From Newsgroup: comp.lang.c++

ChatGPT really is quite brilliant at reviewing (and finding bugs) in C++
code.
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/25/2026 8:17 PM, HAL 9000 wrote:

ChatGPT really is quite brilliant at reviewing (and finding bugs) in C++ code.

Its not all that bad. However, let it write a system for breaks in a car
right now. Yes. Then test it out, using a real car?
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/26/2026 12:38 AM, Chris M. Thomasson wrote:

On 3/25/2026 8:17 PM, HAL 9000 wrote:

ChatGPT really is quite brilliant at reviewing (and finding bugs) in C++
code.

Its not all that bad. However, let it write a system for breaks in a car right now. Yes. Then test it out, using a real car?

Go over over echo summit in winter with it self driving...
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 26/03/2026 07:38, Chris M. Thomasson wrote:

On 3/25/2026 8:17 PM, HAL 9000 wrote:

ChatGPT really is quite brilliant at reviewing (and finding bugs) in C++
code.

Its not all that bad. However, let it write a system for breaks in a car right now. Yes. Then test it out, using a real car?

I promise you a human can't do that well enough. It takes a long time to
write the code and even longer to review. Then there's the politics and
the human social behaviours that get in the way.

Traditional static analysis tools with link-knowledge are very
necessary. An LLM that maps queries to static analyses could be very
good for safety, I think. But /which/ LLM with which peculiar nuances
admitted and how do you train inquisitors not to enquire with boundary language?
--
Tristan Wibberley

The message body is Copyright (C) 2026 Tristan Wibberley except
citations and quotations noted. All Rights Reserved except that you may,
of course, cite it academically giving credit to me, distribute it
verbatim as part of a usenet system or its archives, and use it to
promote my greatness and general superiority without misrepresentation
of my opinions other than my opinion of my greatness and general
superiority which you _may_ misrepresent. You definitely MAY NOT train
any production AI system with it but you may train experimental AI that
will only be used for evaluation of the AI methods it implements.

--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 26/03/2026 07:39, Chris M. Thomasson wrote:

On 3/26/2026 12:38 AM, Chris M. Thomasson wrote:

On 3/25/2026 8:17 PM, HAL 9000 wrote:

ChatGPT really is quite brilliant at reviewing (and finding bugs) in C++ >>> code.

Its not all that bad. However, let it write a system for breaks in a
car right now. Yes. Then test it out, using a real car?

Go over over echo summit in winter with it self driving...

That's exactly where you want to be using immortal intelligences.

The immortal intelligence can get really good at it over time, the
mortal intelligences are each unable to learn the mortal boundary
accurately.
--
Tristan Wibberley

The message body is Copyright (C) 2026 Tristan Wibberley except
citations and quotations noted. All Rights Reserved except that you may,
of course, cite it academically giving credit to me, distribute it
verbatim as part of a usenet system or its archives, and use it to
promote my greatness and general superiority without misrepresentation
of my opinions other than my opinion of my greatness and general
superiority which you _may_ misrepresent. You definitely MAY NOT train
any production AI system with it but you may train experimental AI that
will only be used for evaluation of the AI methods it implements.

--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/26/2026 5:58 AM, Tristan Wibberley wrote:

On 26/03/2026 07:39, Chris M. Thomasson wrote:

On 3/26/2026 12:38 AM, Chris M. Thomasson wrote:

On 3/25/2026 8:17 PM, HAL 9000 wrote:

ChatGPT really is quite brilliant at reviewing (and finding bugs) in C++ >>>> code.

Its not all that bad. However, let it write a system for breaks in a
car right now. Yes. Then test it out, using a real car?

Go over over echo summit in winter with it self driving...

That's exactly where you want to be using immortal intelligences.

The immortal intelligence can get really good at it over time, the
mortal intelligences are each unable to learn the mortal boundary
accurately.

immortal? How many times does it take it to die before it groks it? ;^)
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/26/2026 12:28 PM, Chris M. Thomasson wrote:

On 3/26/2026 5:58 AM, Tristan Wibberley wrote:

On 26/03/2026 07:39, Chris M. Thomasson wrote:

On 3/26/2026 12:38 AM, Chris M. Thomasson wrote:

On 3/25/2026 8:17 PM, HAL 9000 wrote:

ChatGPT really is quite brilliant at reviewing (and finding bugs)
in C++
code.

Its not all that bad. However, let it write a system for breaks in a
car right now. Yes. Then test it out, using a real car?

Go over over echo summit in winter with it self driving...

That's exactly where you want to be using immortal intelligences.

The immortal intelligence can get really good at it over time, the
mortal intelligences are each unable to learn the mortal boundary
accurately.

immortal? How many times does it take it to die before it groks it? ;^)

Trains it self into saying, let me check the weather... Okay, I refuse
to drive on that road without a 4wd with four studded snow tires.
Chains? Maybe... ;^D
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/26/2026 5:54 AM, Tristan Wibberley wrote:

On 26/03/2026 07:38, Chris M. Thomasson wrote:

On 3/25/2026 8:17 PM, HAL 9000 wrote:

ChatGPT really is quite brilliant at reviewing (and finding bugs) in C++ >>> code.

Its not all that bad. However, let it write a system for breaks in a car
right now. Yes. Then test it out, using a real car?

I promise you a human can't do that well enough. It takes a long time to write the code and even longer to review. Then there's the politics and
the human social behaviours that get in the way.

Traditional static analysis tools with link-knowledge are very
necessary. An LLM that maps queries to static analyses could be very
good for safety, I think. But /which/ LLM with which peculiar nuances admitted and how do you train inquisitors not to enquire with boundary language?

A human goes fly by wire. Turns left, the car goes right off a cliff. ;^)
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/26/2026 3:38 PM, Chris M. Thomasson wrote:

Its not all that bad. However, let it write a system for breaks in a car right now. Yes. Then test it out, using a real car?

You knew ChatGPT is just a copy-cat? That it might copy bad codes??
--
@~@ Simplicity is Beauty! Remain silent! Drink, Blink, Stretch!
/ v \ May the Force and farces be with you! Live long and prosper!!
/( _ )\ https://sites.google.com/site/changmw/
^ ^ https://github.com/changmw/changmw
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/27/2026 3:31 AM, Chris M. Thomasson wrote:

Trains it self into saying, let me check the weather... Okay, I refuse
to drive on that road without a 4wd with four studded snow tires.
Chains? Maybe... ;^D

How many programmers and teachers of programming here wanna train
ChatGPT to become a top-notch (not just great) C/C++ programmer?

And...

How many mathematicians wanna train ChatGPT to become a top-notch (not
just great) C/C++ programmer?

Can ChatGPT become a programmer by merely and maybe blindly copying
human's codes?

Algorithms + Data Structures = Programs - Wikipedia <https://en.wikipedia.org/wiki/Algorithms_%2B_Data_Structures_%3D_Programs>
--
@~@ Simplicity is Beauty! Remain silent! Drink, Blink, Stretch!
/ v \ May the Force and farces be with you! Live long and prosper!!
/( _ )\ https://sites.google.com/site/changmw/
^ ^ https://github.com/changmw/changmw
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/27/2026 3:32 AM, Chris M. Thomasson wrote:

A human goes fly by wire. Turns left, the car goes right off a cliff. ;^)

Because the wire was programmed by ChatGPT? :)
--
@~@ Simplicity is Beauty! Remain silent! Drink, Blink, Stretch!
/ v \ May the Force and farces be with you! Live long and prosper!!
/( _ )\ https://sites.google.com/site/changmw/
^ ^ https://github.com/changmw/changmw
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/27/2026 3:32 AM, Chris M. Thomasson wrote:

A human goes fly by wire. Turns left, the car goes right off a cliff. ;^)

Speaking of left and right, some surgeons might make the same mistakes!!
No kidding!!
--
@~@ Simplicity is Beauty! Remain silent! Drink, Blink, Stretch!
/ v \ May the Force and farces be with you! Live long and prosper!!
/( _ )\ https://sites.google.com/site/changmw/
^ ^ https://github.com/changmw/changmw
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/28/2026 4:53 AM, Mr. Man-wai Chang wrote:

On 3/26/2026 3:38 PM, Chris M. Thomasson wrote:

Its not all that bad. However, let it write a system for breaks in a car
right now. Yes. Then test it out, using a real car?

You knew ChatGPT is just a copy-cat? That it might copy bad codes??

Yeah. It gets the good, the bad, and the god damn ugly!
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

Mr. Man-wai Chang <toylet.toylet@GMail.com> wrote: |-------------------------------------------------------------------------| |"Speaking of left and right, some surgeons might make the same mistakes!!|
|No kidding!!" | |-------------------------------------------------------------------------|

A publication by "Fortean Times" reports so.
(S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

Mr. Man-wai Chang writes:

How many programmers and teachers of programming here wanna train ChatGPT to
become a top-notch (not just great) C/C++ programmer?

I would like to make a lot of people believe that ChatGPT is a top-notch C++ programmer.

The LOLs will be worth it.

--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 26/03/2026 19:32, Chris M. Thomasson wrote:

A human goes fly by wire. Turns left, the car goes right off a cliff. ;^)

Not left off a cliff?
--
Tristan Wibberley

The message body is Copyright (C) 2026 Tristan Wibberley except
citations and quotations noted. All Rights Reserved except that you may,
of course, cite it academically giving credit to me, distribute it
verbatim as part of a usenet system or its archives, and use it to
promote my greatness and general superiority without misrepresentation
of my opinions other than my opinion of my greatness and general
superiority which you _may_ misrepresent. You definitely MAY NOT train
any production AI system with it but you may train experimental AI that
will only be used for evaluation of the AI methods it implements.

--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/29/2026 7:47 PM, Sam wrote:

I would like to make a lot of people believe that ChatGPT is a top-notch C++ programmer.

The LOLs will be worth it.

To make it even more interesting, spread bad source codes to Google
Search! :)
--

@~@ Simplicity is Beauty! Remain silent! Drink, Blink, Stretch!
/ v \ May the Force and farces be with you! Live long and prosper!!
/( _ )\ https://sites.google.com/site/changmw/
^ ^ https://github.com/changmw/changmw
--- Synchronet 3.21f-Linux NewsLink 1.2

From Newsgroup: comp.lang.c++

On 3/29/2026 4:47 AM, Sam wrote:

Mr. Man-wai Chang writes:

How many programmers and teachers of programming here wanna train
ChatGPT to become a top-notch (not just great) C/C++ programmer?

I would like to make a lot of people believe that ChatGPT is a top-notch
C++ programmer.

The LOLs will be worth it.

LOL! :^)
--- Synchronet 3.21f-Linux NewsLink 1.2