AI - Whats the point in GZ?

Forums › Geekzone › AI - Whats the point in GZ?

1 | 2 | 3 | 4

BDFL - Memuneh
80806 posts

Uber Geek
+1 received by user: 41411

Administrator

ID Verified

Trusted

Geekzone

Lifetime subscriber

#3422557 7-Oct-2025 09:49

Thanks for well wishes.

For everyone else, if you are man, have your prostate checked. If the specialist gives the options of "either wait another 12 months for another MRI to see if there are changes, or do a biopsy now" always go with the biopsy.

I know some may not have health insurance but do it any way you can.

This might be something for the Health sub-forum.

Referral links: Quic Broadband (free setup code: R587125ERQ6VE) | Samsung | AliExpress | Wise | Sharesies

Support Geekzone by subscribing (browse ads-free), or making a one-off or recurring donation through PressPatron.

timbosan

2201 posts

Uber Geek
+1 received by user: 294

Subscriber

#3422560 7-Oct-2025 09:56

freitasm:

Thanks for well wishes.

For everyone else, if you are man, have your prostate checked. If the specialist gives the options of "either wait another 12 months for another MRI to see if there are changes, or do a biopsy now" always go with the biopsy.

I know some may not have health insurance but do it any way you can.

This might be something for the Health sub-forum.

I can relate to this and say the same - just get it checked, sooner rather than later. Luckily I haven't needed a biopsy (just the finger!) but its so important to have it checked.

Take care @freitasm :-)

eracode

Smpl Mnmlst
9379 posts

Uber Geek
+1 received by user: 6265

ID Verified

Trusted

Lifetime subscriber

#3422561 7-Oct-2025 09:57

eracode:

Gordy7:

Had a chat with Meta AI this morning.

When asked about HAL from 2001, Llama 4 was quick to point out that HAL was fictional - Dave 😀

I thought Llama 4 might reply “I’m sorry, Dave. I’m afraid I can’t do that.”

Disclosure: I used ChatGPT to check my memory on exactly what HAL had said (after all, it was 57 years ago) - then slightly edited my post to show it correctly. 😀

IMO this shows that using AI here is not always subversive - it's a matter of how it's used.

Sometimes I just sit and think. Other times I just sit.

lxsw20

3710 posts

Uber Geek
+1 received by user: 2190

Subscriber

#3422631 7-Oct-2025 10:30

eracode:

I vote for not totally banning AI - not sure if it can be policed anyway.

I'd like to see a requirement that if a post is based on AI info, it should be fully disclosed and acknowledged by the poster. Not sure if that could be policed either.

Agree on not a total ban....hence a policy. But I can't see the point in AI says this or Google says this. That isn't personal experience.

If I don't know something then move on to the next thread on Geekzone - there isn't a need to google / LLM an answer for a person beyond clarifying or backing up your own experience. They can do that themselves.

Bung

6748 posts

Uber Geek
+1 received by user: 2932

Subscriber

#3422662 7-Oct-2025 11:11

lxsw20:

They can do that themselves.

If only they'd do that 1st then ask for for interpretation of the answer. The usual insurance co FUD has doubtful value.

Rikkitic

Awrrr
19125 posts

Uber Geek
+1 received by user: 16413

Lifetime subscriber

#3422746 7-Oct-2025 15:21

In my old age my memory is not always reliable and I use tools like Perplexity as necessary as a jog and to verify things. I am perfectly well aware that it hallucinates and goes down rabbit holes and whatever, but I usually know when what it says is true or not. I find it a useful tool, not a brain substitute.

Plesse igmore amd axxept applogies in adbance fir anu typos

New World

Shop on-line at New World now for your groceries (affiliate link).

TwoSeven

1717 posts

Uber Geek
+1 received by user: 304

Subscriber

#3422860 7-Oct-2025 18:25

I'm not the greatest fan of what is currently referred to as A.I. especially in a social forum where human discourse is encouraged. I don't mind it in a technical subforum about A.I. but even then, I think there should be context.

Software Engineer
(the practice of real science, engineering and management)
A.I. (Automation rebranded)
Gender Neutral
(a person who believes in equality and who does not believe in/use stereotypes. Examples such as gender, binary, nonbinary, male/female etc.)

...they/their/them...

pdh

470 posts

Ultimate Geek
+1 received by user: 316

#3422881 7-Oct-2025 20:42

Ok - here's a real-world example of what I like to think is AI's strength.

A friend & I had a debate about immigration into NZ (as we're both immigrants of 30 & 40 year duration).

He contended that poms still made up a very large proportion (annually) - while I felt that poms were a thing of the past (neither of us is a pom - perhaps you guessed).

So I asked Perplexity to assemble a nice graph (from a recent year) of poms vs other significant migrant groups.

I won't bore you with the results - as you can ask your preferred AI the same question...

But it gave the two of us some hard data to discuss, over the next week's beer.

I'm sure that I could have come up with something like that data - 40 years ago - by hitting the library or writing to NZ's bureaucracy... but it would have taken hours & days - not seconds.

And no - I didn't automatically trust the answers - but I did ask for & get the sources of the data, which I was able to assess for propaganda / bias.

So I / we both felt that we were able to have a semi-intelligent discussion on a subject that I have not seen covered in any depth by our mainstream media.

I feel better for the opportunity to inform myself and think about stuff - which even 20 years ago would have been much, much harder to do.

That's a simple example of using AI (2025 version) as a tool.
Not a source of what to think - but as an aid to thinking...

pdh

470 posts

Ultimate Geek
+1 received by user: 316

#3422882 7-Oct-2025 20:49

Perhaps I should have followed that by saying explicitly that - much as I admire the Geekzone forum members - I don't expect any one of them - or even all of them collectively - could have come up with as much info as Ai was able to muster on recent immigration stats.

Geekzone members have a tremendous amount of personal experience & knowledge - much of which isn't written down anywhere.

Modern AI is just a bunch of Large Language Models - soaking up all written info, distilling it & regurgitating it.

There's obviously a difference - play each to its strengths.

timmmay

20900 posts

Uber Geek
+1 received by user: 5381

Trusted

Lifetime subscriber

#3422886 7-Oct-2025 21:06

I don't think anyone is saying AI isn't useful, just that we shouldn't be pasting AI answers into Geekzone.

pdh

470 posts

Ultimate Geek
+1 received by user: 316

#3422890 7-Oct-2025 21:19

If I paste an AI answer - and you can usefully challenge it...
Haven't we learned something useful ?

Quic Broadband

Move to New Zealand's best fibre broadband service (affiliate link). Free setup code: R587125ERQ6VE. Note that to use Quic Broadband you must be comfortable with configuring your own router.

eracode

Smpl Mnmlst
9379 posts

Uber Geek
+1 received by user: 6265

ID Verified

Trusted

Lifetime subscriber

#3422911 7-Oct-2025 23:22

pdh:

If I paste an AI answer - and you can usefully challenge it...
Haven't we learned something useful ?

Not necessarily - and especially not if it’s passed off as being from the poster’s own experience and thought. Hence my earlier suggestion that for any post copied from, or based on, AI sources that fact should be required to be disclosed and attributed by the poster.

Sometimes I just sit and think. Other times I just sit.

SomeoneSomewhere

1895 posts

Uber Geek
+1 received by user: 1097

Lifetime subscriber

#3422912 7-Oct-2025 23:23

pdh:

If I paste an AI answer - and you can usefully challenge it...
Haven't we learned something useful ?

Only if enough people have the time, inclination, and experience to challenge it when it's wrong.

AI does a great job at making something that looks plausible on the surface, and responding to criticism in a plausible way. But is complete garbage.

It's like arguing with people who design perpetual motion machines (many of which are now turning to AI...) - it doesn't even make enough sense to be debunkable.

It further worsens the issue of it taking more effort to disprove misinformation than to post misinformation.

alasta

6907 posts

Uber Geek
+1 received by user: 3391

Trusted

Subscriber

#3422935 8-Oct-2025 08:56

A while ago my friend was showing me how to use AI and we asked it to write a poem about President Trump. What is came up with was pure brilliance.

It was a different story at work when I asked it to come up with something that involves facts and evidence. I actually documented my experience for the benefit of the rest of my team and I stated that interacting with the AI chatbot was like "an awkward conversation with an under-performing employee".

timmmay

20900 posts

Uber Geek
+1 received by user: 5381

Trusted

Lifetime subscriber

#3422940 8-Oct-2025 09:10

The quality of the output from an AI is directly related to the quality of the prompt and the capability of the model. The best model will perform poorly with a short or poorly defined a prompt. With a large informative prompt you can get surprising amounts of useful information out of an AI. My largest prompt to date has been about three A4 pages, which resulted in a fairly useful document being created including the required sections.

Having said that, sometimes AI is just hopeless.

1 | 2 | 3 | 4