if x<10:
> print(date)
> print('No go!')
>
> Those programs are not identical.

The "implicit else" referred to early Fortran, where a logical IF
statement had the form

IF (something) GOTO somewhere else.

In that case the "else" part came straight after the IF line.

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

On 20/04/24 00:47, Bertel Lund Hansen wrote:
> Hibou wrote:
>
>> Perhaps self-taught AI marks the divide, the point at which humans
>> can no longer unravel its inner workings.
>
> Isn't that the case with the Go program that beat the world champion?
> I read that the decisive move was one that a human never would have
> thought of.

The example that sticks in my mind was a program designed to prove
theorems in mathematics. It proved that the base angles in an isosceles
triangle are equal because ABC is congruent to ACB. That's a proof that
probably would not have occurred to humans.

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

lar3ryca <larry@invalid.ca> wrote:

> On 2024-04-19 02:10, Athel Cornish-Bowden wrote:
> > On 2024-04-18 17:03:19 +0000, occam said:
> >
> >> On 18/04/2024 14:51, Athel Cornish-Bowden wrote:
> >>
> >> <snip>
> >>
> >>>>>>
> >>
> >>>>>> 'Conditional statements'
> >>>>>> (e.g. IF ...THEN ...ELSE instructions) have been the mainstay of
> >>>>>> programming languages forever.
> >>>
> >>> Not forever. When I learned Fortran IV in the 1960s it had arithmetic if
> >>> and logical if, but neither corresponded to if...then...else.
> >>
> >>
> >> Emm...actually they are. Both logical and arithmetic 'if' statements
> >> are of the form 'if...then...else'
> >>
> >> If the condition is true, 'then' follows (e.g. GO to 50)
> >>
> >> If the statement is not true, the 'else' is implicit. You continue down
> >> the list of instructions.
> >
> > Yes, if you go through some mental contortions you can interpret them
> > that way, but I'd be surprised if all early users of Fortran IV did.
> >
> > Leaving the "else" implicit is a great way to confuse people.
>
> I don't understand the confusion.
>
> An if without an else just says:
>
> If whatever <do something>
> or don't
>

Exactly so. Programs are built on the hardware and the 2nd. generation
mainframes I used to fettle in the late '60s certainly did "if, then,
else" just like that. I used to write my own hardware test routines in
m/c code using that instruction.

Sometimes programmers would ask us hardware people "can the m/c do this
and how?" We tried not to giggle in front of them...

--
^Ï^. Sn!pe, PA, FIBS - Professional Crastinator

My pet rock Gordon admires J. Alfred Prufrock.

Bertel Lund Hansen <gadekryds@lundhansen.dk> wrote:

> Sam Plusnet wrote:
>
> >> https://www.youtube.com/watch?v=jOrhjgt-_Qc
> >
> > I'm sure there are lots of technical reasons why that's rock & roll, but
> > I still hear it as (gospel) blues.
> >
>
> This is more clearly rock&roll, but it's later (1964):
>
> https://www.youtube.com/watch?v=Y9a49oFalZE
>

Bill Haley & his Comets (founded 1952) - 'Rock Around The Clock' 1955:-

<https://youtu.be/ZgdufzXvjqw>

--
^Ï^. Sn!pe, PA, FIBS - Professional Crastinator

My pet rock Gordon admires J. Alfred Prufrock.

jerryfriedman <jerry.friedman99@gmail.com> wrote:

[...]

> > * As a learner of English, I want to express the concept "less than
> > several, but more than just two or three". Let's say, between five and
> > ten. Better suggestions are highly welcome.
> >
>
> "Several" is fine there, in my opinion.
>

I'd say "a few" but "several" is pretty close to what you ask.

--
^Ï^. Sn!pe, PA, FIBS - Professional Crastinator

My pet rock Gordon admires J. Alfred Prufrock.

On Fri, 19 Apr 2024 09:32:33 +0200, occam <occam@nowhere.nix> wrote:

>On 19/04/2024 00:27, Rich Ulrich wrote:
>> On Thu, 18 Apr 2024 08:58:36 +0200, occam <occam@nowhere.nix> wrote:
>>
>>> On 18/04/2024 07:47, Rich Ulrich wrote:
>
><snip>
>
>>>>
>>>> Wouldn't it be nice if AI would fill in my credit card number
>>>> for me when I order something?
>>>
>>> You don't need AI for that. Any more than you need AI to open an
>>> automatic door when you approach it.
>>
>> Well, you do need a lookup list. And there is some recognition
>> going on, to know the context, that a program WANTs a credit card.
>
>The point of AI is not just the recognition (which is only a part) but
>also adaptability ('learning') and adjusting to new situations.
>
>>
>>>
>>>> along with my address, etc.?
>>>> Yeah, Amazon knows that because it reads my cookies and
>>>> knows who I am. We don't call it AI.
>>>>
>>>> Amazon has a further assisting function. Once I'm at the
>>>> site, using its search bar, typing two or three letters pops
>>>> up the suggestion of what I'm about to look for. AI sitting
>>>> here with me, sort of, but we don't call it AI.
>>>
>>> Why not? It is using text-based prediction in some language context? It
>>> is therefore making use of AI.
>>
>> I'm defending lookup lists. This looks like another use of lookup,
>> but this time you accept it.
>
>It *is* a look-up table, but it may well be a slightly *different*
>lookup table than the last time you typed in your search term. Why?
>Because the AI has modified the look up table (since your last effort)
>based on your choice last time around. A fixed 'look up table' is
>different to an adaptive lookup table, which adjusts itself every time
>new data goes into its database.

The lookup structure for the user is similar, whether the table
is adaptive or fixed.

'Adaptation' is what seems to account for the tendency of chatbots
to drift into racism, etc. That is a complication, for several sorts
of concerns.

>
>>
>>>
><snip>
>
>
>>
>> Translation of natural language seems to be accepted as AI.
>
>Yes.
>
>>
>> But my impression is that the BEST translation of idiomatic
>> expressions or sayings is done by lookup, enabled by huge
>> amounts of cheap storage and fast, parallel, processors.
>> "The spirit is willing but the flesh is weak" should be found
>> in its entirety.
>
>Yes. (No contradiction here with your statement above.)
>
>>
>> Translation in context, I imagine, is done by scoring other
>> words found in what is being observed. And using huge sets
>> of data for comparison.
>
>Now ask yourself, will the 'scoring' of the other words be the same next
>time you use the AI? In a truly adaptive system, the answer is NO. With
>every addition of new data into the 'training' data set, the scoring
>with change, albeit slightly.
>
>It is important to not get stuck on the term 'look-up table'. You have
>to ask: how often is this table revised/reviewed? And by whom? If it is
>done automatically by the computer, then that is 'AI'.

Thanks for the comments. That is an interesting track.

Now I start thinking of natural, human intelligence. Are we
reviewing our own lookup tables?

--
Rich Ulrich

Silly possum!

Peter Moylan wrote:

> The "implicit else" referred to early Fortran, where a logical IF
> statement had the form
>
> IF (something) GOTO somewhere else.
>
> In that case the "else" part came straight after the IF line.

Thanks. That's just like assembler.

--
Bertel
Kolt, Denmark

Rich Ulrich wrote:

>>It is important to not get stuck on the term 'look-up table'. You have
>>to ask: how often is this table revised/reviewed? And by whom? If it is
>>done automatically by the computer, then that is 'AI'.
>
> Thanks for the comments. That is an interesting track.
>
> Now I start thinking of natural, human intelligence. Are we
> reviewing our own lookup tables?

Constantly. That's the way babies learn to speak - and people who
willingly or by force stay in a foreign country without having learned
the language.

--
Bertel
Kolt, Denmark

On 19/04/2024 23:24, lar3ryca wrote:
> On 2024-04-19 13:39, occam wrote:
>> On 19/04/2024 16:50, Bertel Lund Hansen wrote:
>>> occam wrote:
>>>
>>>> If the statement is not true, the 'else' is implicit. You continue down
>>>> the  list of instructions.
>>>
>>> I have thought a bit about this statement because I don't understand it.
>>>
>>>        if x<10:
>>>            print(date)
>>>        else:
>>>            print('No go!')
>>>
>>> versus:
>>>
>>>        if x<10:
>>>            print(date)
>>>        print('No go!')
>>>
>>> Those programs are not identical.
>>>
>>
>> ..and you are a terrible programmer. You don't know when to HALT your
>> program.
>
> Since we do not see what comes after "print('No go!')", we don't know if
> the programmer wants the program to halt or not.
>

Its a simple example. I assume it is intended to print <whatever>, THEN
stop (or jump to the next part of the program).

On 2024-04-20 01:44:12 +0000, Sn!pe said:

> jerryfriedman <jerry.friedman99@gmail.com> wrote:
>
> [...]
>
>>> * As a learner of English, I want to express the concept "less than
>>> several, but more than just two or three". Let's say, between five and
>>> ten. Better suggestions are highly welcome.
>>>
>>
>> "Several" is fine there, in my opinion.
>>
>
> I'd say "a few" but "several" is pretty close to what you ask.

I'd go for "several".

--
Athel -- French and British, living in Marseilles for 37 years; mainly
in England until 1987.