Page 1 2
Go | New | Find | Notify | Tools | Reply | |
| Member |
I use Perplexity as a research tool. I chose it because it passed my inquiry test, asking complex questions that I already knew the answer. If response was outright wrong, quoted sources such as NYT, TV News etc. I eliminated that AI! Even with Perplexity, I specify using primary sources only: scientific journals, news releases from competent sources, verifiable court cases etc. I like the fact that, in the body of the reply, it gives a hot link to the source document cited! If I have any concerns whatsoever, I pull up the citation and read the article. The AI looks at the purpose of papers, then the conclusions drawn - it does not read the body of fact and check that the data contained therein supports the author’s conclusions!! Example - the D.C. Mayor wanted a dusk till dawn curfew to reduce criminal activity… the City Commission only allowed 11:00 pm until dawn curfew to be put in place, claiming the curfew made no difference. A Social Sciences paper defended their position. That did not pass my smell test! I read the paper; indeed, the number of criminal acts remains the same with 11-7 curfew. What they DID NOT reveal was the crimes simply shifted to the dusk to 11:00 pm time frame. The authors LIED BY OMISSION**! AI is a tremendously powerful research tool - it does not substitute for the human mind! Used as a guide, a research tool- great! As an end point for drawing your conclusions, not so much! **i suspect the authors were influenced by political beliefs, being from one of our Leftist elite universities. No quarter .308/.223 | |||
|
Prepared for the Worst, Providing the Best |
I 100% agree with this. Typically when I ask a question here, it's out of a desire to hear the personal experience and opinions of the members. The banter and discussion between members who I've never met but feel like I've grown to know over the past 20 years helps me form opinions on issues and make decisions in ways that raw technical data from an LLM can't. If I need a torque spec for a bolt my first stop is Google, and typically I get what I'm looking for. But the past couple of months I was looking to buy a motorcycle and spent a lot of time doing my research beforehand. Google and AI could give me all the technical specs and links to videos of sponsored content creators with questionable motives. Sigforum gave me real-world experience from people I trust, and led to an outcome that I'm very happy with. Occasionally I see people on here regurgitating AI results in threads. If it's an answer to a technical question then I guess it has value, but also makes me question why the OP didn't just go to AI in the first place. But when it's done in response to a request for an opinion or personal experience, I question why the AI poster didn't just stay out of the thread if they didn't have anything personal to add, because that response isn't offering the value that the OP is looking for. AI bots are another thing altogether, and I'm grateful that Para has been on top of culling them. I imagine that is going to become harder and harder to do as they become more prevalent and developed. ----------------------------------------------------------- Any comments made by this poster are my own and do not reflect the views or opinions of my employer. | |||
|
| Savor the limelight |
181 lb-ft or 242 lb-ft? One is AI’s answer while the other is from the service manual. 2012 Honda Odyssey CV axle nut torque? | |||
|
Political Cynic |
I’m not a fan of AI and in my narrow opinion I think it’s a huge step along the path of dumbing down the species. We’ve already given up trade schools and lowered standards for sciences and engineering. I really don’t like where this is going. It’s not that I don’t like change, I am a huge supporter of change but when it’s for the betterment. We are going to have a reckoning soon and AI is going to cost lives with a bridge or a building because someone wants to take a short cut or doesn’t want to do the real work. I hope we can find a way to ultimately filter and exclude AI from social media and more importantly from news feeds | |||
|
| Do No Harm, Do Know Harm |
The only other board I really ever spend any time on is Quora. On that one I just scroll for mindless things that interests me, generally history topics. It’s so full of AI crap using AI pictures that it has lost me. I don’t fear that here. Not a fan of the AI search answers though. I feel like I can always tell by the tone that it was generated through one of those systems. Knowing what one is talking about is widely admired but not strictly required here. Although sometimes distracting, there is often a certain entertainment value to this easy standard. -JALLEN "All I need is a WAR ON DRUGS reference and I got myself a police thread BINGO." -jljones | |||
|
| Raised Hands Surround Us Three Nails To Protect Us  |
I also hate when folks say AI Source says… then add really no further info. I would love for the post to have the thoughts of the poster and then at the bottom bring in the AI source. If someone leads with an AI source I skip that post. My wife uses it for everything. It’s wild how often it is wrong. I have a shirt that is 25 years old and it is finally ready to call it a day. I want to remake the shirt as it is just simple text on a dark green shirt. I asked my wife to ask ChatGPT what font the text was. The answer Chat gave was not even close. Not even remotely similar. I ended up scrolling through the font list in Word and found the right one. ———————————————— The world's not perfect, but it's not that bad. If we got each other, and that's all we have. I will be your brother, and I'll hold your hand. You should know I'll be there for you! | |||
|
| No More Mr. Nice Guy |
As have electronic calculators. We had to do math by hand until high school, when we learned to use a slide rule. The slide rule augmented one's understanding of how math worked, plus sped up calculations. A circular slide rule gives visual representation of trigonometry while solving problems, so it too augments understanding. My kids had graphing calculators in school. A very cool tool but it bypassed learning how equations become graphs. Likewise, the internet is a nice tool to speed up getting an answer, but at the cost of learning what is behind that answer. One might argue that the average human doesn't need to know any deeper than what time the two trains pass, but that definitely brings everyone down to lowest common denominator (and they won't even know what that term refers to). | |||
|
| probably a good thing I don't have a cut |
I had an argument with the Duck.ai on Duck Duck Go yesterday. I asked it who the congressional candidates were for my district and it only returned the Democrat candidate. I asked if she was running unopposed and then it told me the Republican candidates name. I told it that it was trying to influence my vote by only telling me who the (D) candidate was. It denied it and said it could provide all the filing information for the candidates. I asked how that proves you weren't trying to influence my vote by only providing the one candidates name. It had no answer but could provide me the filing information. It got ugly. We're not talking right now. | |||
|
| Thank you Very little  |
Questions here, and input from interested but not so experienced parties in threads have always had input from internet searches. The fact is you can't get away from the AI input on searching since all the major search engines use it and provide a quick summation of the information they have found. Not all AI answers are good for a forum response, but they can lead you to better understanding on where to research and can provide good information, you just have to understand that. It will take a bit of time for people to understand that. Just for kicks, wondered how ChatGPT would answer the OP's question that is being discussed.  it would be interesting to see how it compares on Grok, but I don't have a subscription plan and Grok was "busy" ie gimme money.... The Question as posed to AI: summarize how will ai negatively impact discussion boards in the future with an elevator presentation format Elevator Presentation: How AI Could Negatively Impact Discussion Boards Opening (15 seconds) Discussion boards have long been places where people share experiences, ask questions, and build communities. As AI-generated content becomes widespread, these platforms may face significant challenges that affect trust, participation, and information quality. Key Concern #1: Loss of Authenticity (20 seconds) AI can generate convincing posts, comments, and responses at scale. Users may no longer know whether they are interacting with real people or automated systems, reducing the sense of genuine human connection that makes discussion boards valuable. Key Concern #2: Information Overload (20 seconds) AI can produce large volumes of content quickly. Discussion boards may become flooded with repetitive, low-value, or automatically generated posts, making it harder for users to find useful information and meaningful conversations. Key Concern #3: Increased Manipulation (20 seconds) Organizations, advertisers, political groups, or bad actors could use AI to influence discussions, amplify certain viewpoints, or create the appearance of consensus. This can distort public opinion and undermine trust in online communities. Key Concern #4: Declining Participation (20 seconds) When users suspect that many posts are AI-generated, they may be less motivated to contribute their own experiences and expertise. Communities thrive on personal knowledge, and reduced participation can weaken the overall value of the platform. Closing (15 seconds) The future challenge for discussion boards will not be AI itself, but maintaining trust, authenticity, and meaningful human interaction. Platforms that successfully identify AI content, verify contributors, and encourage genuine engagement will be better positioned to preserve healthy online communities. One-Sentence Takeaway: AI could transform discussion boards from communities built on human experience into spaces where authenticity, trust, and meaningful participation become increasingly difficult to maintain | |||
|
| Peace through superior firepower  |
Same question posed to Grok, and if you read through this and don't see the value of its perspective (plus some interesting trivia) then, you just don't want to see it: Yes, AI is already severely impacting discussion boards, and the effects are likely to intensify, though the outcome depends on the platform, moderation, and user behavior. It's not a total collapse yet, but authenticity, engagement, and signal-to-noise ratio are declining in many spaces. Key Ways AI Is Disrupting Discussion Boards Counterpoints: Not All Doom Some platforms adapt or benefit: Likely Future Trajectory In open, low-moderation boards: Severe degradation—more noise, less value, potential "ghost towns" of bot conversations. Closed or well-moderated ones: Adaptation and survival, possibly with AI-assisted features (e.g., AI helpers that disclose themselves).This mirrors broader internet trends: AI amplifies scale but dilutes soul. Users are already voting with their feet toward smaller, verifiable groups. The core challenge is preserving spaces where real people connect, think, and disagree—without everything turning into echoey AI slop. Platforms that solve detection, incentives for humans, and transparency will fare best. ___________ Notice this tidbit: And, AI moderators?? I can just see it: Me: Reopen the thread on Best Value in Pod Bay Doors, HAL. HAL9000: I'm sorry, Dave. I'm afraid I can't do that. Me: Who the fuck is Dave? HAL9000: Dave, this conversation can serve no purpose anymore. Goodbye. (bans me) | |||
|
| Freethinker |
Like it or not, any search engine like Google or Bing is an artificial intelligence source. If I want to know who is credited with the definition of insanity that’s so often quoted on Internet forums, how does Google give me an answer? It’s not because the question shows up on the computer screen of some guy sitting in a dark room someplace who then pulls down a book of quotations to try to find the answer and reply back in a second or two. The artificial (not human) intelligence process is to search through sources that are accessible on the Internet and then respond back with links to sources, assuming it found an answer among those sources. Google searches are artificial and intelligent-like because they must sift through literally countless sources to try to find something that might be relevant. No genuine-intelligent humans involved. “But, but, but … what about mistakes?” The things (some, anyway) that are called AI today readily admit that their responses can be mistaken. Something like a Google search, however, doesn’t, and of course that’s because it’s only providing links to sources, not attesting to their accuracy. So, what if I ask Google about the “rough men” and sleeping well in the night quotation that we see frequently? It will direct me to things like lists of quotations or other places it appears. Is a list necessarily correct? What if I’m directed to a post in a discussion forum? I go there, see who posted the quotation, and now it’s up to me to decide whether it’s correct. “Oh, yeah: I found it on the Internet so it must be true.” What of course is now different with an “AI” source like Copilot or Gemini is that because it doesn’t just link to other sources, it may add a comment like, “That quotation is often attributed to George Orwell, but there is no record he ever said it. The earliest record of its having been stated was by Jane Doe, a journalist for XYZ news in an article she wrote for Pacific Monthly in 1972.” A response like that, however, spooks some people because they may forget that it’s still just an Internet search result, albeit one that extracts information and presents it in a conversation-like manner. AI has been with us for a long time. ► 6.0/94.0 “I can’t give you brains, but I can give you a diploma.” — The Wizard of Oz | |||
|
| Thank you Very little |
With AI's approach t scans all internet sources for answers and everything on the net isn't true of course, so it can provide false answers based on the false data online. Surely they are working to correct that, or maybe, AI is working on AI to correct AI.... Shall we play a game? For now Humans can figure out AI from Human responses at least now easily. The AI videos are for now pretty easy to spot, pictures as well, who know how much self learning it will have to improve the software. Open the pod bay doors, Hal...... | |||
|
Hop head  |
gunboards has it's own AI member, don't recall the name but it was activated a year or so ago, it is called forafrank, https://chandlersfirearms.com/chesterfield-armament/ | |||
|
A Grateful American |
I asked AI how World War Eleven started and it said it was a result of the Germans bombing Pearl Harbor. Seems legit. "the meaning of life, is to give life meaning" ✡ Ani Yehudi אני יהודי Le'olam lo shuv לעולם לא עוד | |||
|
| Political Cynic |
Sounds about right. | |||
|
Member |
My extensive use of AI is focused on challenging questions, software scripts and wide-ranging inquiries. I find the performance of the paid tiers to far outclass the free. I enjoy the totally fake "dialog" as if it is a person, but that makes communication simpler. For discussion boards, and social media groups, I admit to frustration that simple questions are posed of a group, instead of AI. For example, pictures of birds, insects, snakes, foxes, bobcats, etc. are posted as if this is an incredibly difficult ID to make. And, the presumption is that humans can ID the object. I am interested in what a person believes, thinks, opines, and counsels. For that need discussion boards will serve well. ------- Trying to simplify my life... | |||
|
| No More Mr. Nice Guy |
I enjoy the identification discussions as much for the discussion as for the answer. If someone makes an incorrect identification and is corrected, I learn even more. People post their humorous or un-humorous experiences. If the subject is an object such as an old mechanical part, there is usually an interesting series of stories about great-grandpa MacGyvering old machines. AI can answer how to accomplish a repair or maintenance task in a sterile manner, but people share little tricks or warnings learned with experience that AI doesn't know. I've learned though to avoid the simple questions on other boards, such as which brand of guitar strings to use, or can a different weight of oil be used in a particular application. Not here, but on several other boards those are AI generated posts created by the forum owner to generate interaction and, most likely, boost ad revenues. I can hardly stand to visit some of my previously favorite forums. | |||
|
| my existence, while grotesque and incomprehensible to you, saves lives |
So they got me Claude AI at work. To test it, I asked it a moderately difficult question, “how do you convert a P226 SAO to short reset?” After having its answer corrected twice, it refered me to sigforum. Seriously, the entire chat is below: This message has been edited. Last edited by: car541, ***************************** "I don't own the night, I only operate a small franchise" - Author unknown | |||
|
| Member |
I found Claude making up stuff again. I hate when that happens. I asked for the classic fingerings for a three-octave g-scale played on the violin from the g-string. It created utter nonsense. Listed the entirety of lettered notes, from G to G to G. Listed them organized by violin strings. But, the rest of it was utter garbage. I explained its shortcomings most bluntly, and it replied that AI "has a specific weakness with things like fingering charts, tablature, and the detailed physical technique notation -- content that exists heavily in books, pdfs and videos rather than clean indexable text." So, at least AI admitted it was wrong, apologized, and explained itself. ------- Trying to simplify my life... | |||
|
Told cops where to go for over 29 years… |
AI certainly has its place. My son is an Operations Manager at a county level 911 agency. He used Claude AI and some other tools a bit over a year ago to come up with AI handling of non-emergency calls. That project partnered him with the Aurelian company and he now has a side hustle with that company. Fantastic results, including other agencies signing on to use the “product” and he has now been involved in developing an AI tool to assist call receivers handling 911 calls. I’m pretty proud of the kid… He followed in my foot steps as Call Receiver at the age of 18, then dispatch and Dispatch Supervisor. Picked up a Criminal Justice degree and moved to the management side of the center. Has been a self-taught techie and worked his way into getting it added to his job description, and now there are making a new tech specific manager position tailored to him. Here is an article about the tech: https://www.msn.com/en-us/news...11-calls/ar-AA1X0fA8 Not sure about it for “interacting” on discussion boards, but as others noted it is great for getting data and more useful info than simple internet searches. My first use of AI was using ChatGPT to “help” write Amazon reviews (I’m in the Vine program, free stuff for review). I give it the link and tell it write a good, mediocre, or negative review based on three or four key points and it spits out a review for me to tweak and post. Most recently I used ChatGPT to come up with a math equation to calculate power stats for my e-bike rides. The app for my new bike doesn’t have as much info as my old bike’s app did. My old bike would report my power output along with how much power was contributed by the motor. New bike didn’t give any power data at all. I know there are workout apps like Strava the give you an estimated power output so I I figured there had to be a way to calculate for it. I already track my ride with Apple Watch and the bike app and didn’t want to bring a third app into the mix. Figured I cold take my available data, plug it into an Excel spreadsheet and get the power numbers. I was impressed with the way the “conversation” went. I told it want I wanted it to do and what variables I had. We went back and forth a bit tweaking things like bike type, tire type, etc. I tried the formula it suggested, and it didn’t work. I told it what my answer was compared to what it came up with and it discovered its own error (a constant of .000376 vs .376 for rolling resistance), it apologized and corrected. Did this a second round and after correcting itself a second time (constant of .043 instead of .43 for aero drag), it was fixed and gave me a formula formatted to use in Excel. Definitely came to the answer faster and easier than I would have through a simple search. What part of "...Shall not be infringed" don't you understand???  | |||
|
| Powered by Social Strata | Page 1 2 |
 | Please Wait. Your request is being processed... |
|
© SIGforum 2026