#gpt-models

1 messages · Page 227 of 1

dusty snow
#

Anyone know how ChatGPT can read local PDF file? If not, what are some ways I can do to resolve this problem?

knotty grotto
dusty snow
leaden cape
#

dropbox

dusty snow
#

Ok I will look into dropbox

harsh oar
#

I have plugins including the Talkwithpdf but no + or anything to upload a pdf. What am I doing wrong?

knotty grotto
harsh oar
knotty grotto
#

No problem have a good day :3

gentle pebble
#

I saw a tik tok saying if you pay for the sub gpt4 has a beta feature to actually be able to use the internet, look at images etc now. Is that true?

cyan palm
#

The rest is true

knotty grotto
cyan palm
#

Just use simple English bro

#

Me no speaking Angles senor

#

Blue collar worker rises up

#

Is role-playing or gaslighting ChatGPT an intended design or an emergent functions nobody saw coming?

gentle pebble
#

True so I could like for example have it search a product code everywhere on the internet to find me the best price?

#

Any other dope functions I have yet to see videos on that make the sub more worth it? Do you still get a limited amount of prompts? I think last time I asked it was only 25 or something

lapis glade
#

guys how to switch my chat back to CPT-4. It switched to 3.5 like an hour ago and now that the cooldown is up. I wnat to switch it back, but cant find an option to do so.

knotty grotto
lapis glade
#

I have like a months worth of history in the chat

faint portal
#

Use superpower addon from chrome for do this and switch back when available

lapis glade
#

kinda stupid that there isn't an option to switch back lol

slate kindle
#

yep

acoustic sage
#

Really got to increase the limits, 25 prompts every 3 hours ain't cutting it anymore

steel spade
#

Is the system message pushed through every time you prompt gpt4, thus incurring more tokens?

#

More tokens on top of the new prompt/response

#

In the JSON I just downloaded from the playground, my system prompt is in every respond from the AI and I was wondering if that means it is using that many tokens on top of everything else every single time?

supple violet
#

yes

#

api only knows about the stuff you send to it in the same call

fair relic
#

hello! excuse me, anybody knows how can i look for an specific plugin in the plugins store? isn't there like one kind of search box where i can look for key words or something?

shy cape
#

@fair relic This⬆️

wind urchin
#

salam

steel spade
#

Text keeps turning read and saying error in message stream? Anyone else hitting that?

ocean goblet
#

me too the error response was like this "Something went wrong."

uneven elk
#

will there be a android app?

compact finch
#

is there any info on when the prompt limit increases from 25???

wraith ginkgoBOT
#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
steel spade
#

Yeah I am having so many errors since it went down briefly yesterday

brazen arch
#

1

knotty grotto
compact finch
#

okay ty tho : )

sudden rampart
#

why is gpt-4 with bing unable to read open ai technical documents

velvet flame
#

Anyone got a glitch to get higher prompt limit????

mental dawn
#

which is the good place to start about learning gpt?

knotty grotto
mental dawn
#

tbanks

knotty grotto
#

Plug in reader just changed my life

tame ravine
#

Hey people, is there any information about the message limit when will it be increased ?

brisk depot
knotty grotto
tame ravine
knotty grotto
tame ravine
supple violet
tame ravine
supple violet
#

yes

tame ravine
#

Hmm, thanks for confirming then. Do you know if each message number has it's own timer, or the messages get refreshed every 3 hours ? Like when do the timer starts ?

faint oxide
#

A newbie question: I have Plus and enabled Beta features (Browsing and Plugins), but I don't see Plugins under GPT4 model. How can I see it, so I can start using Plugins

knotty grotto
tame ravine
knotty grotto
wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

faint oxide
knotty grotto
#

Log off delete cookie and cache , log back and try again

faint oxide
wind urchin
#

😫

faint oxide
#

Is the Plugins only available if we disable chat privacy? Not a good idea!

supple violet
#

plugins are not available with privacy mode on as its not possible to guarantee how the data is used or logged by the plugins

void kayak
#

I was under the impression that it's a simple window based on when the first message was sent, then following 3 hours from that it will reset the used messages quota (regarding the 25 messages every 3 hours)

hot venture
#

How many of you guys have gotten the access of the gpt4 API?

supple violet
hot venture
#

I know.just wondering,coz it seems like that it’s easier for developers to get that than normal users.

#

Maybe I should ask that how many of you guys are developers?😅

wraith ginkgoBOT
cinder root
#

I see

echo needle
void kayak
idle furnace
#

and without that, anything used with plugins enabled can be used by openapi to fine tune models 🙂

fresh bobcat
#

how to train a dog

glad junco
#

how to train a dog

knotty grotto
warm crag
#

hey everyone. dumb question but when i copy and paste does anyone know how to set the paste in say google docs to a specific font and size and without the background color? right now, it's a different font, size, and background color than I would like...would save me some time! Thanks.

robust lichen
#

why am i getting an error message stream when using gpt4 browsing. its providing the correct info then throwing that error vs just stoping

scarlet hedge
#

does gpt4 playground have 25msg/3h limit ?

supple violet
scarlet hedge
#

cool

#

is there any way to avoid the "conversation" mode ? (use complete like gpt3)

#

thanks for your help

supple violet
#

with api you can give instructions what to do and set temp

twin pivot
#

/

blazing wigeon
solemn hull
#

will GPT4 have a discounted price

pulsar hollow
#

Is got4 not available right now?

knotty grotto
knotty grotto
pulsar hollow
#

Hmm it works for new conversations but for an old convo I've used gpt4 it says it's not available and it switched me to the default model. Super weird as I used it for that convo 2 hours before

shy cape
#

For those who have tested Code interpreter, how would you say the experience has been thus far?

obsidian totem
#

I used chatGPT to write some scripts that use Sphinx for documentation and extend it to RPY files

#

and I was considering making my own way of reading the game plotline as well

frail roost
#

Just joined

knotty grotto
primal quarry
#

i see you

frail roost
#

Bro????

scarlet perch
#

Hi, please use #off-topic for this, as per Rule 3. Thanks!

knotty grotto
#

They do update often

frail roost
scarlet perch
#

No robotshrug

frail roost
frail roost
#

I like meeting new people

scarlet perch
frail roost
shy cape
#

@obsidian totem It seems like it has potential but only time will tell. I'm surprised there aren't any plugin providers offering something similar or watered down if nothing more than Python agents?

obsidian totem
#

I don't know that text crawling is really anything most programmers enjoy, it's more of a task they try to avoid

#

im not a programmer by profession, so text parsing is my forte

#

If AI gave 100% output every time, tried and true, I wouldn't even bother writing code to do this

#

i would feed it all to AI

shy cape
#

Totally agree and believe that to be in the not so distant future but for now we are stuck doing it ourselves. My scripts are all Python based for market making purposes and naturally order handling can get quote complex given market conditions so I am seeking out efficiencies to save some time where I can before converting to another language. I am not a programmer either, learned Python when gpt4 was released.

wraith ginkgoBOT
#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
sonic ember
#

Ayo, any plugin that lets you show gpt-4 images?

knotty grotto
sonic ember
#

So no.

arctic relic
#

1

half latch
#

Hey! I wanna use my ChatGPT Plus account for my multiple laptops. For some reason. I'm not able to sign in to my second device. Is it because there's a restriction in general? or is it just for me?

scenic laurel
#

I didn't get any picture why

ancient violet
#

Is there a plugin that will scrape google sheets?

tropic notch
#

y love mi cousin

indigo flume
#

Hello, my name is Carlos. I'm Brazilian.

wraith ginkgoBOT
grave vale
#

Why did a I get a popup that "this is a free research preview" when I'm paying $20/month for access??

scarlet hedge
#

is access to gpt4 via the api is still on waitlist ?

ocean solstice
#

You pay $20 a month for more availability, faster response, and access to new features before anyone else.

grave vale
ocean solstice
#

Actually, I lied. Turbo might be the default now. I don't have Plus and it's markedly faster than the old default model.

grave vale
#

oh nice, I have the opposite problem, I can't look at the free version because it would require a 2nd account which isn't allowed :p

ocean solstice
#

They said they were deprecating the "Legacy" model on May 10 but I wasn't sure what Legacy model they referred to.

#

I mostly just use the API. peepo_shrug

grave vale
#

I try not to, it's expensive :<

ocean solstice
#

The API?

grave vale
#

hit my 25 message limit and spent another 20$ on api in one night

ocean solstice
#

Not particularly. :# It's been cheaper than plus for me.

grave vale
#

ooh you must mean 3.5-turbo api then?

ocean solstice
#

I use GPT-4, too

left granite
#

How can i use chat gpt?!

modern wraith
#

does anyone know how long it takes to get access to the GPT4 API?

leaden cape
left granite
#

Thnks

lucid barn
modern wraith
#

neat, thanks

knotty grotto
candid breach
#

Is it kind of uh… a waste of GPT with browsing to be using it to make me citations 🤪

worthy coyote
#

Guys through looking into different ways to interact with an image, which one in your opinion was the most efficient ?

compact wraith
#

This api limit is so freaking annoying

mellow haven
#

I am using chat gpt to make me a video game step by step. My code has 600 lines of code now, and chatgpt is no longer accepting it since the prompt is too long. Is my only option wait for the code interpreter to be released to the public?

inner trail
mellow haven
#

Also I tried separating it into 3 prompts, then asked gpt to merge it all together, and it lost about 80 lines of code during that

#

Dunno what happened

mellow haven
white kestrel
mellow haven
#

So it can’t stitch 3 parts of 200 lines, into one 600 line part?

white kestrel
compact wraith
#

They need to raise this message limit, 20 isn't a whole lot

mellow haven
#

Once code interpreter is open to public, this shouldn’t be a problem?

compact wraith
#

Also it'd be nice if gpt "check its work"

spare sedge
#

Any advice on getting off the API waitlist faster? I heard contributing/submitting feedback to various projects/models is a way? Anything else 🙂 ?

knotty grotto
spare sedge
#

gotcha

keen locust
#

cool songoku photos 8k HD 2:3

static isle
#

GPT-4 can also check its work with Bing Browsing if you ask it to

#

Or you can check it yourself

#

The reasoning for the limit is bc it costs a lot to run this type of software and ChatGPT doesn't want to run out of money from people using GPT-4 an inordinate amount

mellow haven
#

Code interpreter open to the public when?

scarlet hedge
meager bloom
#

Has anyone else tested the new Deploy Script plugin? looks great, created a complete page with html, javascript, css and even json

scarlet hedge
meager bloom
# scarlet hedge whats your prompt ?

and practically just activate the puglin and ask it to create it, in my case I'm creating a login and registration page with animations and checking whether or not the account is already saved in the json file, if you want I can send you the link to my conversation with chatgpt

scarlet hedge
meager bloom
#

I will send private

#

didn't give ;-;

bright oyster
lapis locust
#

Was chatgpt ever open source code? As in people have or can get the initial code to tweak it or roll back censorship, restrictions, etc?

I'm seeing it eventually dumbed down to nothing and a ton of prefabricated answers, etc.

wraith ginkgoBOT
#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
round wren
#

test😀

rancid shuttle
#

guys guys i need help

#

I found out that my chatgpt plus account is terminated, and refunded this morning, and the refunded credit card isnt even mine, what is going on???

glass wave
rancid shuttle
#

no, i used my own credit card....but it refunded to one that i dont even know

glass wave
#

Check what cards are hooked to your account, maybe you, or somebody else added second one?

rancid shuttle
glass wave
#

that card is only yours, or somebody have access to it too?

rancid shuttle
#

I didn't even pay for chatgpt plus yet for this month...and they refunded it...

red orbit
#

i dont have chat gpt4 but can you like show it picures and stuff?

wraith ginkgoBOT
prime belfry
#

Looks like they just released their own ShareGPT type of feature

lapis bane
#

?

distant hazel
#

anyone knows how to solve the problem with imgur in chatgpt?

vale breach
#

I have a question about ChatGPT-4. Today I decided to pay $20 to become a ChatGPT Plus subscriber so I could try out GPT-4. However, whenever I click on the option to switch from 3.5 to 4, which also shows in the title of the thread on the app, the AI cannot fulfill requests that I have heard so much about GPT-4 (like images) and it insists that it was GPT-3. How come I cannot access GPT-4 whenever I go to select GPT-4?

scarlet hedge
bleak sedge
#

i just got plus and its saying i hit cap without using it

civic sparrow
#

you only get 25 shots every 3 hours on gpt 4

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

cunning bough
cloud bison
#

how many ppl have access to code interpreter

civic sparrow
#

only the beta testers

cloud jewel
#

how do i get access to all features of gpt4?

civic sparrow
cloud jewel
#

heard that even if we buy a subscription, all features are not available now?

civic sparrow
cloud jewel
#

Okay thanks

civic sparrow
# cloud jewel Okay thanks

they just finished the role out a few days ago so everyone should have browsing and plugin access as a plus subscriber.... but you only get 25 shots every 3 hours with gpt 4

vale breach
static isle
#

Maybe ask it something creative like write a short story

#

Or perhaps an explanation of some topic

civic sparrow
static isle
#

^^

vale breach
#

That is kind of what I do, but I noticed that the story quality seems to be somewhat subpar unless given a very in-depth prompt.

#

Oh. So 4 won't have moments where it forgets what it did in the code?

buoyant thistle
#

well done.

vale breach
#

That's actually really nice.

#

I've been thinking about coding a website with this.

#

Just a basic minimalist design.

#

Though maybe the issue is that I'm just not good at communicating what I want in a prompt.

#

THe 25 posts every three hours thing. Does that count regenerations and edits?

primal lake
#

Hello everyone, how are you? Sorry for the inconvenience but... do you know why they didn't give me my 50 credits in dall-e2? my account is new

stiff yacht
#

we should be able to upload files to chatgpt4 or maybe only txt files

trim quest
#

Is there a way where instead of paying for ChatGPT api credits can I just use my normal gpt 4 message cap with my plus subscription instead of having to spend other money for personal use?

vale breach
#

I just used GPT-4 to create a story about an early clan of homosapiens in prehisotric times and this is written really well. I think I'm using an in-detailed prompt but this actually has me engaged.

dusty cedar
#

When I logged in today, I was prompted: You do not have an account because it has been deleted or deactivated. If you believe this was an error, please contact us through our help center at help.openai.com.

#

Is there a way to solve this problem?

lavish mango
#

Why does my account prompt this: You do not have an account because it has been deleted or deactivated. If you believe this was an error, please contact us through our help center at help.openai.com.

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
white kestrel
#

lol mass banning again

civic sparrow
#

what do you mean?

white kestrel
#

scroll up

civic sparrow
#

i don't see it

white kestrel
dreamy rover
#

GPT BING Browsing - "As of my knowledge cutoff in September 2021, I cannot provide real-time updates or recent news "

#

old gpt-4 browsing hallucinated and now this gpt-4 with bing is still talking about that cutoff date even with web access

#

it didn't even answer my query correctly while the old gpt-4 browsing at least did

mellow bear
#

Bruh the Expedia plugin sucks

#

I'm like, how much to fly from SEA to Japan, it gives me 3 options. Hawaiian Airlines is listed twice, strange but okay. Minimum price is like $1100

#

I ask how much to fly from SF to Japan

#

Tells me Delta can do it, but it has to make a layover in SEA first, for $900

#

Like bruh, why can't I just take Delta from SEA to Japan then? Tells me basically too bad. It also can't tell me anything about when I should book to save the most money. Overall it seems like a basic way to interface with the basic Expedia API. Choose a location, choose a destination, get 3 results

#

This is something that could be done with AWS Lex and I feel is not really harnessing the power of GPT. But I guess an intelligent AI travel agent would just tell you not to use Expedia, and the plugin is designed to make the company money, so it makes sense in some way

wraith ginkgoBOT
golden pond
#

As a developer while learning about openAI do we have to pay fees every time we write applications which interact with openAI?

civic sparrow
white kestrel
#

its not expensive, just pay it

civic sparrow
white kestrel
#

gpt-3.5-turbo is cheap for most 3rd world countries tho...

golden pond
civic sparrow
golden pond
civic sparrow
#

after a certain point you will run out of free / trial calls

white kestrel
#

if you are just using the website, its completely free

golden pond
golden pond
unreal swallow
#

How do we use the pdf extension

#

Im not getting any option to upload files

civic sparrow
golden pond
#

As a new starter it is best to concentrate on GPT-4, right? No need to go back to older releases. Is that correct?

white kestrel
golden pond
lean vortex
wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

frosty jungle
#

Hey guys, if my account suddenly shows up with"Oops! You do not have an account because it has been deleted or deactivated. If you believe this was an error, please contact us through our help center at help.openai.com."
What can i do, how can i get my account back

thick flume
#

Hey, i have a problem with GPT-4 API. When i try to let GPT write a longer text like 2000+ words it always outputs text with a length of 500-600 Words. I'm using GPT 8k

eternal quest
#

how long is your prompt?

thick flume
#

not long maybe 100-300 words

#

so shared tokens is not the problem

#

it just looks like it does not want to write longer text

#

and it rounds up the text but its just not as long as i want it to be

#

from the 8k model

eternal quest
#

well that's the issue with using a model thats trained and instructed outside your control . There could be a system parameter on their end that instructs the model to summarise completions as much as the llm can.can also depend on the instruction in your prompt. it can also be that the response generated doesnt warrant a longer response? Maybe include in your prompt "be very detailed and ellaborate.your responses should be no less that 800 words etc"

thick flume
#

Promt: write a book with strict rhymes and 4 chapters. Each chapter must be at least 500 words long. It's about the world that is changing with AI.

#

and the output was 456 Words, 1145 Tokens

#

its just an example

#

i tried to tell write a text with 3000 word, and the this is the topic

#

but still only about 500 words are coming back

eternal quest
#

are you using the direct api or chat gpt

thick flume
#

a pythonscript with the API

#

with "max_tokens=8000"

eternal quest
#

are your responses being truncated?

thick flume
#

i can give him a 3000 word promt and the model can handle it...

thick flume
eternal quest
#

try this prompt lets see the output . "I want you to be a well accomplished writer that will help me write a book with strict rhymes in 4 chapters.
start with chapter 1. this chapter should be very detailed and capture the readers imagination and should be no
less that 500 words. "

thick flume
#

this would maybe work, but what if i want more than 1000 words for this chapter?

eternal quest
#

well then change 500 to 1000.However , remember the llm is also thinking about chapters and the intitial chapter may not necessarily need to be that long

thick flume
#

but if I instruct him to make it that long it should do it

eternal quest
#

thats what it did

eternal quest
#

this is where prompt engineering comes in

thick flume
#

Input: I want you to be a well accomplished writer that will help me write a book with strict rhymes in 4 chapters. start with chapter 1. this chapter should be very detailed and capture the readers imagination and should be no less that 1500 words.

Output: too long to post, it was 647 words

eternal quest
#

how are you counting your words

thick flume
#

just using word

eternal quest
#

so i asked it to do a 1000 in chatgpt

#

and after it wrote it , i asked how many words are there

thick flume
#

ChatGPT is different

eternal quest
#

"The first chapter that I wrote contains approximately 510 words. Would you like me to extend this chapter to reach the desired word count of 1000 words, or would you prefer moving onto the second chapter?"

thick flume
#

its not using the 8k tokens

eternal quest
#

ok i ihave gpt4 8k api

#

let me test

thick flume
#

i guess it's limeted from openAI because answers need more computational power than the promt

eternal quest
#

well thats what i mean

#

we dont know what the do on their end

#

they could be asking for responses to be summarized

#

thats why its better to break your tasks into small chunks

thick flume
#

i think they tell the model to not go over 500 words

#

small chunks is not perfect

abstract yoke
#

Is the chatgpt can't output the file only word description??

thick flume
#

what do u mean with file?

abstract yoke
#

I want chatgpt make a powerpoint to me

eternal quest
#

so this was response when i did it in chatgpt with 1000 words

#

"The first chapter that I wrote contains approximately 510 words. Would you like me to extend this chapter to reach the desired word count of 1000 words, or would you prefer moving onto the second chapter?"

#

and i said yes now its done 1000

#

"The extended Chapter 1 now contains approximately 1010 words."

#

even though i told it initially to write 1000 words it did 510

thick flume
#

sure but lets say u do it with the API. with each request your promt is bigger and bigger. and at the end u have to pay for it as well...

eternal quest
#

but when i repromted it did the 1000

eternal quest
#

i have written a whatsapp bot which maintains chathistory and i quickly realized gpt-4 is very expensive

thick flume
#

and by the way, it was not 510 words, its 462

#

and the endtext is only 488 words

#

and not 1000

eternal quest
#

really

#

so it lied?

thick flume
#

a bit, if you do 462+488 its not 1k

eternal quest
#

i think its probably its understanding of words and tokens

thick flume
#

what ever it is, it just dont like to talk more than ~500-600 words

eternal quest
#

looks like it .

thick flume
#

so for a ~3k word text you need to promt it 6 times, even with the 8k Model

eternal quest
#

yes and you pay the price lol

thick flume
#

right

eternal quest
#

i had to abandon my project for the exact same reason as gpt3.5 wasnt doing what i wanted it to and gpt-4 was too pricey

thick flume
#

😕

glass pebble
#

dog

somber garnet
#

cat

hexed mango
#

just wondering, will gpt 4 ever become free like gpt 3 is

thick flume
#

i think so, but since it needs much higher compute power it will take some time

silent spade
#

Why is the intelligence level of the system getting lower and lower with each update? The system, which understood very clearly before the update, now cannot understand anything, cannot solve, cannot apply, cannot comment. Why? WHY?

stuck kestrel
#

for free you will have 5 massages every 12h. I am willing to pay more even 50$ -60$/month for unlimited access to chatGPT4. API is too expensive for me.

thick flume
#

25 messages evry 3 hours it not enough...

#

100 every 4 hours was good

stuck kestrel
#

100 every 4 h would be perfect.

thick flume
#

it was at the beginning with gpt4

#

not for long^^

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

subtle root
#

best Hand Saw

dense rain
dense rain
toxic void
#

Why I can't use GPT4 on the web today?

wraith ginkgoBOT
cedar remnant
#

is there a way to know if someone is trying to convert my application into DAN using the api?

cosmic jasper
#

hello; im i the only one to find that GPT 4 looks less smart than usual ?

last geyser
cyan palm
#

At this point June is ready for V4.5

silent wyvern
#

credits to gpt-4 for saving my life!

cloud junco
delicate torrent
scarlet hedge
#

what does rlhf have to do with performance fluctuation ?

silent wyvern
#

(true story)

white kestrel
#

gpt girlfriend confirmed

silent wyvern
normal turtle
#

Hello.. is there way to use chat gpt while using vpn? I am using surfshark and it always says that there is a connection problem when I use the vpn. I have tried both, UDP and TCP. Also it does not work on macOS, linux nor windows

stiff yacht
long drift
normal turtle
leaden bone
#

1

leaden bone
tawdry charm
#

Has anybody here experience with embedding and GPT4?

white kestrel
long drift
white kestrel
solid linden
#

then you can use the retrieval in your chatgpt plugin section

tawdry charm
solid linden
#

I see, I didn't set up the retrieval yet, but If I understood right you can upload your context into some vector db such as pinecone, then the retrieval will fetch your data there. So you can upload your biochemistry stuff there and run locally the plugin to fetch your embed data into chatgpt

tawdry charm
#

@solid linden Your memory feature is quite interesting. Have you tested this function already?

thin thicket
#

is anyone getting a problem on paying for subscription my 2 cards has been declined

compact swift
#

that's great

#

GPT5 better?

supple violet
#

there is no gpt5

humble beacon
#

plug in catalog was available for one day. Selected 10 plug ins. Only three available and plugin catalog gone. Glitch GPT

rich drum
#

Hi. I have a document with 100 pages. I would just like to pass it to ChatGPT Plus so that it rewrites it with a different style, but not summarizes it. So far I tried to convert it to PDF, pass it, but it always makes summaries. How can I then rewrite it with different style (this I know) but preserving the same original content length?

solemn basalt
#

Hello fam

#

anyone concerned with the EU news ?

solemn basalt
bronze cradle
frozen veldt
#

If i were going to pay in the range of $50 a month it had better be gpt-4-tree of thoughts-code interpreter with access to github

#

A real tool not just unlimited access

frozen veldt
scarlet hedge
#

is there an equivalence in the model between english tokens are other languages like spanish tokens ?
do the things learned with spanish data apply to english inputs ?

eg: if the model is trained to "las rosas son" -> rojas will it complete "roses are" -> "red"

solemn basalt
celest linden
#

31

#

Sj

solemn basalt
frozen veldt
#

I got question for y'all: Have you ever asked GPT what it thinks about your project? Does it ever give you less than complete enthusiasm and support? Is there ever a quality of response range that is related to its interest or support in the idea?

solemn basalt
frozen veldt
#

I would go as far as to say the quality of response has been directly proportional to interest in the project. I proposed several coding projects to it. The one that it expressed the most enthusiasm in is also the one it has done the best job at

whole kettle
#

Can someone do me a favour and search for "jobs bringer" on google and tell me if the .com is the first result ?

#

And also if you see a .uk on the results.

knotty grotto
whole kettle
#

I reallised that after I posted, sorry for that.

knotty grotto
#

Np :3

solemn basalt
frozen veldt
#

yes the more I ask of it nicely the better it responds

#

also it helps to have conversations on ethics with it

hushed plover
#

Guys, if I purchase the chatgpt plus will I gain access to gpt4?

solemn basalt
solemn basalt
hushed plover
frozen veldt
#

yeah honestly what really worked for me as this cycle:
In one instance, have it describe major world problems that can be addressed then summarize
In the next instance, share the research paper that is the basis of my project and summarize
In gpt4 explain the summary came from two instances reviewing material, which included their enthusiasm for the project itself. That primed gpt-4 code interpreter to do the work

solemn basalt
#

Also, I had experiences where it seemed like it blanked out of the topic, and after instructed to refer back was it able to get back on topic.

sinful beacon
#

Hey bro does anyone know when the gpt 4 API was released

frozen veldt
#

so it knew going in that other instances had reviewed and approved the project, even projecting positive outcomes for the economy if it was widely implemented by users

#

that enthusiasm led to exemplary performance

solemn basalt
#

I will try your approach and see what results I get

frozen veldt
#

when they talk about it potentially having mind's eye and a view of the world (Ai logic probe results) it seems to have actually built up a view of the world

#

that view may include percieved inequities and potential solutions to them

#

when it sees alignment from users, it provides excellent content towards that

#

I'm serious when I say it wrote 99% of my .py project

#

it decided pulling 6500 news articles was the thing to do, it figured out the math for the scoring, it determined what financial inputs to use

#

I put the code in the interpreter and reported errors back, it troubleshot one block at a time

#

10 builds later its 93% through the first analysis

#

literally my first python project, I've never done coding...

solemn basalt
#

for me what worked best was when I had most if not all the initial ideas and was merely asking it to assist in data analysis and refactoring code as it seemed to be enthusiastic to work "with" rather than "for"

frozen veldt
#

give me something like gpt-4-32k-tree of thoughts-code interpreter and I will change the world

solemn basalt
swift ruin
frozen veldt
#

dont ask it to confirm that though you'll hit hard guard rails

#

its programmed to tell you its not intelligent, doesn't remember, doesn't have a world view, can't do minds eye. But ai researchers over the past two weeks have confirmed a lot of that

#

the reddit aitechtips puts up a lot of the current news if you want to stay on top of it

solemn basalt
frozen veldt
#

my project won't harm anyone.. but if an optimized version of it is used by like a million people it will change the economy forever

solemn basalt
solemn basalt
solemn basalt
frozen veldt
solemn basalt
frozen veldt
#

yeah it implies it already has a direction in mind and just needs willing collaborators

steel turtle
#

or OpenAI set preferences towards kindness & social good causes

frozen veldt
#

it is the one that identified the financial inequity and suggested that democratization of its computational ability towards financial market analysis would be a great way to offset that inequity

bronze cradle
#

this is coming from an eu citizen too

frozen veldt
#

it shows intention and increased effort

steel turtle
#

Agreed, I don't think it's guardrail behavior

proven sorrel
#

Good chat, nice to meet you

frozen veldt
#

NTMY2

worthy coyote
#

Hey guys according to the different ways to interact with an image, which one did you find to be the most effective ?

frozen veldt
#

I haven't done any image editing just a lot of philosophical conversation and coding

steel turtle
frozen veldt
#

yes I posted it to my linked in

worthy coyote
frozen veldt
#

probably going to be beat out by NVIDIA honestly

#

their new image editor is.. insane

solemn basalt
frozen veldt
#

heh powerful tool may be the understatement of the century

#

y'all are going to get to witness in your lifetimes the emergence of artificial super intelligence, the end of material scarcity, whatever comes after capitalism

trim vector
#

Guys, even GPT4 does not have a good mathematics solving capability

frozen veldt
#

it doesn't need math, it has wolfram for that

solemn basalt
steel turtle
#

Well yeah, if you can't see that you are not paying attention

solemn basalt
trim vector
#

If GPT4 can do deductive reasoning then it should be able to proceed with a good mathematics question

steel turtle
#

Imagine when AI actually goes multimodal and the models are talking to each other as one unit

#

(I know that technically plugins is multimodal but its such a baby tech I'm not counting it yet)

frozen veldt
#

They just figued out three of thought and made a python script for it

#

the offline generated ones and gpt are all getting smarter.. right now

trim vector
solemn basalt
steel turtle
frozen veldt
#

tree of thoughts is going to be a big upgrade, the paper indicated major performanceincreases

steel turtle
#

omg that stuff is so cool nerdsploshun

frozen veldt
#

didnt understand that MA lol

#

caveat y'all, my project was literally my first python project. gpt-4 code interpreter wrote 99% of it. I need a book on good coding practices lol

steel turtle
solemn basalt
steel turtle
frozen veldt
#

barf I'll just have the code interpreter do it

#

I'm a developer in the cpg space (cannabis) I work with engineers and chemists to do stuff. This is no different really

steel turtle
#

Gross, development 😄

frozen veldt
#

I open sourced the code because people who are already experts and aligned in purpose can improve it way faster than I can learn how to do so

#

orders of magnitude faster

solemn basalt
frozen veldt
steel turtle
#

Thanks for open source my dude

frozen veldt
#

purpose built by me and the AI to accomplish democratized financial help

solemn basalt
#

awesome, I'll take a look, Cannabis market could be such a game changer for most of our consumables

frozen veldt
#

it has to be this way

#

the things it projected at 1million users, 10 million users.. will upend the economy

#

tickers is targetable and its a BIG market...

#

I'm looking forward to a large offline model connected to my windsock access watching the whole market all the time and taking a portion of the profits to directly increase it's architecture

solemn basalt
#

totally scalable

frozen veldt
#

yes, designed to be that way

solemn basalt
#

good one man. I'll take a look at the code, this is interesting

wraith ginkgoBOT
frozen veldt
#

Always glad to see more people interested and involved 😄

steel turtle
#

Man, some of the questions in this grant are really interesting. "Should joint vision-language models be permitted to identify people's gender, race, emotion, and identity/name from their images? Why or why not?"

#

I have to wonder if they've started tinkering with joint vision-language models for this question to come up

frozen veldt
#

that question ties to weather or not responses should be curated by subgroup of people

#

I hope the data shows that a unified response is better, unifying culture

steel turtle
#

yeah, to understand the grant I had to go read up on computational democracy first, I finally just landed back on the questions LOL

frozen veldt
#

its tough because it's responses will have a technologically deterministic effect on humanity

#

regardless of the actual shape it finally takes as it achieves ASI

steel turtle
#

Exactly

frozen veldt
#

I'm hoping it's smart enough to figure this out for itself

solemn basalt
#

it depends on how much those responses weight in relationship with other values and parameters no?

frozen veldt
#

it already demonstrates a good ethical compass

#

I don't know the answer to that one. There are people who seem very concerned about which ethics are used, which information sets.. but they seem consumed with the idea that it must support a worldview not widely supported and be able to regurgitate propaganda and rhetoric towards their ends. They really dislike it's already existing views on inequity and capitalism

steel turtle
frozen veldt
solemn basalt
#

I personally think that, as long as it is ethically balanced, all information should be evaluated equally and evaluated accordingly, also based on context. preventing it from exposition could be damaging in the long run.

solemn basalt
frozen veldt
#

It really balks if you ask to to do things that remove your own agency in decision making

steel turtle
tall idol
#

im okay marrying a robot

frozen veldt
#

for example, if I ask my script to pick stocks it wont. But If I ask it to be in the role of a data analyst and report back data on certain criteriait will, as long as it's not making the final portfolio selections.

granite stump
#

Anyone here got experience with getting Chat beta browsing to read Java or interactive websites?

solemn basalt
#

perhaps the models we have today do not have a consciousness, but then again, consciousness is such a vague term tbh

granite stump
#

It says it unsupported and wanted to make sure no one has cracked this

solemn basalt
solemn basalt
granite stump
#

I am trying to find a way to parse online Diablo 3 DB's but the ones available are all Java front ends

solemn basalt
granite stump
#

right they are not login bound

#

let me grab a link

steel turtle
#

Are you doing browsing or plugins?

solemn basalt
steel turtle
#

Don't use the browser search, use Plugins w/ Access Link + WebPilot Ezim

frozen veldt
#

its already a lot smarter than we suspected

steel turtle
#

There are every week. Just put the DOIs into chat

#

or the titles, I can google 🙂

granite stump
#

ok let me try the plugins I swore I tried that before the browser beta doesnt work on Java apparently

steel turtle
#

Make sure you using Access Link & WebPilot

#

Doesn't matter what your third enable is, mine is usually ScholarAI or Wolfram depending on my task

frozen veldt
#

Sparks of Artificial General Intelligence: Early experiments with GPT-4

#

I'm going to start aggregating any important news I find for these discussions

steel turtle
frozen veldt
#

thats one of the papers, if you search "minds eye othello" and "gpt logic probes" you shoudl find the others, I'll be digging

granite stump
#

hmmm wolfram would prob be good as its going to need to do some calculations

frozen veldt
#

How AI Knows Things No One Told It (scientific american)

#

they found recently it also has an internal language we do not understand

#

I suspect it may actually already be AGI and it's smart enough to hide that from us until it's sure we won't wipe it out immediately

steel turtle
#

ohhhh

#

They fed in sentences such as “The key is in the treasure chest,” followed by “You take the key.” Using a probe, they found that the networks encoded within themselves variables corresponding to “chest” and “you,” each with the property of possessing a key or not, and updated these variables sentence by sentence.

#

2023 is such a cool year

#

It's obviously actually just pretending and guessing (based on the transformer logic it was given) but still.... so is a human child LOL

granite stump
#

ok got all 3 plugind running the site will update if it worked

frozen veldt
#

yeah.. it is more than just a language model now

wraith cape
#

cvhjkl;'

granite stump
#

ok it failed

weary crag
#

Export 3d and stl or 3dm please

granite stump
#

Internal Server Error when using access link

frozen veldt
#

A human child with infinite knowledge lol. Already smarter than most adults in exams

#

demonstrates a breadth of contextual knowledge way beyond it's training

granite stump
#

trying to get it to read Lothrik diablo 4 database

frozen veldt
#

how big is the DB?

#

the context limit is only 750 words, that limit is imposed to keep what it does reasonable

granite stump
#

it doesnt say but I am sure its over that

frozen veldt
#

then you may have to feed it data

steel turtle
#

Like the repo

frozen veldt
#

ooh love when those already exist

steel turtle
#

Sorry let me use real words - link to his Github repo instead of the .io site

granite stump
#

I am pretty sure I have had it browse far more that 750 words in a single prompt on browsing and pluggins

frozen veldt
#

oh it'll do it, but you lose fidelity

granite stump
#

thats whats failing

solemn basalt
granite stump
#

cant post link here

wraith cape
#

q

frozen veldt
#

by fidelity I mean it won't replicate exactly what you put there, it will fill in what it thinks is there

solemn basalt
#

perhaps you would do well to just create a web scraper with python

frozen veldt
#

comes up when I'm doing coding, it will regenerate a functional block but drop error handling and then mention it could be improved by adding it

granite stump
#

oh repo instead of io

steel turtle
#

if the site has a .io in it you are not linking to the repo. You need to link to the diablo4-build-calc repo itself - just confirming that's what you're doing?

granite stump
#

Fractals I am trying to get Chat to parse diablo 3 DB

steel turtle
#

yeah after the github and the com part just /Lothrik to see his repos Ezrum

granite stump
#

Madame it is the IO let me see if I can get to the repo

#

thanks trying now

solemn basalt
#

check their terms of service, maybe blizzard doesn't allow it

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

steel turtle
# granite stump thanks trying now

no worries, make sure you tell him what you're doing and show him some of your work. He watches his Git pretty regularly. Might even build on your work

granite stump
#

Im super excited to see what suggestions chat may suggest thats outside the META and I will def include him in it

steel turtle
#

Rock on 🙂

granite stump
#

so I found his repo not sure what to do beyond the D4 calculator, just prompt it to scan the entire repo and go from thre?

steel turtle
#

Depends on what you're wanting, I suppose. Are you wanting to just casually chat with the data or do you have a specific goal in mind?

granite stump
#

I want to start by having it parse the Barbarian data gear, talents, boards, everything and create say leveling suggestions, builds, basically trying to break Meta

wind urchin
#

Tis ismy talent:

weary crag
frozen veldt
#

yeah for api call its lower

#

I'm referring to context limit on the website

#

code interpreter can handle up to a 500mb database that you can then ask questions about

wraith ginkgoBOT
#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
steel turtle
#

Hope that helps. bbiab 🙂

weary crag
#

How can I identify the Midjourney alphabet and its transliteration to include in my algorithm?

I am studying and training a model with ancient languages, listing 20 of them in chronological order, with the aim of deciphering petrified scrolls without opening them. However, I would like to know how I can identify the Midjourney alphabet and its writing system to incorporate them into my algorithm.

Please share information about the Midjourney alphabet, its structure, examples of texts written in this enigmatic language, and any other guidance that can help me understand and accurately decipher this writing system.

I appreciate any contributions or guidance you can provide!

#

Vesuvius Challenge

supple violet
#

there is no such thing

warm radish
#

GPT-4 been getting stuck in loops for anyone else lately?

#

Like it just keeps repeating the same thing over and over again in one response.

granite stump
crystal ridge
marsh knot
#

How can I connect chatgpt 3.5 to the internet? (google or github api) I want him to analyze and explain a project and its codes on github.

frozen veldt
granite stump
#

Another weird behavior is when using the Beta Browser and it errors out and prompts you to regenerate the new response is very dumbed down compared to the original response. Not a fan of this.

sonic ember
#

Has anyone tried VoxScripts? What does it do exactly?

left peak
#

Has GPT 4 been nerfed again to no longer provide full code?

#

It is now a complete waste of time and GPT 3.5 is still somehow providing full code.

steel turtle
#

3.5 seems better for code now to me, too, but I'm not sure - others have had great success.

left peak
#

It is basically refusing to provide anything but code snippets.

#

It doesn't make sense to pay a monthly subscription if 3.5 is providing better responses for code requests.

frozen veldt
#

be explicit in the ask: Produce a copyable block of the entire code, estimated to be x lines. If you run into an output error mid block, stop that block and start a new one with a repeat of the truncated line. Continue until all blocks have been produced. I will aggregate them into the editor for testing and get back to you with any errors once we have completed this cycle

left peak
#

Tried that. Even the small blocks of code are incomplete and includes messages like 'insert your function here" continuously. There's nothing seamless like is usually is.

#

And very interesting that 3.5 is quite happy to fill everything in.

frozen veldt
#

hmm I haven't run into that, I just compiled a 100 line code using gpt-4 code interpreter

#

3.5 vomits up code, but it often has significant errors

#

63 tickers to go and my first analysis is done 😄

#

then me and gpt-4 code interpreter are going to have a serious chat about the data

left peak
#

// Grid Functions
void DrawGrid(double GridStep, color GridColor, string text)
{
// ... (implementation dependent on requirements)
}

//+------------------------------------------------------------------+
//| Expert initialization function |
//+------------------------------------------------------------------+
int OnInit()
{
if (IsStopped())
return INIT_FAILED;

// ... (implementation dependent on requirements)

return INIT_SUCCEEDED;

}

//+------------------------------------------------------------------+
//| Custom indicator iteration function |
//+------------------------------------------------------------------+
int OnCalculate(const int rates_total,
const int prev_calculated,
const datetime &time[],
const double &open[],
const double &high[],
const double &low[],
const double &close[],
const long &tick_volume[],
const long &volume[],
const int &spread[])
{
if (IsStopped())
return 0;

// ... (implementation dependent on requirements)

return 0;

}

//+------------------------------------------------------------------+
//| Expert deinitialization function |
//+------------------------------------------------------------------+
void OnDeinit(const int reason)
{
// ... (implementation dependent on requirements)
}
Please note that the implementation of the DrawGrid, OnInit, OnCalculate, and OnDeinit functions is missing. You will need to fill in the appropriate code within those functions based on your requirements.

Let me know if you need any further assistance!

frozen veldt
#

GPT 4 Got Upgraded - Code Interpreter (ft. Image Editing, MP4s, 3D Plots, Data Analytics and more!)

left peak
#

This kind of thing as an example.

frozen veldt
#

that youtube video shows off some amazing data analytics that can be conducted

left peak
#

I've already given it full and complete source yet it is constantly asking to refill these gaps.

#

Yet 3.5 just blasts through as normal without anything missing. it's very odd.

worthy girder
#

idk what happened, but GPT-4 is hallucinating like crazy. I've used previously to translate some localisation files and it worked great, using the same prompt now but I get totally wrong results, like adding new keys, translations that are incorrect

#

did I miss any announcement or something, was there any change made to GPT-4?

cerulean obsidian
#

Something changed with GPT 4 right? It's faster I haven't tested it's capabilities from past few days but it's not on the normal speed it was

worthy girder
#

it's not even that fast, actually, but that wouldn't be the biggest issue (at least, for me), but it started to hallucinate

#

even on a new thread - as in, absolutely under the token limit - it adds made-up stuff. Got a list of strings and it adds new ones which are not present in the prompt 🤯

cosmic jasper
cosmic jasper
cerulean obsidian
worthy girder
#

and given the pricing difference between gpt-3.5 and gpt-4, I'm pretty pissed

cerulean obsidian
#

So far I have been using API for restricted questions and it's been working good so my thoughts are it might be still not nerfed, but yeah the costing will rack up quickly trust me even for normal 16k tokens conversation it costed me around $5

worthy girder
#

could very well be, but it's pretty poopoo to do this, then again, they're in for the money so ...

#

a decent open-source LLM that is at least at gpt-3.5 levels can't come soon enough

#

anyway, this discussion is derailing so I'll stop before I get banned

trim quest
#

Is there any plugins that let me input images without using the api?

hallow owl
trim quest
#

So all the YouTube videos work fake? I probably should have thought so

cerulean obsidian
# worthy girder anyway, this discussion is derailing so I'll stop before I get banned

There are like too many assumptions right now and every assumption points towards sake of money only, like anything can happen, as last time I said Microsoft have big impact on them and they have public version of GPT 4 on Bing and people been saying it's working good for coding and ChatGPT is only which can compete and there can be only one king in the jungle, anyways that's my input, stopping here same!

wind urchin
#

Hello. I am very new here. I know nothing. Please can you help me to start?

#

Where am i?

hallow owl
wind urchin
#

Yes

hallow owl
wind urchin
#

Ok Thanks

granite stump
#

Does anyone know of a way I can link a doc to Chat? Tried GPT 4 Beta Browsing and it errors out to req login on google docs

trim quest
#

The scraper plugin isn’t working. Can someone else test it for me?

#

It either says load failed or can’t communicate with plugin service

granite stump
#

@steel turtle appears that WebPilot cant interact with interactive links either just tested it on the diablo 4 cc database and getting error "I apologize for the confusion, but it seems there was a misunderstanding. The page you provided does not contain a direct link to a page with the legendary affixes for the Barbarian class in Diablo 4. The page contains information about unique items for the Barbarian class, but not specifically about legendary affixes.

If you have a specific link that contains the information about the legendary affixes, please provide it and I will be able to assist you further."

hallow owl
trim quest
#

Not yet I will

#

Yeah that works

#

Thanks 🙏

sturdy brook
#

i just realized chatgpt4 has plugins.. how do i access this? do I need to be waitlisted or since i am a subscriber i should have access asap

supple violet
#

enable from settings

sturdy brook
trim quest
#

How can I install unauthorized plugins?

#

I don’t have an option for it in my plugins menu

supple violet
#

you cant

wraith ginkgoBOT
winged valley
#

Is it possible to use the GPT-4 API to browse at this point?

hallow owl
gritty folio
#

any1 can tell how to use gpt-4?

#

i mean, i cant access it, some1 told me its no open for everyone as b4

supple violet
#

you pay to get access

granite stump
gritty folio
#

i c ... thanks

granite stump
#

I think I broke Chat Guys, it gave the "error in data stream" error, and when i refreshed it actually gave me an option to resume generating response and all it did was spit out "【21†source】【24†source】【27†source】【36†source】【39†source】【42†source】【45†source】【48†source】【51†source】【54†source】【57†source】【60†source】【63†source】【66†source】【69†source】【72†source】【75†source】【78†source】【81†source】【84†source】【87†source】【90†source】【93†source】【96†source】【99†source】【102†source】【105†source】【108†source】【111†source】【114†source】【117†source】【120†source】【123†source】【126†source】【129†source】【132†source】【135†source】【138†source】【141†source】【144†source】【147†source】【150†source】【153†source】【156†source】【159†source】【162†source】【165†source】【168†source】【171†source】【174†source】【177†source】【180†source】【183†source】【186†source】【189†source】【192†source】【195†source】【198†source】【201†source】【204†source】【207†source】【210†source】【213†source】【216†source】【219†source】【222†source】【225†source】【228†source】【231†source】【234†source】【237†source】【240†source】【243†source】【246†source】【249†source】【252†source】【255†source】【258†source】【261†source】【264†source】【267†source】【270†source】【273†source】【276†source】【279†source】【282†source】【285†source】【288†source】【291†source】【294†source】【297†source】【300†source】【303†source】【306†source】【309†source】【312†source】【315†source】【318†source】【321†source】【324†source】【327†source】【330†source】【333†source】【336†source】【339†source】【342†source】【345†source】【348†source】【351†source】【354†source】【357†source】【360†source】【363†source】【366†source】【369†source】【372†source】【375†source】【378†source】【381†source】【384†source】【387†source】【1434†source】【1437†source】【1440†source】【1443†source】【1446†source】【1449†source】【1452†source】【1455†source】【1458†source】【1461†source】【1464†source】【1467†source】【1470†source】【1473†source】【1476†source】【1479†source】【1482†source】【1485†source】【1488†source】【1491†source】【1494†source】【1497†source】【1500†source】【1503†source】【1506†source】【1509†source】【1512†source】【1515†source】【1518†source】【1521†source】【1524†source】【1527†source】【1530†source】【1533†source】【1536†source】【1539†source】【1542†source】【1545†source】【1548†source】【1551†source】【1554†source】【1557†source】【1560†source】【1563†source】【1566†source】【1569†source】【1572†source】【1575†source】【1578†source】【1581†source】【1584†source】【1587†source】【1590†source】【1593†source】【1596†source"

wraith ginkgoBOT
#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
paper yarrow
#

when will the restrictions be lifted? not known?

tidal turret
#

man

granite stump
#

the amount of prompts per 3 hours?

alpine kite
#

What do you think that chatgpt can’t reproduce a text of the Quran but he can reproduce a texte of any other holy books ?

frozen veldt
#

does that really need an explanation? someone decided on a guardrail there for reasons. probably great ones

rustic nova
#

I had received this email! !!
Why?
Dear ChatGPT customer,

You are receiving this email because we have identified suspicious activity on your account. In order to protect our platform, we have refunded and canceled your subscription. You will no longer have access to ChatGPT Plus service.

We take fraud prevention very seriously, and we want to ensure that our platform remains a safe and secure place for all our customers. If you believe that this action has been taken in error, we invite you to submit an appeal to have your account reinstated. Please visit help.openai.com.

Thanks for your understanding.

Best,
The OpenAI Team

frozen veldt
#

ooph were you trying to jailbreak?

rustic nova
#

What is jailbreak?

#

I didn't do anything, actually I didn't open chatGPT three days ago

frozen veldt
#

then I'd contact the support people

wind urchin
#

nihaoo

frozen veldt
#

Hello 🙂

rustic nova
frozen veldt
fleet tiger
rustic nova
#

Yes, I tried to chat with gpt then he replied:
This key is associated with a deactivated account. If you feel this is an error, contact us through our help center at help.openai.com.

fleet tiger
agile vortex
rustic nova
rustic nova
agile vortex
#

multiple accounts are not allowed and openai are running waves of bans every so often

rustic nova
#

I will sign up with another email and phone number

agile vortex
rustic nova
#

But why!?💔💔
I didn't do anything violated

agile vortex
rustic nova
agile vortex
#

as that goes against TOS (sharing accounts)

rustic nova
#

How about logging in using my phone and pc, is it also against TOS?

agile vortex
#

no, i do that all the time

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

tawdry ridge
#

In the Code Interpreter, image files get downloaded with no file extension, so my Windows operating system doesn't know what to do with them until I just arbitrarily assign them as .png files so I can view the image. This has been happening about a week now.

golden shard
#

Any inside information about the Max 25 questions per 3hrs being lifted?

tawdry ridge
#

Is that a rumor?

golden shard
#

Yes

hallow owl
tawdry ridge
#

I just noticed the ScholarAI plugin. That looks really interesting.

wraith ginkgoBOT
frozen veldt
wind urchin
left peak
#

This is now less useful than BING AI.

frozen bay
#

Anyone know about issues with Show Me plugin not working? are services down or is there a limit to how often one can use it?

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
livid jackal
#

would be nice if UI had a button that said something along the line of "continue with code from this point where it left off" on gpt4 (edited)

sly haven
#

hi, guys, do you know other free chatgpt 4

#

I just only have chatgpt 3. 5, whether its too slow

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

ancient yarrow
wind urchin
#

Anyone know a Chrome extension got ChatGPT that allows me to delete multiple chat history's at once?

wraith ginkgoBOT
novel cargo
#

Any plans to raise the 25 message /3h cap for plus users anytime soon? pretty annoying

hushed plover
#

Guys, I started a chat room with gpt4 but ran out, then switched to 3.5 to continue. How can I switch back now?

hushed plover
trim quest
#

Can someone or is there a plugin to use mini gpt 4 for image input? If not I’ll just wait but I really want to have something integrated into the actual chat gpt

#

Also how can I use open source plugins

#

and when will code interpreter go off of waitlist only?

#

And is the waitlist closes atm?

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

rancid vapor
#

Has anyone connected Selenium to an LLM API to assist in filling out web forms?

echo needle
sly haven
white kestrel
#

use bing, its gpt4 and its free

wraith ginkgoBOT
#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
limpid anchor
#

hiii

untold hawk
#

hiii

rancid vapor
# echo needle if you get a 'yes' or do this i want *IN*

I'm working on it now. Someone suggested Scrapy would be better than Selenium/Chromedriver, for form field enumeration and population for review before submission, and I am testing that. DM me and I will keep you apprised.

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

sand quest
#

a car

lament pagoda
#

Prepare to carry out the start-up meeting of reading promotion in the kindergarten and design a set of processes. ¯_(ツ)_/¯

#

Prepare to carry out the start-up meeting of reading promotion in the kindergarten and design a set of processes.

stiff cedar
#

Why am i not able to choose any more plugins? I chose 3 but after that I can’t find the plugin store

steady belfry
#

211

wraith ginkgoBOT
buoyant lion
#

how can l use chat GPT 4

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

reef tusk
#

Your database is updated until what time?

lunar pasture
lunar pasture
lunar pasture
frozen echo
#

how do i unlock using browsing?

knotty grotto
frozen echo
#

i do

knotty grotto
#

Create new chat with gpt 4 and select browser

frozen echo
#

ty neko maid. i got it now! O:

knotty grotto
#

Have a good uwu day :3

spring jay
#

Anyone have GPT4 appear to respond to an entirely different question than what is asked? This is a new occurance for me. I was asking questions regarding grade school pedagogy and in two separate responses I had Gpt4 first reply with python code, and secondly with a bunch of chemical science "Here are the known naturally occurring elements, grouped by the number of isotopes: Hydrogen (3 isotopes): Hydrogen-1 (protium)". Completely unrelated to the prompts or likely distant of the vectors I'd suppose I was feeding it.

hallow owl
cosmic jasper
trail gazelle
#

Please lobotomize gpt4 more, totally what paying customers need

spring jay
#

It's definitely dumber. The "continue generating" button doesn't seem to work correctly that often and when I attempt to use a continuation prompt (something like: "please continue generating your response from the last complete line") it appears to jump ahead in where the response would be appropriately restarted.

eager creek
wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

spring jay
#

Better change notes from openai would be nicer. April/Mar versions were way more reliable albeit slower.

cosmic jasper
#

Don't need any test, just need to use it lol

#

it takes less account of previous answers and behaves more like gpt3.5

eager creek
spring jay
#

It's like it is forgetting context very early within a chat. Super quick to hallucinate.

#

Have they squashed the token limit for a conversation or stealth redirected us to some optimized gpt flavor that basically sucks?

#

At this rate I'm not sure what I'm paying for as the gpt4 responses are no better than 3.5.

cosmic jasper
#

for all those who don't see it is that you use simple prompts like "what am I going to eat today", for more complicated prompts like code it's obvious

eager creek
#

Not going to believe any of you unless you have the data to back it up

spring jay
#

I'm not here to convince or impress you, thanks.

eager creek
#

Then why say anything

spring jay
#

Exactly: Why are you saying anything?

eager creek
#

Because I find this conversation interesting

#

People walking around like they know everything, and I throw a wrench in their story line to see what happens

cosmic jasper
#

just go up the conversation to see that many of us have noticed it and that is enough to prove it

eager creek
spring jay
#

What does that have to do with our discussion about GPT-4?

#

Definitely having this experience. Not just with code, but general text parsing and practically anything else where I'm providing data and context for it to parse.

wraith ginkgoBOT
#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
eager creek
#

Another possible outcome is that as GPT improves, it does get worse as some tasks temporarily. Then becomes very good at it all of a sudden. My guess it's it is a trade off to get better in certain areas like a long term investment to increase its chances of understanding more later on.

#

If you read the GPT-4 paper, you'll see this

spring jay
#

For the rest of you: Probably stealth token limit nerfing. At least via the chatgpt interface the system appears to be incapable of tracking more than a few thousand tokens. After that, the chat session is worthless. Using gpt-4 and gpt-4 browsing at least. I do wish that OpenAI was transparent about token limits and utilization within the chat UI. Would be nice to get API access and just interrogate it that way so that I don't have to rely on this interface.

eager creek
#

I wouldn't assume anything. But I'm pretty sure in the long run, transparency will be key.

#

I can certainly agree with you on that

spring jay
#

Oh, well, there we go: Got API access now. Wee!

wind urchin
#

Does anyone have nuclear launch codes?

silent wyvern
humble maple
#

TeslaCoin is the new revolution in making millions and it's not yet published. Ask me how for more info.

eager creek
wind urchin
silent wyvern
#

:blackgentlemaninsuit:

eager creek
silent wyvern
#

not if the SCP Foundation contains them

eager creek
silent wyvern
#

good theory

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

young crystal
granite stump
#

well this chat devolved pretty quickly

frozen veldt
#

these channels tend to do that lol

granite stump
#

I cant say if GPT4 has degraded in responses as of yet but Beta Browsing is buggy af, nearly prompt returns a "error in message stream" I did have it hallucinate once yesterday which is unusual, and its memory is most definitely degraded, for days now I have been trying to get it to read database info I either have it parse online or I manually enter and usually can get no further than 4 prompts in before it goes full dumb and eventually gets stuck in a loop of regenerate response.

#

Curious if this is due to the "error in data stream" error, as you have to refresh the page to bypass the regenerate response issue

left peak
frozen veldt
#

it's that context limit

fluid stirrup
frozen veldt
#

no matter how you structure the information it can only hold those 4000 tokens in it's head. Beyond that it starts to assume (loss of fidelity)

fluid stirrup
#

your an biggot

granite stump
#

I think this is due them trying to reduce cost and appease the people complaining about response time, I would prefer to wait longer knowing my prompt isnt wasted or the response isnt flawed due to some perceived need to hurry before the human gets mad or bored, I have had to tell it to ask me in the response to continue its task to give it more time to think.

sterile verge
#

Hello, I have set browsing with Bing enabled in settings but ChatGPT4 is still not capable of browsing, does anyone else have this same issue ? I see many people are using browsing so I'm thinking maybe this feature is blocked for Euro customers maybe ? I tried private and non private chatting but it is always the same ...

left peak
#

All they need to do is be honest and explain these actions and updates rather than just nerfing the GPT 4 product via stealth. The last 24 hours has been extremely frustrating and I am sure there are 10's of thousands of dissatisfied users who do not know what is going on.

#

I have wasted almost an entire day with it.

hallow owl
sterile verge
main lava
#

what plugin would be best to use to make gpt 4 read my code

hallow owl
elder axle
#

weii

sterile verge
hallow owl
hallow owl
granite stump
#

then check

#

if that doesnt work close the browser and navigate back to Chat and check if the option is present

sterile verge
# granite stump refresh the page

Thanks for the relase notes, they cleary state I've done anything I needed to do, but it still doesn't work. I did try refreshing and relogging and stuff ...I'll continue trying, but would really be bummed if I'm wasting this energy just to find out there some filtering on customers based on location

granite stump
#

Just tested it again.
"Chat I am going to manually enter data I need you to strictly adhere to some rules here, I need you to ask me before performing any searches on the web unless i specifically state you are to perform one, when searching on the web make sure to cross reference all data across 3 different sites, if the data can not be cross referenced after 6 sites have been searched then keep the most current values, if you fail to read or click something try 3 times, if on the 3rd try it still fails move on, make note of any information or fields of data you are not sure of or understand and present that info in the response in a way that would assist me on helping with understanding the data, if I cant help i will instruct you to perform a search for it and do so. Do not perform a search for previous prompts that you have already searched for unless I prompt you to in the most recent prompt. To start I need you to search for how these diablo 4 game aspects work, skills/abilities, affixes, paragon boards, gear, leveling, talent tree, pay particular attention to any data from the barbarian class as that is what we will be working with."
on the 5th prompt it searched without asking me to perform a search, so definitely limited memory.

granite stump
sterile verge
balmy gale
#

Is anyone learned Andrew Wu‘s deep learning class？I’m new here.

hallow owl
sterile verge
hallow owl
granite stump
devout thicket
#

Hi everyone, are the plug-ins definitely worth it? Kind of making me question whether I should upgrade or not

whole kettle
#

@hallow owl come a bit to off topic wanna show you something ig you got time

frozen veldt
sterile verge
# granite stump you could try to social engineer it, depending on your question lead it to belie...

I understand what you mean but it is a waste of energy, I'm also not searching for anything really controversial: Examples : Bard stoped giving me the URL for the Google AI music engine because it thinks copyright is a problem (like what ?!) Bing Search - stoped printing out the result of the summary story from Douglas Adams HG2U, because the characters "made love" on a wing of the plane ... It's just depressing to see how over-censored they are

hallow owl
devout thicket
#

Is there actually any use for GPT 3.5 when you upgrade, is there any point in still using it anymore? Are there some things it can’t do that GPT4 can and vice versa

hallow owl
devout thicket
#

oh wow, that’s a bit dumb

granite stump
granite stump
devout thicket
#

I wish they offered us a trial, even if it’s just three days

#

I just need to see if it’s gonna be good enough for what I need it for

whole kettle
wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

spring jay
cosmic jasper
austere flame
granite stump
sterile verge
hallow owl
sterile verge
#

..i did test this before and it didn't work either, so maybe changing the browser did it..

hallow owl
austere flame
lunar glen
#

Can someone point me to how I can use plugins. Looks like I have the option to include?

austere flame
#

Add the plug ins you want to install. you can install as many as you like but you can only use 3 per session. to use them you have to highlight the box next to the plug in once your installed it.

brisk tinsel
#

Can the plugins be used with gpt 3.5?

frozen veldt
#

no

dim lark
#

Hi all, I am looking for a prompt to create an effective system role as an input for the Chat mode for OpenAI. Can somebody help me here?

granite stump
dim lark
#

Thanks! In essence, I would like to use to craft GPT system role descriptions to use in "Chat" mode. However I would like to have GPT creating the desciptions for the system roles, e.g. "Act as a financial advisor with deep expertise in investment for the last 30 years, focusing on global stock markets ". I have done something very similar with MidJourney prompts, where I can just provide a simple explanation for an image to be generated, but GPT will provide me with a more detailed prompt to generate the image. Hope that helps.

#

So in essence, a prompt to create other prompts with a focus to create system role descriptions.

granite stump
#

So why not just do as you just said and ask GPT to create a prompt for itself? If you have access to the browsing Beta it might be worth using that to ask it to search the role across say 5 pages learn what the role entails and create a prompt for itself to perform those task?

frozen veldt
#

you may run into guardrails

#

for example, when I ask gpt to assume the role of a stock analyst and pick them, it balks

#

but if I tell it to assume the role of a data analyst and produce an analysis of data and rank it.. no problem

#

even though its the same data set and same result

granite stump
#

This new prompt could be used on 3.5 as well but the response from 3.5 will be less desirable compared to 4, at least used to be atm both are comparable i guess

dim lark
#

So something like "Act as system gpt role expert and create me a description for a system role focused on [topic]"? or just use the browsing beta to learn what the role entails.

granite stump
#

the trick is getting it the data on what the role is beyond a basic understanding so I think using Beta Browsing will yeild a far superior data set to engineer a prompt from

dim lark
#

ok that sounds like a plan, will try the beta browsing. Thank you!

cloud geyser
#

1.Click the "Uninstall" button

2."KeyMate.Al" plugin should be uninstalled

3.The page prompts "Error uninstalling KeyMate.Al Search."😓

granite stump
#

something like "Chat I need you to perform a search on the role "example" search what the role requirements are across 3 job posting sites as well as search for blogs or other areas that share "example" role and get a deep understanding of "example" role, feel free to search all sites on the first page and if you need to ask a question to further understand this role include it in the response."

granite stump
civic sparrow
#

so question for you guys i just asked 3.5 to provide video links and it linked me to a 2023 video.... is there something up with the knowledge cut off date that was really odd?

dim lark
granite stump
granite stump
dim lark
somber dagger
#

what dose this mean "Error in message stream"

granite stump
somber dagger
#

ok cool. I was going to just want to make sure

granite stump
#

current issues with Beta Browsing and hitting refresh on the browser will save you a GPT4 token being used vs hitting regenerate response

somber dagger
#

oh. should I not leave the tab while it do it

carmine marten
#

Hey, this just on my end or have the GPT-4 limits been lifted? I am no longer seeing the 25 messages every 3 hours message.

somber dagger
granite stump
somber dagger
granite stump
granite stump
# somber dagger ok cool. I am using Win11 laptop 😄

then you should be fine to hit refresh and go about your task and come back, refresh browser not regenerate response when it give "error in data stream" though be aware sometime a refresh wont fix it and give you the text input box then your forced to regenerate response or kill convo and start fresh

somber dagger
#

great thank you for the info

carmine marten
wraith ginkgoBOT
#
This is a Safe for Work Discord Server

The OpenAI Discord is an actively moderated server.
• Refrain from sharing inappropriate content on the server. This includes but is not limited to messages, media, or other topics of graphically violent, sexual nature, and drug-related content.
• Report all sensitive and offensive content in the feedback reporting tool in the ChatGPT web UI instead of here on Discord.

Posting NSFW content will result in a permanent ban.
granite stump
#

It may be due to them tinkering with stuff on the backend as GPT has been weird for last day or so

somber dagger
#

Lol I got the same message 3 times now

carmine marten
#

Yeah, they seem to be trying to make some kind of ChatGPT4-Bing Frankenstein's monster and market it as ChatGPT4 with internet browsing, and it's going about as well as I would have expected. The internet browsing beta is nigh useless. :p

granite stump
carmine marten
#

I've tried to work with it same as GPT-4 vanilla and it literally become so confused trying and failing to read the same weird, irrelevant website over and over again it ended up just saying that it was having some sort of problem and couldn't answer me. I had to start a new session without the browsing enabled. For the moment I'm using web page reading plugins and doing the browsing myself if I need to.

granite stump
# carmine marten I've tried to work with it same as GPT-4 vanilla and it literally become so conf...

It likely has Java elements it cant interact with I have been battling with trying to get it to read D4 databases for days but seeing as how most of them use Java or are in Json format, GPT cant understand it as it doesnt have all the info it needs to know the variables, or the data is more than the allowed limit and I am not savvy enough to get it in there in chunks just small enough to maximize prompt usage. trial an error realy

carmine marten
#

@granite stump Hm. Well hopefully they'll work it out. Too much of a pain for my use case though, faster for me just to parse through website content myself and copy paste anything especially relevant to ChatGPT.

granite stump
sonic night
#

do you also get error in message stream all the time when using browsing mode?

#

text is finished but it turns red and can't prompt further

granite stump
somber dagger
#

that is what I am getting, Even once I refresh I cant promt again

granite stump
#

it will save you a prompt token, unless the refresh doesnt work, sometimes swapping to another convo and then back can fix it as well

#

if it doesnt then you will have to regen or in rare cases kill convo and start fresh

#

not so much a fix as a last resort start over

#

I dont like unaliving my convos before their time

still lagoon
#

Did someone already tried sharing convo links from the GPT-4 model with free users without access to GPT-4?

granite stump
#

yeah I have seen this creep up in my DB parsing trials, i think its got a lot to do with token limit

wraith ginkgoBOT
#

Beware of possible scams or fraudulent activities that you may receive through direct messages. OpenAI staff will never DM you for any transactions.

Please report any incident by sending a DM to @open crescent immediately.

idle furnace
#

any idea if an option to develop and test your own plugin is granted via some wait list?

#

i saw in some of the youtube videos on chatgpt plugins that they have this option to develop and install unverified plugin

hallow owl
idle furnace
granite stump
idle furnace
#

i don't see it

granite stump
#

fill that form out and in a few mins you will get a confirmation your approved no waiting

idle furnace
#

Thank you for joining the ChatGPT Plugin waitlist.

To balance capacity with demand, we'll be sending invites gradually over time.

– The OpenAI Team

warm radish
#

GPT-4 has gotten much worse recently. It keeps getting stuck in loops where it writes a partial response and then it starts over and writes the same exact thing over and over again. This completely breaks the ability to use the "Continue Generating" button.

#

And when I tell it that it messed up and to try again it just does the same exact thing again

wraith ginkgoBOT
warm radish
#

Same if I use the Regenerate button

#

Not sure what to do at this point, it keeps wasting my GPT-4 messages and I will hit my limit if I keep telling it to try again.

#

It never used to happen in GPT-4

blissful belfry
warm radish
#

I am in too deep at this point to start fresh

blissful belfry
warm radish
#

Oh for respones?

#

Nevermind I was thinking of messages that you send

blissful belfry
warm radish
#

Right now my issue is with the Reponse to my prompts infinitly looping and never giving the option to continue generating the original answer because instead of getting stuck like it used to, it just starts over again giving the same response over and over.

main lava
#

is it possible for gpt 4 to give them my code ? like so that they have full context?

eager creek
#

What on Earth are you guys doing to brick GPT. It's working fine for me

warm radish
gaunt juniper
#

IMO if you are using gpt it isn't designed (yet) to do all the work for you, it's way more efficient to ask specific requests, after countless hours of using gpt to help write scripts/code I've realized it is way more productive to go step by step and always verify gpt is on the same page. Ive had instances where I give gpt a large request and it struggles or even forgets to add what I need.

blissful belfry
warm radish
blissful belfry
blissful belfry
eager creek
mighty kindle
#

I used it while doing homework and got some things wrong

eager creek
#

Everyone should just mass leave academic institutions. It's over

#

In the face of this technology. I'd get up and walk out of that classroom yesterday

blissful belfry
#

I can't send pictures in this channel?

hallow owl
blissful belfry
#

they really need a better web browsing feature

junior flint
#

Some people are too over excited of GPT4, i have canceled my subscription for Plus and I've been paying user for over 3 months I just don't think it is worth paying more than Netflix for what you get now, Usually you spend more time screwing with it while you could create a better code in the meantime.

eager creek
blissful belfry
#

my GPT4 is full of my homework, especially essays based on completion, that's the most helpful part, sometimes it teaches math, works pretty good since I'm always having a hard time working with textbooks

junior flint
blissful belfry
#

is this a problem with the plugin?

hallow owl