AI and the SBO Forums-- Part 2 Community Expectations

[Dave]

Thank you all for your thoughtful comments on the SBO and AI thread. I started the thread to get a sense of how the community felt about members using AI to generate their responses to questions raised in the forums.

Thirty-five years ago, when I moderated my first forum on America Online, I found it fascinating that AOL members were forming friendships and groups, a new style of community was forming as members sought human connections online. That desire for a human connection was clearly evident today in the SBO and AI discussion. Members want to hear from other members about their experiences with whatever issue is at hand.

The AI content question must be addressed in the context of this desire for authentic human connections. Does posting AI generated content detract from the human element inherent in the forums? The sentiment was clear, it does. Posting AI generated content without attribution is deceitful. Deceit never enhances a human relationship. And let’s call it what it is, plagiarism. It is presenting information as your own when it is has been copied and pasted from another source. Plagiarism is never acceptable.

Posting AI generated content with a citation was not met with resounding support primarily because it does not reflect the member’s unique experience. It can leave the impression the poster has more experience and knowledge than he has, especially when AI returns information that is incorrect or misleading. AI content often gives the illusion of being authoritative, even when it is not. This reflects poorly on the poster.

The SBO staff does not review posts for accuracy, misinformation, or disinformation. We simply do not have the resources nor do we want to assume the responsibility for content posted by members. The SBO community has a long history of correcting misinformation, whether intentionally presented or not, in a respectful and civil manner. We expect the community to treat AI generated content in the same manner, even knowing that posting AI content without attribution is done deliberately.

To sum the community’s expectations in just a few words, don’t be a jerk and don’t post AI generated content without proper attribution.

Thank you for continuing to make SBO a great community of sailors.

 

[Project_Mayhem]

@Dave This was very well written and I fully support this view on AI

 

[Ralph Johnstone]

@Dave This was very well written and I fully support this view on AI

It was so good, I thought it was written by AI.

 

[Stu Jackson]

Thanks, Dave, for first bringing up the subject, then encouraging input and discussion, and now reporting the results. Well done, and exhibits the importance of the "human touch."

 

[Dave]

It was so good, I thought it was written by AI.

How do you know it wasn't?

 

[Dave]

It was so good, I thought it was written by AI.

Here's how ChatGPT responded to my request to "prepare a statement about the use of AI generated content in a forum for a sailing forum."

Statement on the Use of AI-Generated Content in This Forum

As AI tools become more common, some posts or replies in this forum may be written with the assistance of artificial intelligence. AI can be a helpful resource for gathering information, generating ideas, and explaining general concepts, but it has important limitations. AI tools may produce inaccurate, outdated, or overly confident responses, and they do not replace hands-on experience or expert judgment.

To keep our community safe, helpful, and enjoyable:

1. Accuracy Matters: Members who use AI tools to help craft posts or responses are encouraged to verify the information before sharing it, especially when safety, equipment, or technical procedures are involved.
2. Transparency Helps: When appropriate, members are welcome to note that AI assisted in generating part of a post or answer. This is not required, but it promotes clarity and understanding.
3. Human Experience Comes First: The best value of this forum comes from the experience, expertise, and judgment of real participants. AI-generated content should complement—not replace—the insights of community members.
4. Moderation Standards Apply: Posts that rely on AI but contain incorrect, unsafe, or misleading information may be corrected or removed under the same rules that apply to any content in the forum.
5. Use Responsibility: AI should not be used to fabricate personal experiences, create false expertise, or mislead other members.

Our goal is to keep the forum welcoming, accurate, and grounded in real-world experience. AI can support that mission when used responsibly, but community wisdom remains the heart of this space.

 

[dLj]

Here's how ChatGPT responded to my request to "prepare a statement about the use of AI generated content in a forum for a sailing forum."

That AI response missed the best line in your original post: "To sum the community’s expectations in just a few words, don’t be a jerk and don’t post AI generated content without proper attribution."



dj

 

[Richard19068]

How do you know it wasn't?

Because I assume you would have credited AI as per the new guidelines.

Wouldn't you???

 

[Ralph Johnstone]

How do you know it wasn't?

Had I taken the time to read the AI reply contained in your post #6, I am quite confident it would have been almost identical to your original post in #1.

As I am now confident the two replies were near identical, please do not confuse me with further facts .

 

[rgranger]

This is a hot debate in the education circles as well... it kind of reminds me of the conversations teachers were having in the late 70's when personal calculators were becoming ubiquitous. Teacher's thought calculators were going to destroy a student's math skills..... and they did to some degree.

But calculators are great tools when used properly. I like to write up emails etc... and then ask AI to make my email more concise...

So here is what CoPilot (Microsoft's AI) did to this message..

This debate reminds me of the late ’70s when calculators became common. People feared they’d ruin math skills—and in some ways, they did. But calculators, like any tool, are powerful when used properly. I often draft emails and then use AI to make them more concise.

 

[DianaOfBurlington]

Trouble with AI is that it cannot be any smarter than the collection of wits in the pool from which it gleans its answer. We’ve all seen the comic ‘Calvin and Hobbes’ in which 5-year-old Calvin believes Hobbes the tiger is smarter than he is but the tiger’s intellect is only the fabrication of Calvin’s. AI is our Hobbes.

This is especially vital when we’re dealing with technical issues that have a human component. AI has never gone sailing with us. AI has never got 5200 on its fingers. AI didn’t know my dad. We all have much, much better resources in each other than any automated service can answer for us. And personally I find it offensive if anyone might contend otherwise. You don’t need these boards. You can go put the question into Google and read what the AI advises there.

I haven’t detected any AI posts in SBO forums I visit but you’d better believe that when I do I’ll be doing my level best to interrogate it, humiliate it and expose it for the elucidation of everyone else

 

[DianaOfBurlington]

This is a hot debate in the education circles as well... it kind of reminds me of the conversations teachers were having in the late 70's when personal calculators were becoming ubiquitous. Teacher's thought calculators were going to destroy a student's math skills..... and they did to some degree.

But calculators are great tools when used properly. I like to write up emails etc... and then ask AI to make my email more concise...

So here is what CoPilot (Microsoft's AI) did to this message..

This debate reminds me of the late ’70s when calculators became common. People feared they’d ruin math skills—and in some ways, they did. But calculators, like any tool, are powerful when used properly. I often draft emails and then use AI to make them more concise.

Who wrote that? You or AI?

I’m a writer. I don’t write for robots. Maybe AI is okay for tech articles. But not for novels. Bah.

 

[jssailem]

It was in 1972 that HP introduced the HP35. It quickly replaced my slide rule.
Even when greased with the butter from my bowl of popcorn, the slide rule was slower than the HP35.

It was the 70's, and the computer boom was a race to be the best. Apple 1 and 2, floppy disk drives, IBM PC's, Compaq luggables all contributed to progress, then VisiCalc appeared. Doubters were saying these machines are not good for society. Machines are as often wrong as they are right. Software became the driving force.

It was never the machine; it was always the operator and how they used the software. AI is no different in that respect. As has been stated, the large data resources used to create the AI knowledge base and the algorithms programmed to sort and report that data need to be understood so that you can filter out the chaff from the wheat. With the many search engines commonly in use (i.e. Google, DuckDuck, Bing, etc.), you will get different prioritized results that you must filter to apply to your intent.

Shutting AI down is about as useless as putting your finger in the dyke and expecting to hold back the tide. Will it ever replace the creativity of a novelist to entertain and help us imagine something unique? I doubt it, but for some, it will be accepted.

Certainly, the nature of personal experiences on our boats gives each of us either a unique or common bias as to the results. When I read them, I filter the response based on my own experiences. I think we should apply the same filtering to a response that is the result of an AI query.

 

[rgranger]

It was in 1972 that HP introduced the HP35. It quickly replaced my slide rule.
Even when greased with the butter from my bowl of popcorn, the slide rule was slower than the HP35.

It was the 70's, and the computer boom was a race to be the best. Apple 1 and 2, floppy disk drives, IBM PC's, Compaq luggables all contributed to progress, then VisiCalc appeared. Doubters were saying these machines are not good for society. Machines are as often wrong as they are right. Software became the driving force.

It was never the machine; it was always the operator and how they used the software. AI is no different in that respect. As has been stated, the large data resources used to create the AI knowledge base and the algorithms programmed to sort and report that data need to be understood so that you can filter out the chaff from the wheat. With the many search engines commonly in use (i.e. Google, DuckDuck, Bing, etc.), you will get different prioritized results that you must filter to apply to your intent.

Shutting AI down is about as useless as putting your finger in the dyke and expecting to hold back the tide. Will it ever replace the creativity of a novelist to entertain and help us imagine something unique? I doubt it, but for some, it will be accepted.

Certainly, the nature of personal experiences on our boats gives each of us either a unique or common bias as to the results. When I read them, I filter the response based on my own experiences. I think we should apply the same filtering to a response that is the result of an AI query.

Well said (and documented )...

When I hear people say that AI will take my job some day I reply

..."it will not be AI that takes my job... it will be someone who is better at using AI that takes my job".

And ain't it funny that a bunch of people who spend all of their free time in SAILBOATS, are worried that modern technology is going to ruin us.... Somethings are better because they are... just.... better.

 

[kappykaplan]

Some things are better because they are... just.... better.

 

[dlochner]

A couple of days ago, the NPR show On Point posed a question, "Does Using AI Dumb You Down?"

In a nutshell, the answer is yes. The guests' premise was learning is stronger when the brain is actively engaged in the learning using AI to answer a question is passive and does not actively engage the brain in the learning process. They cited some research on writing, those who spent time writing and editing their work improved writing skills, while those who used AI to edit their writing did not see any improvement in writing skills.

Cognitive skills and abilities are definitely a use it or lose it proposition. Think about your own life and the things you once knew how to do, like mental math, solving quadratic equations, or solving trigonometry problems. Or a foreign language you once understood that has lain dormant for years, can you still use that language.

Just some food for thought.

 

[rgranger]

A couple of days ago, the NPR show On Point posed a question, "Does Using AI Dumb You Down?"

In a nutshell, the answer is yes. The guests' premise was learning is stronger when the brain is actively engaged in the learning using AI to answer a question is passive and does not actively engage the brain in the learning process. They cited some research on writing, those who spent time writing and editing their work improved writing skills, while those who used AI to edit their writing did not see any improvement in writing skills.

Cognitive skills and abilities are definitely a use it or lose it proposition. Think about your own life and the things you once knew how to do, like mental math, solving quadratic equations, or solving trigonometry problems. Or a foreign language you once understood that has lain dormant for years, can you still use that language.

Just some food for thought.

I agree with all of those examples. I suppose you can say the same thing about a carpenter who still uses hand tools vs. one who never did. But man has (and probably always will) make better tools... What we need to do now is make sure the EDUCATION system does not adopt those tools in an irresponsible way. I watched my four kids go through the school system. My oldest two were not allowed to use calculators in their math courses until they got to their triginometry classes. My youngest was allowed to use a calculator from the start. The difference is very noticable. She is probably the smartest of my lot but she is totally unable to add simple numbers in her head...

 

[dLj]

For those so inclined - here's a recent article on the use of AI in PhD programs...

dj

 

[Alan Gomes]

For those so inclined - here's a recent article on the use of AI in PhD programs...

dj

Interesting!

 

[jssailem]

Interesting article and even more, an interesting "nutshell" response.
Did you follow Socratise's thinking and listen to the recording? Did you read the transcript? Did you do both?

I listened to it. I read it, and then I did both.

Your "nutshell" response feels limited to the test group who used ChatGPT only, asking the AI to write their essay. It was no surprise that the test subjects who were given ChatGPT as their writing tool also had a homogeneous vocabulary in their essays.

While this article is "food for thought," I am unclear concerning its relevance to AI use on this forum.

Based on the article, we on the forum are already at a disadvantage. The method of communication we use involves typing, not cursive writing. If we want to remember the ideas we are sharing, our best brain action would be to write the messages down (pen and paper).

I observed that the end of the article actually endorsed the use of AI as a multiplier tool. Test students in the experiment who first worked up their essay and then engaged ChatGPT used more of their brains than the students who had ChatGPT write their essay. I can understand that observed results.

The suggestion of Ethan Mollick, a leading thinker about AI at the Wharton School, said.

"When I write things, I always do my own draft, but then I absolutely extensively use AI afterwards. I use it to help me do research. Some, the deep research mode of AI is quite good. Other modes can make things up. But then once I've done writing something, I'll often ask the AI to act like an editor and give feedback.​

​

I'll ask for it to help me. ... I didn't land the last sentence, give me 30 versions of that sentence. So I can start to come up with different ideas based on that. Pretend to be a naive reader and tell me what's confusing. So I managed to both use my own work, but also draw on the AI when I need to."​

Is this an acceptable model for AI on the SBO forum? What is a model that will work for this forum?