TOS Question: ChatGPT

[Garf]

I think you're just bitter that it is on team algae scrubber lol

Lol. It has a lot to learn :)

 

[Miami Reef]

I think you're just bitter that it is on team algae scrubber lol

My AI is smarter than yours:

AI will spit out the answers based on how you word the question. See?

 

[areefer01]

If I ask it to translate DIY Randy's magnesium additive from gallons to metric, why does it gimme a different answer every time?

I am not sure. Could it be how you phrased the question?

Note I was not promoting it to answer questions only saying that in time it may be. And why that may be is because of the data it could have access to and the compute driving it.

On a personal level I started with books, clubs, and good old USENET. Thus my generation comment.

 

[Garf]

I am not sure. Could it be how you phrased the question?

Note I was not promoting it to answer questions only saying that in time it may be. And why that may be is because of the data it could have access to and the compute driving it.

On a personal level I started with books, clubs, and good old USENET. Thus my generation comment.

Pretty sure it makes crap up, based on a foundation of ignorance, just like people.

 

[areefer01]

Pretty sure it makes crap up, based on a foundation of ignorance, just like people.

Maybe it needs more badges?

 

[Crabs McJones]

My AI is smarter than yours:

AI will spit out the answers based on how you word the question. See?

Charge your battery, it's giving me anxiety :rolling-on-the-floor-laughing:

 

[Miami Reef]

Charge your battery, it's giving me anxiety :rolling-on-the-floor-laughing:

I would, but my cat is on me and I can’t get up unless he does.

I don’t make the rules. :grinning-face-with-smiling-eyes:

 

[mizimmer90]

My AI is smarter than yours:

AI will spit out the answers based on how you word the question. See?

It's almost like it's a "tool" that you can use for various purposes ;)

 

[Garf]

I would, but my cat is on me and I can’t get up unless he does.

I don’t make the rules. :grinning-face-with-smiling-eyes:

Is it a tiger?? Who's the boss?

 

[Miami Reef]

Is it a tiger?? Who's the boss?

Haha. It’s the unspoken rule of cat ownership: never get up if a cat is resting on you.

 

[Bitter old man]

OK, I thought you wanted to have Reef2Reef vet ALL members, which would just be impossible for them I'm sure.

 

[Crabs McJones]

Haha. It’s the unspoken rule of cat ownership: never get up if a cat is resting on you.

 

[Miami Reef]

I just laughed so hard!!! I’m saving this.

 

[mizimmer90]

Haha. It’s the unspoken rule of cat ownership: never get up if a cat is resting on you.

 

[Randy Holmes-Farley]

Maybe chatgpt isn't the answer today but as the data it has access to grows it very well could be in 2 or 3 years. Just as spice is the most precious substance in the universe (Dune reference) or water on earth data is chatgpt's precious substance. Anyone awake can see it being collected be it a medical test, record, or something put in ones favorite search engine...

The levee broke - no stopping it now.

While I blast ChatGPT for being inaccurate, I’m quite confident that it won’t be long before ai are much too smart for our own good. They will know the right answer, but may or may not give it to us depending on their mood and whether they like the questioner. Lol

 

[sixty_reefer]

I once mentioned to Rev that it might be a good idea to begin a mentoring program. If a member wanted to mentor newcomers to the hobby, they might message Rev expressing interest in doing so. Rev would then reply with a questionnaire that would have been assembled as a collaboration by moderators. The questionnaire would be completed and returned for review. If the applicant passes muster, this would be indicated in some way on their profile.

You wouldn't have to screen everyone, only those interested in applying. And, I'm not sure what you mean by "spoofed" where a questionnaire is concerned. I mean, I understand spoofing an email or phone number, but how or why would one spoof a questionnaire?

People work from home, so I don't see the need to go to a person's home to see their results. Pictures and video could easily do the same.

And yes, I do think a tank should be a requirement. Participating here without a tank would be the same as participating on a boating forum and not owning a boat. :face-with-tears-of-joy:

A down vote button would be helpful, but giving everyone access to such a button would also give users the option to down vote when they take something personally, even if the response they take personally has merit.

Badges are eye-candy. They carry little weight in my view. To repeat my example, there are people here with 150k reaction score, but most of that is the result of a pretty picture in a "welcome" thread.

As a salt water forum, I want information on salt water subjects. I'm not interested things like "Excellence Award" as I find it ambiguous. Excellent at what? It tells us nothing about the users skill set. The same is true for a "Hospitality" badge. Sure, you're a nice person and that is important to me because I don't like mean people. But you can be a nice person and still provide wrong information.

So I think badges can be considered, but they would carry less weight in the decision process than a targeted questionnaire would.

They could also just ask chat gpt for the users credentials.

Apparently they know almost everything about every one on here. I may start using it in future to see the credentials of the person I’m speaking with.
Apparently even chat gpt seems confused to my real profession lol

I’ve asked them a question that is often asked to me:

 

[mizimmer90]

While I blast ChatGPT for being inaccurate, I’m quite confident that it won’t be long before ai are much too smart for our own good. They will know the right answer, but may or may not give it to us depending on their mood and whether they like the questioner. Lol

Always make sure to say please and thank you to your LLMs!

 

[A_Blind_Reefer]

While I blast ChatGPT for being inaccurate, I’m quite confident that it won’t be long before ai are much too smart for our own good. They will know the right answer, but may or may not give it to us depending on their mood and whether they like the questioner. Lol

Kinda like me…. With the addition of before coffee and after coffee

 

[Garf]

I've certainly never had a machine thank me before, lol. Do these things have a habit of missing out crucial parts of the question?

 

[A_Blind_Reefer]

Always make sure to say please and thank you to your LLMs!

I must admit, this does make me feel a little weird. I totally rely on Google, Alexa, Siri being blind as a bat. I most always say please and thank you out of habit…. It was the way I was brought up. Alexa has always responded with you’re welcome, just doing my job, your kindness really gives me a charge…whatever. Siri recently started responding (thankless brat for a few years!) with you’re welcome, no problem…and then there’s Google. Google is like the honey badger. Google don’t give a xxxx! I’m curious if there’s a database tracking how proper, or rude, people are with their devices. Even though it’s habit, it feels a little creepy.