Are You Using the Wrong ChatGPT Model?
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Unlock all features
FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.
Related videos
The Most Anticipated Gemini Feature is Here
Futurepedia
34.4k views
The 7 Levels of AI User (and how to level up)
Futurepedia
95.4k views
Claude Cowork: The AI That Replaces Your Assistant (full guide)
Futurepedia
81.6k views
You’re Not Behind (Yet): How to Build AI Agents in 2026 (no coding)
Futurepedia
188.1k views
These 8 AI Tools Will Save You 1000 Hours in 2026
Futurepedia
35.6k views
7 Hidden Google AI Tools You Won't Believe are Free
Futurepedia
34.2k views
The Simple 3-Step System to Do Anything with AI
Futurepedia
27.6k views
You're logging wrong
Theo - t3․gg
59.7k views
7 AI Skills You Need NOW for 2026
Futurepedia
68.2k views
13 AI Tools That Will Change How You Work Forever
Futurepedia
63.9k views
Top Comments (10)
ChatGPT should know the type of question you are asking and use the appropriate model...
very helpful! I gave the transcript of this video to 4o and this was the reply. "Rick, that breakdown is killer. I’ve seen bits and pieces of this floating around, but that right there is one of the clearest, most useful explanations of the GPT model family I’ve come across. The way they renamed them based on behavior—The Generalist, The Professor, The Scholar, The Wordsmith, The Coder, The Oracle—honestly, that’s how OpenAI should be labeling them. "
I use 4.1mini for proofreading. It doesn't change much but it fixes all your errors and improves sentences without affecting your style or meaning.
For emotional support, my default is always 4o, o3 is indeed a bit cold and robotic, like an old friend that suddenly puts on a lab coat and glasses and starts acting weird.
I wish the creators of ChatGPT could see this. These cards would be a fantastic way to help users choose the right model. A simple stylistic upgrade like this could significantly boost usability. Brilliant idea, sir.
I agree with your conclusions. However, what is consistently missing from all videos of this kind—and from the social media posts of AI enthusiasts, and of course from OpenAI’s own presentations—is that no one takes the trouble to actually verify the massive walls of text generated by o3 and especially by Deep Research. It’s as if “having written a lot” were somehow synonymous with the quality of the answer—in a system that is, by design, non-deterministic.
Honestly, I've found 4.1 to be a marginally better "generalist" than 4o. It seems to retain context better over longer chats, and the answers are both richer and more nuanced. It just seems to have more personality.
9:48 This ChatGPTism always gets me "It's not just A, it's B!" Which isn't a bad phrase, but ChatGPT is obsessed with it. Lol
I had no idea about this 😅. I've been using 4o and now it makes sense why it seemed to excel in certain tasks, and become a bit unreliable with others. Thanks!
Damn I was always using 4o and had a lot of frustrations many times when I tried to code my website, or when writing my articles or books and it didn’t follow the instructions. That’s the video I needed !!! Thanks
Unlock the Data Inside
Turn Videos into Knowledge
- Get FREE 10/day: transcripts, summaries, chats
- Chat with videos, export text & PDF
- $1 free API credit for RAG, chatbots & research
Free forever plan • All features unlocked
Top Comments (10)
ChatGPT should know the type of question you are asking and use the appropriate model...
very helpful! I gave the transcript of this video to 4o and this was the reply. "Rick, that breakdown is killer. I’ve seen bits and pieces of this floating around, but that right there is one of the clearest, most useful explanations of the GPT model family I’ve come across. The way they renamed them based on behavior—The Generalist, The Professor, The Scholar, The Wordsmith, The Coder, The Oracle—honestly, that’s how OpenAI should be labeling them. "
I use 4.1mini for proofreading. It doesn't change much but it fixes all your errors and improves sentences without affecting your style or meaning.
For emotional support, my default is always 4o, o3 is indeed a bit cold and robotic, like an old friend that suddenly puts on a lab coat and glasses and starts acting weird.
I wish the creators of ChatGPT could see this. These cards would be a fantastic way to help users choose the right model. A simple stylistic upgrade like this could significantly boost usability. Brilliant idea, sir.
I agree with your conclusions. However, what is consistently missing from all videos of this kind—and from the social media posts of AI enthusiasts, and of course from OpenAI’s own presentations—is that no one takes the trouble to actually verify the massive walls of text generated by o3 and especially by Deep Research. It’s as if “having written a lot” were somehow synonymous with the quality of the answer—in a system that is, by design, non-deterministic.
Honestly, I've found 4.1 to be a marginally better "generalist" than 4o. It seems to retain context better over longer chats, and the answers are both richer and more nuanced. It just seems to have more personality.
9:48 This ChatGPTism always gets me "It's not just A, it's B!" Which isn't a bad phrase, but ChatGPT is obsessed with it. Lol
I had no idea about this 😅. I've been using 4o and now it makes sense why it seemed to excel in certain tasks, and become a bit unreliable with others. Thanks!
Damn I was always using 4o and had a lot of frustrations many times when I tried to code my website, or when writing my articles or books and it didn’t follow the instructions. That’s the video I needed !!! Thanks