XPost: nl.politiek, nl.wetenschap, nl.religie
XPost: nl.gezondheid.psychiatrie
On 4/21/2025 12:19 AM, Pancho Sanza wrote:
Jos Bergervoet <[email protected]> wrote:
Historical Oversimplification: The claim that Gödel "torpedoed" Hilbert's >>> program is an exaggeration, as mathematics continued to thrive post-Gödel. >>
Het feit dat Grok zelf vrijwel niets van wiskunde snapt (wat hij zelf
ook altijd toegeeft) maakt zijn commentaar hier weinig relevant,
PS: Dat hij het zelf altijd toegeeft is wel opmerkelijk (zie voorbeeld
aan het eind).
dat hij mijn woorden opzettelijk verdraait is erger! Wat ik letterlijk
schreef was "volledige axiomatisatie van de gehele wiskunde", en dat
aka "Hilbert's program".
In je eigen Wikipedialink:
<
https://en.wikipedia.org/wiki/Hilbert%27s_program> lezen we
"Hilbert proposed to ground all existing theories to a finite,
complete set of axioms,"
en drie zinnen verder:
"Gödel showed that any consistent system with a computable set of
axioms which is capable of expressing arithmetic can never be complete"
...
droombeeld van Hilbert viel dus inderdaad volkomen in het water.
Niet "volkomen",
Wiskundig gezien wel. Hilbert wilde dat het "complete" was, en Gödel
bewees dat het niet "complete" kon zijn. Hilbert wilde dat iets "waar"
was waarvan Gödel bewees dat het "niet waar" was. Erger kan niet, voor
een wiskundige. Het is dus "torpedoed!" Grok moet dat niet ontkennen.
als ik de volgende wikipedia passage mag geloven:
Waarom zou je? Met het bovenstaande is de zaak beslecht. Maar goed:
|Many current lines of research in mathematical logic, such as proof theory |and reverse mathematics, can be viewed as natural continuations of |Hilbert's original program. Much of it can be salvaged by changing its |goals slightly (Zach 2005), and with the following modifications some of it |was successfully completed: [enz]
"Changing its goals slightly" is niet toegestaan als het om wiskundige
bewijzen gaat. Anders kun je ook bewijzen dat 2+2 gelijk aan 5 is.
...
Wil je bij dit soort grove fouten je AI vriendjes voortaan zelf even
terechtwijzen? Je mag ze gerust een beetje helpen, dat wekt uiteindelijk
een veel betere indruk in je postjes!
Ik zie hier niet zozeer een grove fout alswel een verschil in interpre- tatie. En/of een hobbeltje in de Engelse vertaling.
Grok negeert dat in de wiskunde de regels meer rigide zijn dan elders.
Ironisch eigenlijk. In vroegere science fiction stelde men de eerste
AI's altijd voor als strikt logisch redenerende nerds! Maar nu blijkt
dat ze juist uitblinken in vaag geklets. (Achteraf gezien begrijpelijk,
vaag geklets is wat hen bij uitstek de menselijke trekjes geeft.)
Maar goed, genoeg spijkers op laag water gevonden voor vandaag.
Wacht nou, nog even dat voorbeeld, van een week geleden ongeveer. Kijk
hoe poeslief de AI zijn fouten steeds toegeeft! (Geloof maar niet dat
dat zo zal blijven..)
=======================================================================
| Jos: For which rational numbers q does the polynomial x^3+3x-2q have
| a rational root?
Grok: To determine for which rational numbers ( q ) the cubic polynomial
x^3 + 3x - 2q = 0
has a rational root, we use the Rational Root Theorem. The polynomial is: p(x) = x^3 + 3x - 2q,
with coefficients in the rationals, ...
| Hier maakt Grok een fout, de coefficienten moeten integers zijn om
| het rational root theorem te mogen gebruiken. En dus trekt Grok
| vervolgens een foute conclusie:
... , so the root must be an integer \frac{p}{1} = p
. Thus, any rational root must be an integer.
| Na een aantal alinea's gedraai komt Grok dan (noodgedwongen) met een
| geheel andere eindconclusie, Dus ik vraag dan:
| Jos: In your last answer the statement in step 1 that "any rational
| root must be an integer" is in conflict with your final answer.
Grok: You're absolutely right, and I apologize for the oversight in Step
1 of my previous response. The statement "any rational root must be an
integer" is incorrect and inconsistent with the final answer, which
correctly accounts for both integer and non-integer rational roots. Let
me clarify and provide a corrected and streamlined answer to your
question: For which rational numbers ( q ) does the polynomial x^3+3x−2q
have a rational root?
[ ... ]
| Zelfs al wil je mijn gelijk hier betwijfelen (wat ik betwijfel), het
| is toch frappant hoe ruiterlijk Grok toegeeft dat hij ongelijk heeft.
| En het gaat verder want Grok maakt dan opnieuw dezelfde fout, dus
| ik leg het uit...
| Jos: In your last two answers you are incorrectly using the Rational
| Root Theorem. It only applies to polynomials with integer
| coefficients, so since the constant term is -2a/b you first have to
| multiply the whole expression by b. Can you correct your answer?
Grok: Thank you for pointing out the error in my application of the
Rational Root Theorem. You’re absolutely correct: the Rational Root
Theorem applies to polynomials with integer coefficients, and since the constant term in x3+3x−2q=0 is −2q
, where ( q ) is rational (e.g., q = \frac{a}{b}
), the polynomial must be adjusted to have integer coefficients before
applying the theorem. Let’s correct this and provide a clear, accurate
answer to the question: For which rational numbers ( q ) does the
polynomial x^3 + 3x - 2q have a rational root?
...
=======================================================================
Kortom, een vriendelijke conversatie en een AI die bereidwillig zijn
fouten toegeeft. We do not have to worry...
--
Jos
--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)