#💬general

1 messages · Page 101 of 1

dusky juniper
#

What the

unborn ridge
#

From using ai to check and evaluate my own thoughts and ideas, I find it's correct about 80-85% of the time, and incorrect in small or large ways 15-20%. Not catching these small errors would lead to very large errors over time, so you really need a human in there to evaluate things and catch mistakes, especially in an iterated process. And due to current ai operating on probability essentially, it's destined to make a certain amount of mistakes as they output educated guesses not hard algorithmic proofs. You can maybe have competing ai to try and catch errors, but it might not catch enough errors and its probably very expensive.

simple plank
#

i love how llm can waffle, but can also summarise waffle

AI is useful, but human oversight prevents costly compounding mistakes.

unborn ridge
simple plank
#

99% of stats on the internet is not evidence

cunning nimbus
simple plank
#

naw mine is the 1% trust me

#

generating infinite naughty pictures

#

and coding

#

look at them naughty function signature shapes

#

I didn't use strict type enforcement

simple plank
#

llm's love using the words, "neon" and "nebula" [and poof the self-bot disappears]

placid crest
#

hello

tawny trellis
trim bone
#

probably been asked to death but, any tips for chatgpt to stop mixing my attached codes for older ones, or telling me I uploaded the wrong code/document when it's not like that?

tawny trellis
trim bone
#

well, it's been happening, i upload three times the same code/document, and still insist is the wrong one, or some older that doesn't have the newer changes (when in fact it is the most recent updated one)

#

I didn't understand a single thing you said, sorry lol

pliant blaze
#

prompt engineering courses are a scam. the only way to get better is to use common sense and describe clearly what you want, and also to practice

#

you can read 100 prompts, but its not usefull until you actually learn to apply it to your own usecases

#

"yeah im not self promoing"

#

least obvious alt account reviews

ebon gyro
#

A few experiments made with Uisato Studio’s Seedance 2 "intelligent" implementation, exploring how far generative video + fine-tuned orchestration layers can be pushed in rhythm, camera language, body transformation, and audiovisual sync.

https://uisato.studio/

Treating AI as a new choreographic tool beyond the limits of expensive gear and t...

▶ Play video
teal spruceBOT
#

test again

crude nexus
#

@obsidian torrent

cunning nimbus
simple plank
jaunty flame
#

Does anyone here use Copilot Tasks or Projects?

pliant blaze
viscid nest
#

hi guys

sonic smelt
hard blaze
#

Hi, is there anyone who is looking for a developer?

latent hemlock
#

Hello guys , i have a final exam in like 3 days and I really really need to learn how to use chat gpt so that he can analyse all the old exams and understand the structure then help me to revise effectively

jaunty flame
pliant blaze
#

yeah

elfin mirage
#

Hello, Idiosia by Entropy

pliant blaze
#

@oak shell???

unborn axle
#

🚀 I am looking for startup founders, AI entrepreneurs, and tech CEOs to be featured in founder specials, media articles, and co-founder specials. I am available to chat anytime if you need help in the fields of technology or AI. Let's exchange experiences by discussing the validity of new ideas and the skills needed to actually execute this vision Please send me your startup information, website, etc. via DM

low patrol
#

Does anyone know where/how to shop or find cool clothes

#

Like with cybersigilism

oak shell
cunning nimbus
pliant blaze
#

@uncut island @dark void

latent hemlock
stark tree
stark tree
stray wolf
#

any suggestions in tools or prompts i could use/resources for writing a professsional CV? never had a job, been self employed since 11 (very successful)

timber yoke
#

Hey guys, i just joined!

simple plank
#

liar... (no clover)

#

that wasnt there a minute ago

#

hacker!