2 ms·

It’s not that difficult to tell when output for prompts (on a paid account) changes consistently from: “I’ve understood and performed the task you’ve requested

It’s not that difficult to tell when output for prompts (on a paid account) changes consistently from: “I’ve understood and performed the task you’ve requested. Let me know if this comprehensive annotated output is correct and if there’s any additional modification you would like to make.” to “In order to do what you’ve requested, you will need to do and consider the following vague high-level steps in a numbered list. Feel free to ask me to do it, but I’m going to spend the next dozen responses apologizing for not following a simple, explicit, unambiguous instruction, saying I’ve corrected the mistake, then making the exact same mistake over and over until I give up and claim it’s too complicated while throwing Error Analyzing messages that don’t need to be shown to the user.”

yacine_ 3y ago
I'd imagine you feel quite upset that you are paying to be silently toyed with by a software company :P I definitely do Infuriating and is only accelerating my basement cluster
soulofmischief 3y ago
Nailed it. This is just getting infuriating: https://chat.openai.com/share/03069d31-06d3-42e2-bfdc-ff6eaf6112d6 https://chat.openai.com/share/03069d31-06d3-42e2-bfdc-ff6eaf... GPT-4 went from being absolutely awe-inspiring to next to useless. This is the kind of response I expect from GPT-3.5-Turbo, not GPT-4. Comparing the first offered code solution to the last offered solution, it's just insane how over-complicated it made things. It used to not be this way. The cream on top is that at the end it offered two different apologies and asked me to select which one I preferred more. The lack of transparency behind these changes, and the gaslighting around the clearly obvious changes to the system's output are just ridiculously patronizing, and I will run for the hills the moment a company which actually values its users creates a competitive product.
- mdekkers 3y ago
  I suspect they are slowly making 4 worse so that when they launch 5 it will appear so much better than 4

yacine_ 3y ago

I'd imagine you feel quite upset that you are paying to be silently toyed with by a software company :P I definitely do Infuriating and is only accelerating my basement cluster

soulofmischief 3y ago

Nailed it. This is just getting infuriating: https://chat.openai.com/share/03069d31-06d3-42e2-bfdc-ff6eaf6112d6 https://chat.openai.com/share/03069d31-06d3-42e2-bfdc-ff6eaf... GPT-4 went from being absolutely awe-inspiring to next to useless. This is the kind of response I expect from GPT-3.5-Turbo, not GPT-4. Comparing the first offered code solution to the last offered solution, it's just insane how over-complicated it made things. It used to not be this way. The cream on top is that at the end it offered two different apologies and asked me to select which one I preferred more. The lack of transparency behind these changes, and the gaslighting around the clearly obvious changes to the system's output are just ridiculously patronizing, and I will run for the hills the moment a company which actually values its users creates a competitive product.

mdekkers 3y ago
I suspect they are slowly making 4 worse so that when they launch 5 it will appear so much better than 4