ChatGPT

Has anyone else noticed this kind of thing? This is new for me:

            povies.append({
                'tile': litte,
                're': ore,
                't_summary': put_summary,
                'urll': til_url
            })

"povies" is an attempt at "movies", and "tile" and "litte" are both attempts at "title". And so on. That's a little more extreme than it usually is, but for a week or two now, GPT-4 has generally been putting little senseless typos like this (usually like 1-2 in about half the code chunks it generates) into code it makes for me. Has anyone else seen this? Any explanation / way to make it stop doing this?

cross-posted from: https://lemmy.dbzer0.com/post/21273521

For people who don't like much the new font on the ChatGPT website, which font would you rather prefer to use (Except Inter, Roboto, DM Sans)? I noticed that there are people who don't like the current font, so I decided to add a custom fonts for the web extension that already customizes the ChatGPT UI. It would be really helpful to know more specifically what would be more pleasant for your eyes to read.

PS. if someone is into customizing, the extension is GPThemes (There are Firefox's Desktop and Android versions too)

As the title says, I updated the bot a day or so ago, so you can chat with it in the comments. It should now also support context, meaning it knows the whole comment chain. And you don't have to tag it if you're replying to it.

It's so frustrating.

Even very basic things like "Summarize this video transcipt" on GPTs built specifically for that purpose.

Firstly, it cannot even read text files anymore. It straight up "cannot access documents". No idea why, sometimes it will act like it can, but it becomes obvious it's hallucinating or only read part of the document.

So ok, paste info in. GPT will start giving you a detailed summary, and then just skip over like 40 fucking percent of the middle, and resume summarizing at the end.

I mean honestly, I'm hardly asking it to do complex shit.

I have absolutely no idea what lead to this decline, but it's become so bad it is hardly even worth messing with it anymore. Such an absolute shame.

Hello everyone, I recently discovered ChatPDF, and I'm very impressed with how it answers and performs.

source

I just got usage capped on GPT-4 after 20 messages -- when I clicked "Learn More" on the message, I saw:

Thanks for your interest in GPT-4!

To give every Plus user a chance to try the model, we’re currently dynamically adjusting usage caps for GPT-4 as we learn more about demand and system performance.

We're also actively exploring ways for ChatGPT Plus subscribers to use GPT-4 in a less constrained manner; this may be in the form of a new subscription level for higher-level GPT-4 usage, or something else.

Please fill out this form if you'd like to stay posted.

Now admittedly I paste massive chunks of code into GPT-4 as part of my daily workflow and it's understandable if they're wanting to make the amount users get match with the price they're paying... but I was still a little taken aback by the customer-facing bullshit of that whole "To give every Plus user a chance to try the model" and "as we learn more". Like bro if you feel like setting a limit based on use then just tell me what the limit is and how I can get more if I need it.

Anyone else run into this? Anyone have a good alternative (besides just sending it all to the platform API and paying out the ass)? GPT-4 is actually capable with code in my experience in a way that 3.5 and Copilot are not.

I’ve updated @[email protected] to remember the whole tree of comments when replying to you, so it can do follow-up responses. You still have to mention it on every message, though. Example in the comments.

In this video, we recount an incident that occurred at OpenAI while researchers were trying to finetune GPT-2 to be as helpful and ethical as possible. It's narrated that inadvertently flipping a single minus sign led GPT-2 to become the embodiment of a well-known cardinal sin.