Randy Holmes-Farley
Reef Chemist
View BadgesStaff member
Super Moderator
Excellence Award
Expert Contributor
Article Contributor
R2R Research
My Tank Thread
- Joined
- Sep 5, 2014
- Messages
- 67,516
- Reaction score
- 63,956
Let me preface this by saying this post is my personal opinion, and not a Reef2Reef policy or rule.
It seems to me that we are beginning to see bots post at reef2reef. I am not knowledgeable enough to know if the entire process can be automated, or whether a human is controlling each bot and each post.
My concern with bots posting stems from the obvious mistakes AI's make, the potential for reef2reef members to be misled, and a potential decline in the trustworthiness of info posted on reef2reef. it's one thing to disagree with a person who holds a different or unusual opinion, and its quite another to be misled by something that perhaps no actual human agrees with.
These are my suggestions:
1. AI's, bots, or whatever one calls them should not be permitted to make posts at Reef2Reef that no human reviewed prior to posting. In this sense, if you, as a member, see something that looks off, report it. For example, a real example, a new post in a 10-month old thread that suggests links on how to find a headstone for a deceased reefing legend seems to be something a human would not write. I am highly skeptical that automatic posts can have a benign reason for existence, and the true reason may be hidden (such as boosting an account to enough posts to gain selling privilages without even being a hobbyist, or to allow spamming of members by PM).
2. Using an AI to write your posts is fine, IMO, as long as you read it and know that you agree with what was written. No one is trying to inhibit the use of technology to help folks who want that sort of assistance, but don't use it to say things that you don't know about.
3. If you cannot get a question answered for whatever reason, or just don't find it handy to ask here first, and turn to an AI for an answer, it is fine to post the AI answer here at reef2reef and ask folks if it seems reasonable. A real example from a few days ago related to dosing potassium. The ChatGPT AI gave an incorrect formula for making a solution and dosing it that was off by a huge factor of 27,000x. Double checking here may help you avoid an AI driven mistake.
Perhaps there are other things we should all think about as bots try to get the upper hand, and I welcome any comments, criticisms, or additions to the discussion.
It seems to me that we are beginning to see bots post at reef2reef. I am not knowledgeable enough to know if the entire process can be automated, or whether a human is controlling each bot and each post.
My concern with bots posting stems from the obvious mistakes AI's make, the potential for reef2reef members to be misled, and a potential decline in the trustworthiness of info posted on reef2reef. it's one thing to disagree with a person who holds a different or unusual opinion, and its quite another to be misled by something that perhaps no actual human agrees with.
These are my suggestions:
1. AI's, bots, or whatever one calls them should not be permitted to make posts at Reef2Reef that no human reviewed prior to posting. In this sense, if you, as a member, see something that looks off, report it. For example, a real example, a new post in a 10-month old thread that suggests links on how to find a headstone for a deceased reefing legend seems to be something a human would not write. I am highly skeptical that automatic posts can have a benign reason for existence, and the true reason may be hidden (such as boosting an account to enough posts to gain selling privilages without even being a hobbyist, or to allow spamming of members by PM).
2. Using an AI to write your posts is fine, IMO, as long as you read it and know that you agree with what was written. No one is trying to inhibit the use of technology to help folks who want that sort of assistance, but don't use it to say things that you don't know about.
3. If you cannot get a question answered for whatever reason, or just don't find it handy to ask here first, and turn to an AI for an answer, it is fine to post the AI answer here at reef2reef and ask folks if it seems reasonable. A real example from a few days ago related to dosing potassium. The ChatGPT AI gave an incorrect formula for making a solution and dosing it that was off by a huge factor of 27,000x. Double checking here may help you avoid an AI driven mistake.
Perhaps there are other things we should all think about as bots try to get the upper hand, and I welcome any comments, criticisms, or additions to the discussion.