TOS Question: ChatGPT

[AKReefing]

Because we’re not going to let AI answers take over here. :)

Love ya, Rev.

 

[ISpeakForTheSeas]

As someone who doesn't have a saltwater aquarium (yet) but who has been around for awhile, I just want to add that the issues I've seen people taking aren't so much with the fact that AI is being used in your responses; it's that:

1 ) The responses are being presented as firsthand experience rather than as AI generated content (which, speaking as someone with a background in psychology, is a lot more important of a distinction than one might initially expect, as the projected confidence of AI being presented as firsthand knowledge is likely to sway a lot of people to view it as correct and trustworthy even if it's egregiously wrong).

2a ) The responses aren't being quality checked appropriately (what qualifies an appropriate quality check is somewhat subjective, but generally speaking the answers should probably align at least somewhat with information presented by either literal professionals or be demonstrated by hobbyists; as someone with no aquarium, my personal experience is pretty much worthless - I almost exclusively cite scholarly sources and firsthand accounts if I feel the information I'm providing needs supporting evidence).

2b ) Piggybacking off of 2a, because of how AI, particularly LLM-based AI (like ChatGPT), work, the information they provide will often be the most readily available, the most commonly repeated (but not necessarily true), and some just plain made up (yes, AI just makes things up sometimes).

And (really just hammering down on the same vein as 2a and 2b):

3 ) AI cannot vet information. AI is a fantastic tool, but also horribly deeply flawed at the moment (and possibly forever, though it will definitely get better), because it lacks the vetting capabilities of humans. For example, I primarily do hitchhiker ID's here, and sometimes I'll use Google Lens on an image to help with the ID process (I use it to help point me in the right general direction to help narrow down some ID's; sometimes it's helpful, and sometimes it's not - honestly, the more I learn, the less I use AI for direction, but sometimes it gets close enough with some of the results [almost never at the top] to help me figure out a taxonomic family, superfamily. etc.); it's a tool in my repertoire. However, trusting AI (Google Lens, in this case) to figure out the right ID would lead to incorrect identification pretty well every single time (the amount of times it tries to tell me that a crab, for example, is a whale shark or a cuttlefish is somewhat mind-boggling; and it's very confident in its replies, even going so far as to rebut my prompts to point it in the right direction and to tell me "No, this picture is not of a crab, it's of a whale shark"). AI is a tool to use to pursue the answers needed for our questions at this point, its answers are not actually the answers in the vast majority of cases (and not just due to the complexity of all the different circumstances/variables of each aquarium).


TLDR; trusting AI to be right is a major (likely unwise) leap of faith at this point, but AI is a good tool to get ideas or direction from in order to figure out an appropriate, correct (i.e. probably not AI) response through your own, human, vetting process and experience.

[Edit: To add, OP, if you want to keep using AI, you'd probably get a better response from the forum if you specifically noted in bold at the top of your posts that your answer is AI generated, not experience-based; the fact that your answers don't state openly that they are AI generated may be leaving some people feeling like you're trying to deceive others on the forum by presenting things as personal experience and fact when you don't actually know these things for sure, even if your purpose is just to help.]

 

[EeyoreIsMySpiritAnimal]

Wow some people… I’m going through threads where people are left hanging with no replies, and I’m using ChatGPT to help provide them with useful information. What’s the problem with that? It’s not like I’m trying to replace anyone, I'm just making sure people get answers when they need them. Whether you like it or not, ChatGPT is an insanely powerful tool, packed with way more knowledge than any one person could ever hold. It's the future of how we share information, and pretending like it’s some evil thing isn’t going to change that.

I’m using the tool to help fill in the gaps and actually get people answers, not to “cheat” or “fake” anything. If you can’t see the value in that, that’s on you. People are getting the info they need, and that’s all that matters. If you think using ChatGPT is a bad thing, maybe it’s time to step into the present and stop acting like it’s some threat to your outdated ways of doing things.

You just don't get it...

 

[AKReefing]

You just don't get it...

Wait 'til AI tells him to use table salt.

 

[FindinNemo]

Wait 'til AI tells him to use table salt.

Tomato tomato

 

[BeanAnimal]

Wow some people… I’m going through threads where people are left hanging with no replies, and I’m using ChatGPT to help provide them with useful information. What’s the problem with that?

But most of it is not useful, lacks meaningful context and is generally just bothersome to the bulk of the people exposed to it.

What you are missing is that even if it was 100% accurate and contextually relevant, it is not what people come here for and, in-fact appears to be almost universally bothersome to most.

I’m using the tool to help fill in the gaps and actually get people answers, not to “cheat” or “fake” anything. If you can’t see the value in that, that’s on you.

Most of us don't see the value in it. Any of us can use google or AI to search for answers. The spirit of the forum is sharing personal knowledge and conversation. We don't need you or anybody else running around filling in gaps with AI or acting as our ChatGPT concierge. To that end, the forum owners could simply add an AI plugin to answer all responses in all threads.

I suggest maybe asking your AI how to read a room.

 

[Randy Holmes-Farley]

Wow some people… I’m going through threads where people are left hanging with no replies, and I’m using ChatGPT to help provide them with useful information. What’s the problem with that?

Really? You jumped into this thread in less than 3 h with an ai answer that I called out as a poor answer. 3 h is not too long to wait for a responder to a highly specific question that you failed to actually answer.

Invasive blue macro algea

Hi, I am new to marine tanks and would like some advice/ suggestion please. I introduced blue macro algea into the tank, wish I had not, as very invasive. I have tried to remove as much as possible by hand, but grows back very fast. I have an emerald crab, but he does not seem to eat it. I am...

www.reef2reef.com

 

[FindinNemo]

I understand the concerns and value learning from the community’s personal experiences. As a newcomer to saltwater aquariums, I recognize how important real world insights are. I truly appreciate the wealth of knowledge here, and I know it’s something you can’t get from just looking up answers online.

Moving forward, I’ll focus on learning directly from all of you and contributing with my own experiences. I’m grateful to be part of this community and will engage in a way that’s respectful, helpful, and focused on meaningful exchange. Let’s all move on from this and grow!

 

[twentyleagues]

As someone who doesn't have a saltwater aquarium (yet) but who has been around for awhile, I just want to add that the issues I've seen people taking aren't so much with the fact that AI is being used in your responses; it's that:

1 ) The responses are being presented as firsthand experience rather than as AI generated content (which, speaking as someone with a background in psychology, is a lot more important of a distinction than one might initially expect, as the projected confidence of AI being presented as firsthand knowledge is likely to sway a lot of people to view it as correct and trustworthy even if it's egregiously wrong).

2a ) The responses aren't being quality checked appropriately (what qualifies an appropriate quality check is somewhat subjective, but generally speaking the answers should probably align at least somewhat with information presented by either literal professionals or be demonstrated by hobbyists; as someone with no aquarium, my personal experience is pretty much worthless - I almost exclusively cite scholarly sources and firsthand accounts if I feel the information I'm providing needs supporting evidence).

2b ) Piggybacking off of 2a, because of how AI, particularly LLM-based AI (like ChatGPT), work, the information they provide will often be the most readily available, the most commonly repeated (but not necessarily true), and some just plain made up (yes, AI just makes things up sometimes).

And (really just hammering down on the same vein as 2a and 2b):

3 ) AI cannot vet information. AI is a fantastic tool, but also horribly deeply flawed at the moment (and possibly forever, though it will definitely get better), because it lacks the vetting capabilities of humans. For example, I primarily do hitchhiker ID's here, and sometimes I'll use Google Lens on an image to help with the ID process (I use it to help point me in the right general direction to help narrow down some ID's; sometimes it's helpful, and sometimes it's not - honestly, the more I learn, the less I use AI for direction, but sometimes it gets close enough with some of the results [almost never at the top] to help me figure out a taxonomic family, superfamily. etc.); it's a tool in my repertoire. However, trusting AI (Google Lens, in this case) to figure out the right ID would lead to incorrect identification pretty well every single time (the amount of times it tries to tell me that a crab, for example, is a whale shark or a cuttlefish is somewhat mind-boggling; and it's very confident in its replies, even going so far as to rebut my prompts to point it in the right direction and to tell me "No, this picture is not of a crab, it's of a whale shark"). AI is a tool to use to pursue the answers needed for our questions at this point, its answers are not actually the answers in the vast majority of cases (and not just due to the complexity of all the different circumstances/variables of each aquarium).


TLDR; trusting AI to be right is a major (likely unwise) leap of faith at this point, but AI is a good tool to get ideas or direction from in order to figure out an appropriate, correct (i.e. probably not AI) response through your own, human, vetting process and experience.

[Edit: To add, OP, if you want to keep using AI, you'd probably get a better response from the forum if you specifically noted in bold at the top of your posts that your answer is AI generated, not experience-based; the fact that your answers don't state openly that they are AI generated may be leaving some people feeling like you're trying to deceive others on the forum by presenting things as personal experience and fact when you don't actually know these things for sure, even if your purpose is just to help.]

Perfect. I was having a hard time expressing my thoughts but you covered them perfectly. I deleted my proposed post.

 

[Reefering1]

I’m using the tool to help fill in the gaps and actually get people answers, not to “cheat” or “fake” anything. If you can’t see the value in that, that’s on you. People are getting the info they need, and that’s all that matters. If you think using ChatGPT is a bad thing, maybe it’s time to step into the present and stop acting like it’s some threat to your outdated ways of doing things.

I’m pretty sure you missed the consensus of the thread??? If you want to leave here to “let the bots and AI fan peeps have their day in the sun” then you’ll be happy to know that you can stay. Because we’re not going to let AI answers take over here. :)

Do you understand yet? That's the owner of this forum telling you how it is and will be.

 

[AKReefing]

I suggest maybe asking your AI how to read a room.

WHOOMP! THERE IT IS! I'd say that's probably the best advice I've seen.

 

[Fish Fan]

I understand the concerns and value learning from the community’s personal experiences. As a newcomer to saltwater aquariums, I recognize how important real world insights are. I truly appreciate the wealth of knowledge here, and I know it’s something you can’t get from just looking up answers online.

Moving forward, I’ll focus on learning directly from all of you and contributing with my own experiences. I’m grateful to be part of this community and will engage in a way that’s respectful, helpful, and focused on meaningful exchange. Let’s all move on from this and grow!

If that’s sincere, that’s the best post you’ve made on this forum :)

You don’t need to be an expert contributor. Simply asking questions is valuable because the replies you get will undoubtably help others too. Just engaging with the community and helping to welcome new members are others ways to be valuable, as well as genuine. I just think no one wants to feel like they’re talking to a computer when we post for help here ;-)

 

[FindinNemo]

If that’s sincere, that’s the best post you’ve made on this forum :)

You don’t need to be an expert contributor. Simply asking questions is valuable because the replies you get will undoubtably help others too. Just engaging with the community and helping to welcome new members are others ways to be valuable, as well as genuine. I just think no one wants to feel like they’re talking to a computer when we post for help here ;-)

Totally understood and yes it is sincere!

 

[revhtree]

I understand the concerns and value learning from the community’s personal experiences. As a newcomer to saltwater aquariums, I recognize how important real world insights are. I truly appreciate the wealth of knowledge here, and I know it’s something you can’t get from just looking up answers online.

Moving forward, I’ll focus on learning directly from all of you and contributing with my own experiences. I’m grateful to be part of this community and will engage in a way that’s respectful, helpful, and focused on meaningful exchange. Let’s all move on from this and grow!

Thank you and thank you to everyone who is passionate about this hobby and this forum!

 

[Fish Fan]

Thank you and thank you to everyone who is passionate about this hobby and this forum!

Thank you for being on top of this kind of stuff! Hope you’re feeling better today!

 

[Reefering1]

Thank you and thank you to everyone who is passionate about this hobby and this forum!

Soooo... im all for 2nd chances, but i also appreciate clear boundaries.. will this user be allowed to continue with the chatgtp responses? How many strikes before action is taken?

 

[revhtree]

Thank you for being on top of this kind of stuff! Hope you’re feeling better today!

Thank you and I’ve turned the corner although I still have a slight fever.

 

[revhtree]

Soooo... im all for 2nd chances, but i also appreciate clear boundaries.. will this user be allowed to continue with the chatgtp responses? How many strikes before action is taken?

He stated that he’s not going to do it anymore and we’re going to define it in our TOS soon. We’ll go from there.

 

[Reefering1]

He stated that he’s not going to do it anymore and we’re going to define it in our TOS soon. We’ll go from there.

Perfect. Appreciate it boss.
I assume there will be a announcement thread?

 

[Crabs McJones]

I understand the concerns and value learning from the community’s personal experiences. As a newcomer to saltwater aquariums, I recognize how important real world insights are. I truly appreciate the wealth of knowledge here, and I know it’s something you can’t get from just looking up answers online.

Moving forward, I’ll focus on learning directly from all of you and contributing with my own experiences. I’m grateful to be part of this community and will engage in a way that’s respectful, helpful, and focused on meaningful exchange. Let’s all move on from this and grow!

Love it, thank you for your understanding :)