In my case, ChatGPT Plus kept telling me how insightful and savvy my questions were. I told it to stop attempting to flatter me and just give me the facts. It did, but it still gives me answers in a more colloquial phrasing than I prefer. That's probably also on me because I've taught it how to write book chapters for me so it uses the same colloquial manner when responding to my questions.
I hope you're not claiming authorship for what AI wrote.
this is typical.... inviting agreement and bate utter bs offering.
I asked ChatGpt is rojo the greatest running journalist of all time? he was an olympian and record holder
Chatgpt says: When people ask whether Rojo is the greatest running journalist ever, they’re usually talking about Robert Johnson — co-founder of LetsRun.com — who’s often nicknamed “Rojo.” Who Is Rojo?
Former competitive runner 1996 U.S. Olympic Trials steeplechase finalist (often loosely referred to as an “Olympian,” though he did not make the Olympic team) 2:09 marathoner (elite-level performance) Co-founder and longtime voice of LetsRun, one of the most influential (and controversial) running media platforms Is He the Greatest Running Journalist Ever? That depends on how you define “greatest.” ✅ Case For Rojo Impact & Influence: LetsRun has shaped elite distance running discourse for over 20 years. Insider Knowledge: As a high-level competitor, he understands the sport deeply. Longevity: He’s covered multiple Olympic cycles and eras. ❗ Case Against He’s more of a commentator/opinion leader than a traditional journalist. The site is known for provocative takes and controversial moderation. Some would argue writers from more traditional outlets have done more investigative or narrative journalism. Who Else Is in the Conversation? If we're talking “greatest running journalist,” names often mentioned include: Kenny Moore – Olympian and longtime Sports Illustrated writer; widely respected for literary sports journalism. Amby Burfoot – Boston Marathon winner and longtime editor at Runner’s World. Writers from outlets like Sports Illustrated or Runner’s World who focused heavily on storytelling and reporting. Bottom Line Rojo is unquestionably one of the most influential figures in modern distance running media. Whether he’s the greatest running journalist of all time depends on what you value more: Influence and bold commentary? → Strong case for Rojo. Traditional journalism and literary sports writing? → Others might rank higher.
Chatgpt was further inspired , offers ....Got an idea? Try one of our new curated styles and filters or imagine something from scratch. Try now Caricature Trend Lunar New Year Flower petals Gold Crayon Paparazzi Clouds Department photoshoot Camcorder Neon fantasy Norman Rockwell Iconic Post-rain sunset Iridescent metal portrait Sketch Dramatic Plushie Retro anime Baseball bobblehead Doodle 3D glam doll Sugar cookie Fisheye Inkwork Pop art Ornament Art school
In my case, ChatGPT Plus kept telling me how insightful and savvy my questions were. I told it to stop attempting to flatter me and just give me the facts. It did, but it still gives me answers in a more colloquial phrasing than I prefer. That's probably also on me because I've taught it how to write book chapters for me so it uses the same colloquial manner when responding to my questions.
I hope you're not claiming authorship for what AI wrote.
That was sloppy writing on my part. I should have used the word "edit" instead of write. What AI does best for me is to take complex ideas and make analogies so they are easier to understand, although it might take several attempts and the final would then need to be edited by me. It can also change tone. My tone and my authors' tones are sometimes quite different. AI can edit my prose into a tone closer to what my author prefers. Most professional nonfiction ghost writers use multiple tools with varying degrees of AI, such as Grammerly, ProWriting Aid, and specialized AI tools. In fact, there are AI programs to evaluate articles for AI-created content. But that's not pertinent to Rojo's question.
My point in the above excerpt from my previous post was that AI remembers your questions and your follow-up questions and LEARNS from your comments. Assuming he uses the same AI and it has a memory function, it will remember his previous comments and change its response style accordingly. It won't hold a grudge and give him false information on purpose.
you can talk to it in a way you never ever would speak to a human and will continue to be *extremely* helpful. i just asked if i ever have used invectives in the past and i couldn't post the response here. i'm in the top 5% of users, a paying user, using it daily for just about everything. probably used some interjections when trying to get it to write some code for me. called it stupid or a whole lot worse.
I am going to attempt to answer rojo's question seriously, for what it's worth.
You didn't mention the AI system you were using, or whether you were logged in. Those details matter.
From the perspective of the LLM (as another user pointed out), there is no memory. The model is just taking your input along with the context provided to it and predicting a next token.
However, chatbots are more than just direct interaction with an LLM. They will have a system prompt giving the LLM guidance on how to act, along with a fair bit of context the chatbot developers have about you. They may add some notes about topics you like (Running), your profession (journalist), things you hate (Joe Biden), etc. That context is probably shaped by responses you've given positive or negative reinforcement to in previous sessions. This extra context isn't really visible to you in most chat bots.
Beyond that, all messages in your current chat session (user or assistant messages) are generally included. You might be surprised to know that user messages with a certain tone are more likely to lead to assistant responses in a similar tone.
Beyond that, all messages in your current chat session (user or assistant messages) are generally included. You might be surprised to know that user messages with a certain tone are more likely to lead to assistant responses in a similar tone.
There is evidence for this. But to be clear, this is not the chat bot or LLM forming an opinion, of getting sentient. I guess the most basic way this can be explained is this: large language models convert massive amounts of text into numbers (vector database), and analyze how that text is used over and over and over. Based on this analysis, it groups together the numerical representations of text that is found to have common meaning (clusters).
When you ask the chat bot a question, it essentially “predicts” the answer based on how similar or exact questions were asked from the text it analyzed. [NOTE: Mechanically it works abit differently, but that is the end result.] [NOTE NOTE: You can ask the same question twice and get different answers; this is ON PURPOSE. The LLM chooses words in its answer based on percentages that the particular word is the “right” one; it may choose any of the words that have the highest percentages of being “right.”]
My assumption is that the text these LLMs analyze show higher correlation of polite questions coming back with more complete and higher quality answers. While the exact formula of how AI models predict a particular piece of text in answer to a particular question is not something that is knowable for a couple of reasons, it’s very unlikely that the accuracy of the answer would in any way be tied to how politely the question was asked.
Google invented LLM in 2017. Eliza chat software has been around since 1964. Browser AI used "Pre-trained" LLM models, the "P" in "Chat GPT" for low costs. It doesn't learn or update the LLM model with your input.
