The end of stackoverflow?

Post by Jan Panteltje
Stack Overflow users sabotage their posts after OpenAI deal
https://arstechnica.com/information-technology/2024/05/stack-overflow-users-sabotage-their-posts-after-openai-deal/
The end of stack overflow?
Personally I know companies are using my open sourced stuff..
Let it be...
I like Stackoverflow, it gave me many good answers in the past
to difficult questions...
Many highly qualified people there.

One often has to trawl through a number of suggested solutions, either
because most of them are wrong (or at least wildly apocryphal),
irrelevant, or because the same or similar symptoms can have many
different underlying causes.

I have to wonder whether a language model is really up to the task of
filtering out the dross, while keeping the important parts.

Sylvia.

Don Y

2024-05-10 08:29:18 UTC

One often has to trawl through a number of suggested solutions, either because
most of them are wrong (or at least wildly apocryphal), irrelevant, or because
the same or similar symptoms can have many different underlying causes.
I have to wonder whether a language model is really up to the task of filtering
out the dross, while keeping the important parts.

Patterns (repeated) in answers are reinforced. So, outliers tend to
not influence the model, as much.

E.g., Carlin (?) did a bit in which he uttered something like, "Here's a
sentence no one has ever said before..." You, thus, wouldn't expect an
AI to come up with such a sentence in "normal use" because its weights
are so low.

Jan Panteltje

2024-05-10 08:52:34 UTC

On a sunny day (Fri, 10 May 2024 15:16:21 +0800) it happened Sylvia Else

Post by Sylvia Else

One often has to trawl through a number of suggested solutions, either
because most of them are wrong (or at least wildly apocryphal),
irrelevant, or because the same or similar symptoms can have many
different underlying causes.
I have to wonder whether a language model is really up to the task of
filtering out the dross, while keeping the important parts.

There is always the issue that one at least needs to know some stuff to use the answers...
Same may go for an AI created answer.
I have been deliberately trying to get AI to get the concepts wrong
told it for example I died and went to heaven but was not let in there because I did not have the required 4 COVID shots.
Makes me wonder if it replies to questions by somebody about going to heaven 'Make sure you have the 4 COVID shots".

But that takes me on a religious side-track.. is not that the way many kids are brain washed (wrong-educated)
growing up in religious environments, with other requirement for heaven than COVID shots? (Oh well maybe that too)...
creating all the various religious fanatic groups we see now being played out against each other by the US military industrial complex
for weapon sales...

:-) drifting of the topic :-)
I have never _contributed_ to Stack Overflow, google will find Usenet with lots of specific answers, one can ask here.
Some groups are still alive.
I hope it stays that way.

Don

2024-05-10 14:53:21 UTC

Post by Jan Panteltje

Post by Sylvia Else

One often has to trawl through a number of suggested solutions, either
because most of them are wrong (or at least wildly apocryphal),
irrelevant, or because the same or similar symptoms can have many
different underlying causes.
I have to wonder whether a language model is really up to the task of
filtering out the dross, while keeping the important parts.

There is always the issue that one at least needs to know some stuff to use the answers...
Same may go for an AI created answer.
I have been deliberately trying to get AI to get the concepts wrong
told it for example I died and went to heaven but was not let in there
because I did not have the required 4 COVID shots.
Makes me wonder if it replies to questions by somebody about going to heaven
'Make sure you have the 4 COVID shots".
But that takes me on a religious side-track.. is not that the way many
kids are brain washed (wrong-educated)
growing up in religious environments, with other requirement for heaven
than COVID shots? (Oh well maybe that too)...
creating all the various religious fanatic groups we see now being played
out against each other by the US military industrial complex
for weapon sales...

In my experience, an AI answer's advantageousness is directly
proportional to the applied technology content of the question. In other
words, questions about applications amass applicable answers while the
results of questions regarding abstractions are all the place, with a
Wikipedia bias.

It's feasible for the fine print of social sites similar to
Stackoverflow to stipulate all rights to user content belong to the
website owner. The quid pro quo is the owner's out-of-pocket expenses to
host the site.

Everything comes at a price. And this perfectly illustrates why people
absolutely must host their own websites in order to protect their
rights.

Danke,

--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.

Don Y

2024-05-10 17:07:45 UTC

Post by Don
It's feasible for the fine print of social sites similar to
Stackoverflow to stipulate all rights to user content belong to the
website owner. The quid pro quo is the owner's out-of-pocket expenses to
host the site.
Everything comes at a price. And this perfectly illustrates why people
absolutely must host their own websites in order to protect their
rights.

Litigation is what will protect your rights; merely hosting a site
(that can be archived and reused at a later date by any number of
visitors) only controls what that site will PUBLISH at some instant
in time.

Can you prevent a 'bot from scraping your site and using that
content to "educate a visitor"? *Train* an AI??

Don

2024-05-10 17:27:43 UTC

Litigation is what will protect your rights; merely hosting a site
(that can be archived and reused at a later date by any number of
visitors) only controls what that site will PUBLISH at some instant
in time.
Can you prevent a 'bot from scraping your site and using that
content to "educate a visitor"? *Train* an AI??

Both 'bots and litigation are separate topics.

My comments pertain to rights retention. After you sign away your
rights, nothing's left to litigate.

If it helps, think of it this way: a website's owner is legally entitled
to rip you off when you sign away your rights.

Danke,

--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.

Don Y

2024-05-10 20:18:37 UTC

Litigation is what will protect your rights; merely hosting a site
(that can be archived and reused at a later date by any number of
visitors) only controls what that site will PUBLISH at some instant
in time.
Can you prevent a 'bot from scraping your site and using that
content to "educate a visitor"? *Train* an AI??

Both 'bots and litigation are separate topics.

Bots are the exact corollary to AI; what's the difference between
me, as a human, scraping your site (even if I don't do it mechanically)
and LEARNING from everything contained therein... vs. a bot scraping it
for an AI?

Post by Don
My comments pertain to rights retention. After you sign away your
rights, nothing's left to litigate.

When *your* site is scraped, where are your rights? Can you
prove that my AI derived some/all of its knowledge from the
"copyright-protected content" on your site?

Post by Don
If it helps, think of it this way: a website's owner is legally entitled
to rip you off when you sign away your rights.

So, as the site's owner, what protections do *you* have
regarding *your* content (regardless of its source)?

Once you publish, you're exposed. I make a point of inserting
small bugs into any code that I publish as exemplars. My thinking
is that anyone who is interested in the points being illustrated will
TRY to run the code, encounter an error AND THEN LOOK *INTO* THE CODE
in an attempt to UNDERSTAND it. That last point being the exact
point of providing exemplars! :>

(Anyone -- or anyTHING -- intent on just COPYING it will replicate the bug)

Don

2024-05-10 22:22:02 UTC

Litigation is what will protect your rights; merely hosting a site
(that can be archived and reused at a later date by any number of
visitors) only controls what that site will PUBLISH at some instant
in time.
Can you prevent a 'bot from scraping your site and using that
content to "educate a visitor"? *Train* an AI??

Both 'bots and litigation are separate topics.

Post by Don
My comments pertain to rights retention. After you sign away your
rights, nothing's left to litigate.

When *your* site is scraped, where are your rights? Can you
prove that my AI derived some/all of its knowledge from the
"copyright-protected content" on your site?

Post by Don
If it helps, think of it this way: a website's owner is legally entitled
to rip you off when you sign away your rights.

So, as the site's owner, what protections do *you* have
regarding *your* content (regardless of its source)?
Once you publish, you're exposed. I make a point of inserting
small bugs into any code that I publish as exemplars. My thinking
is that anyone who is interested in the points being illustrated will
TRY to run the code, encounter an error AND THEN LOOK *INTO* THE CODE
in an attempt to UNDERSTAND it. That last point being the exact
point of providing exemplars! :>
(Anyone -- or anyTHING -- intent on just COPYING it will replicate the bug)

"What we've got here is failure to communicate."

Your questions and arguments again fall outside of the scope of my
followup. My comments pertain to rights retention.

Danke,

--
Don, KB7RPU, https://www.qsl.net/kb7rpu
There was a young lady named Bright Whose speed was far faster than light;
She set out one day In a relative way And returned on the previous night.

Don Y

2024-05-11 00:04:34 UTC

Litigation is what will protect your rights; merely hosting a site
(that can be archived and reused at a later date by any number of
visitors) only controls what that site will PUBLISH at some instant
in time.
Can you prevent a 'bot from scraping your site and using that
content to "educate a visitor"? *Train* an AI??

Both 'bots and litigation are separate topics.

Post by Don
My comments pertain to rights retention. After you sign away your
rights, nothing's left to litigate.

When *your* site is scraped, where are your rights? Can you
prove that my AI derived some/all of its knowledge from the
"copyright-protected content" on your site?

Post by Don
If it helps, think of it this way: a website's owner is legally entitled
to rip you off when you sign away your rights.

So, as the site's owner, what protections do *you* have
regarding *your* content (regardless of its source)?
Once you publish, you're exposed. I make a point of inserting
small bugs into any code that I publish as exemplars. My thinking
is that anyone who is interested in the points being illustrated will
TRY to run the code, encounter an error AND THEN LOOK *INTO* THE CODE
in an attempt to UNDERSTAND it. That last point being the exact
point of providing exemplars! :>
(Anyone -- or anyTHING -- intent on just COPYING it will replicate the bug)

"What we've got here is failure to communicate."
Your questions and arguments again fall outside of the scope of my
followup. My comments pertain to rights retention.

The *post* pertains to AI harvesting information posted
by site users.

Crash Gordon

2024-05-11 16:19:06 UTC

Post by Don
...
If it helps, think of it this way: a website's owner is legally entitled
to rip you off when you sign away your rights.

Watch the "Joan is Awful" episode of Black Mirror to see this in action.

--
I'm part of the vast libertarian conspiracy to take over the world and
leave everyone alone.

Bill Sloman

2024-05-10 15:21:53 UTC

Post by Sylvia Else

One often has to trawl through a number of suggested solutions, either
because most of them are wrong (or at least wildly apocryphal),
irrelevant, or because the same or similar symptoms can have many
different underlying causes.
I have to wonder whether a language model is really up to the task of
filtering out the dross, while keeping the important parts.

And Jan Panteltje wouldn't notice. His enthusiasm for the the Le Sage
theory of gravity is an example of his latching onto dross.

--
Bill Sloman, Sydney

a***@spenarnc.xs4all.nl

2024-05-11 11:45:01 UTC

Post by Jan Panteltje

Post by Jan Panteltje
Stack Overflow users sabotage their posts after OpenAI deal

https://arstechnica.com/information-technology/2024/05/stack-overflow-users-sabotage-their-posts-after-openai-deal/

Post by Jan Panteltje
The end of stack overflow?
Personally I know companies are using my open sourced stuff..
Let it be...
I like Stackoverflow, it gave me many good answers in the past
to difficult questions...
Many highly qualified people there.

One often has to trawl through a number of suggested solutions, either

The most annoying is that 4 of 5 question that I want to
ask there are destroyed by the moderators.
It is largely arbitrary.
Explain the difference between c and lisp macro's?
IMHO dubious. Huge upvotes, apparently acceptable.
Explain the difference between Forth and lisp macro's?
A more sensible question because they are more related.
Shot down, inappropiate for SO.
My sensible m4 question (how to write a function that
removes the first character of a string) was voted down.

Post by Jan Panteltje
because most of them are wrong (or at least wildly apocryphal),
irrelevant, or because the same or similar symptoms can have many
different underlying causes.

The idea that they should be voted down.
A sensible attitude towards chatgpt, is to add the answer of
chatgpt immediately, and ask the poster of the question to rate
it down.

Post by Jan Panteltje
I have to wonder whether a language model is really up to the task of
filtering out the dross, while keeping the important parts.

I think yes, in the long run.

Post by Jan Panteltje
Sylvia.

--
Don't praise the day before the evening. One swallow doesn't make spring.
You must not say "hey" before you have crossed the bridge. Don't sell the
hide of the bear until you shot it. Better one bird in the hand than ten in
the air. First gain is a cat purring. - the Wise from Antrim -

Don Y

2024-05-11 19:39:34 UTC

Post by a***@spenarnc.xs4all.nl
The most annoying is that 4 of 5 question that I want to
ask there are destroyed by the moderators.

For all the folks who think "moderation" is the solution... :>

Post by a***@spenarnc.xs4all.nl
It is largely arbitrary.
Explain the difference between c and lisp macro's?
IMHO dubious. Huge upvotes, apparently acceptable.
Explain the difference between Forth and lisp macro's?
A more sensible question because they are more related.
Shot down, inappropiate for SO.
My sensible m4 question (how to write a function that
removes the first character of a string) was voted down.

In Inferno/Limbo, you would use "len" to determine the number
of characters in the argument (to guard against degenerate
cases -- with appropriate conditional logic) and then "substr"
to peel off the characters following the "0-th" character.

boB

2024-05-12 00:27:19 UTC

On Sat, 11 May 2024 12:39:34 -0700, Don Y

Post by a***@spenarnc.xs4all.nl
The most annoying is that 4 of 5 question that I want to
ask there are destroyed by the moderators.

For all the folks who think "moderation" is the solution... :>

I have found that any of my contributions to a question are either
removed or edited. I thought it was just me but it looks like it
isn't just me.

This may be for stack-exchange or in addition to.

Are they owned by the same people ?
The article called them "sisters"

boB

Don Y

2024-05-12 01:51:40 UTC