ChatGPT Plus SUCKS! What You Need to Know

Three main reasons why ChatGPT Plus falls short
00:00 Intro
00:16 What even is ChatGPT Plus
01:52 Limitations of ChatGPT Plus
01:56 Rate Limiting Issues
03:02 Model Selection Challenges
03:55 Simpler Tasks and Model Switching
04:39 The Solution
05:56 Conclusion
#ChatGPT #AI #ChatGPTPlus #TechReview #GPT #AIFeedback #TechCritique #AIBenefits #AIImprovements #TechSolutions #AIReview #Chatbot #OpenAI #TechTalk #Innovation #AIAdvancements #TechCommunity #GPTInsights #SmartTech #AIApplications

[@gdedgar]

I get what you’re saying and you make good points, but I think this is ultimately a value decision.
Maybe I’m not your audience, but I have a great time using chatGPT every day and the only time I run into a rate limit is if I spend a whole evening working on something like a world building for a story. It’s not infinite, but I wouldn’t say it sucks.
I don’t mind using developer tools, so I’ll check out your next video to see your perspective better.

[@MrDeSheikh]

It sounds like you’re able to get the desired outcome under the rate limit. Your prompts must be good.
I agree, for quick small questions, its fine but for heavy use-case, iterating multiple times/ using pseudo agents, we can easily hit the rate limit.
Just depends on what we use the models for :D

[@Iammzephyr]

Honestly, I have encountered the same issues with 3.5 as well. It sucks to know that after paying 20 dollars/ month you still face the same " Ohh Pay Me More Money Issue". What are your thoughts on using other AI models that are powered by GPT 4?

[@MrDeSheikh]

Yepp. Unless someone mentions that they fine-tuned a model or made a custom one, under the hood its the regular gpt-4 model which you are already using. The limitations still apply only just wrapped up in a specific prompt/system message.
---
Do not pay for any AI services. Whatever they are offering, you could do it yourself if you want to save money unless you want the connivence :D
Also checkout using Claude 3.5 Sonnet. It maybe better than gpt-4

[@mathiasmuggli1162]

if you want the model to write programming code, gpt 4 is MUCH mor capable in my experience...

[@MrDeSheikh]

Yes. I agree. For coding i never use gpt 3.5. Always the 4. Interestingly, majority of the time I get better results with gpt-4-turbo than the brand new gpt-4o

[@iseptimus]

Don’t have an issue. Even with massive data sets.

[@MrDeSheikh]

Nice! Curious to know what sort of datasets are you working with?

[@RekdReation]

You need to understand how to ask the right questions. GPT is not "smart". It is not "AI". It is an LLM. A scrubber. A bottom feeder. It knows almost everything. You just need to know how to ask. BTW, 3.5 is good for most code and basic queries.