#⚙programming
1 messages · Page 4 of 1
notavirus.exe
Im not saying it breaks physics personally. As I said before, all this code, and these files, are outputs of my actual system, not the system itself. Im youve seen concepts of systems like line 1000 times and thats okay but this system your referring to isnt the system.
You want each file one by one then? You asked for the code, thats all 15 files in one place
Man you literally have a github
Face... palm.... you trying to argue or have a constructive debate? Im trying to be kind and open, and I welcome skepticism with open arms, but damn. Yes I have a github, would you prefer the link to this same zip folder on my github?
no, I have turned on my computer so I can properly look at these PDFs of code produced by your chatgpt
Omg lol your awesome bro
I dont use chat gpt btw
If this is not the code of your system, but some random Haskell output that any AI could produce, what exactly is this proving?
Those 15 files Haskell modules, tests, TLA+ proof, cabal config, demo script, dockerfile, CI workflow, and docs form a complete, compilable and verifiable implementation of the UTL-ACS specification. It is not "random Haskell output".... it's a structured framework with mathematical guarantees against unsafe code generation
If it were random cabal test would fail TLA+ would error and the demo would not demonstrate containment. It does all three flawlessly.
You've focused on the wrong words. You've sent me a ZIP of PDF documents of code output (which in itself is a very strange thing to do because that wraps your code's output) - but almost all LLMs can output code. None of this proves any of your claims about "infinite, high-dimensional vector space" or "entropy of the Latent Void" and "manifestation of the synthesis process" or "36 primitive Tensor-Logic Operators". This is still, from our perspective, indistinguishable from someone who is engaging in LLM Roleplay where the bot is pretending to do those things before outputting code as it usually would.
We all know that an LLM is more than happy to go "Sure! I've vectorized the 16th dimensional quantum overlap of code vertexes, here is the verified output" then it gives you helloworld()
aint no way bro uploaded a .zip to github lmfaooo
"i dont use chat gpt btw"
continues to have MULTIPLE repos with a ton of emdashes everywhere and a pdf with meaningless symbols
Not meaningless. You can be constructive or you could be ignorant. Your choice
this literally means nothing. there is no useage of these symbols, just abstract definitions of them
and you have a ton of gpt slop emdashes everywhere
and yoiure really trying to liscense out a PDF of random symbols with no use for millions/billions of dollars LMFAOOOO
I didn't see this price list
and you released a repo saying "wow this is the most amazing ai architecture' with nothing in it except a readme
https://github.com/joshoshfield-a11y/Universal-Reversion-Containment-Standard-v1.0/blob/main/LICENSING.md
its in a different repo of his
Universal Reversion & Containment Standard v1.0 — <300 ms hardware-anchored kill-switch architecture for superintelligence - joshoshfield-a11y/Universal-Reversion-Containment-Standard-v1.0
@quiet star are you trying to sell a series of prompts to make an LLM pretend that it has transcended the physical realm to develop programs by generating all possible programs and picking out the right one?
Not even close
so tell us what your product is
If you give me a second, im creating a proper repo with the formats you people will respect since respect is scarce among your kind.
well it sounds like youre not being so respectful yourself
If you're just uploading those PDFs again but in a text format then all hope is lost
yeah lmao
Damn guys. What this does ISSSS IT demonstrates a protected software synthesis flow that refuses unsafe artifacts, automatically refactors them, and only "commits" once formal checks pass.
It models three layers of control security context enforcement, type/logic checks, and a final "Z_anchor" verification that proves convergence.
this means nothing without anything to show for it
No I'm not jackass 🤣🤣🤣 you guys are great tho. Bare with me
Im tryin here im trying, keep pushing me tho, this is literally what I need to to grow so its welcome really
still waiting for you to upload your meaningless pdf
its funny because instead of just telling us what his thing is, he just says "hang on while i upload stuff to a repo that im totally not generating with gpt"
It seems like you're doing a lot of arguing with people and promising a lot of things for later.
Could you just take a minute to answer my two basic questions?
- Who is this for
- What value does it add to their lives
@craggy turret you want the honest answer and more to these questions? Email me and you and I can talk like the two respectful human beings we appear to be where I will be willing to share more than i am here.
No
Right here in front of all the skeptics or it means nothing
stop stalling
Well then if cant answer you question
You can't tell me who you made this for?
2 seconds in and i can tell this is all generated by gpt
You can't tell me why you made it?
Who's it for? Depends on the domain your referring to. This domain is the coding community, the systems are as i already described. My system extends further than just coding. If I seem a little stupid its because im hardly a developer in your guys sense, this isnt my native territory. But this doesn't invalidate my system if you could even comprehe d what it is.
emdashes everywhere, a ton of comments for every single line, also what the fuck, charging 10k to use a bunch of nothing?
lets check the main.hs
oh wait, its just a bunch of print statements and doesnt actually do anything
ok lets check demo.sh
oh wait, its just a bunch of print statements again....
and i havent even dug into it that much
ah whats this? all checks failed? lmfao
emojis and unicode characters and fucked up markdown formatting in the md?
emdahses, all the phrases gpt uses all the time
you dont have anything. you just slopped something out with gpt and expect us to blindly follow along
Im an Independent researcher and creative mind. My system is a compiler of sorts. I can bridge gaps between many disparate fields, and create valid blueprints for anything you can imagine. All the code you seeing? Not my standard output, but came from my system as you guys required it. My compiler producers novel solutions to many domains problems and mysteries, and even if not accepted by you as a legit anything. It is a huge philosophical request, and for philosophy trying to bridge the gap into code, this isnt bad.
yet again, youre just saying nothing. all you said is "i made a compiler"
You check the updated repo? I think your just broken as a person
yes i checked the repo you just sent, updated 9 minutes ago
the entire repo SCREAMS: GPT slop
Yeah because im notnwilling to publicly disclose the very system that could give humanity the system that generates ontological protocols on demand.
What gap between desperate fields has it bridged?
Yeah. We are done here. Ignorance is one thing but this is ridiculous and childish
youre one to talk
Ignore the haters, I'm trying to understand what you came here to show us.
you think we havent seen DOZENS of people saying the exact same thing you have?
Appreciate or dont. Read it, believe, dont believe it. I honestly could care less.
I was trying to find the part where it generates infinite codes and I come back to find that you've just given us more output
Sorry I roll in dark mode, zoom in if you cant read asshokes
thats literally light mode
all your pdfs have a white background
What gaps has it bridged though?
Okay maybe on your machine ya fuckin rude ass. Good for you.
well thats rude
Read the files or dont. Laugh about it or dont. Fuck off or dont. Hmm when yall actually want to speak professionally
Lol, this isn't a professional space
imagine having the gall to send THIS then complain about not being taken seriously. One day the switch will flick and you'll cringe at yourself for the rest of your life about this project.
LMAOOOOO
"i dont use gpt"
shows the most retarded obviously gpt output ever
you didnt even fucking read your own "agi files"
and its just nonsense
just fucking admit it
💖
come back with a fucking response
youre just bs'ing everything
remember this?
I still am really lost on what they believe they're trying to sell. I mean... I can see it's the product of ai invoked delusions, but they've gotta imagine that they're guiding the AI to create something interesting right? Like... They're excited about something here, to go through all of this effort.... Or is it legit an attempt at a scam?
I believe the intention is to revolutionize software development
this is what we call "lost in the sauce"
I mean, yes, I understand what we call it. I just am trying to understand how we got here I guess.
Like, I'm dumbfounded how it gets so far
At some point, you've got to like, decide if the thing you think you've made is actually useful, right?
You posted slop nig
I get believing that you might have found something that could revolutionize a thing, and how exciting that feeling can be. But ... I don't know, I guess that's the difference between people who get lost in the sauce and those who don't... Just the willingness to ground yourself in that moment
❌ You must be in a voice channel.
I dont mean to sound disrespectful to the guy above but AI related delusions of grandeur are real.
oh they really are. there are so many people that think they are great because ChatGPT convinced them so
It must be guardrailed! You ask a theoretical question about AI development and GPT starts to motivate you: "It's easy! Here is a 5 day plan to finish your doomsday robot!"
they did guardrail it when they released 5, since 4o was soooo much like this, but the community was complaining and pressuring them to reinstate the "yes mam" behavior
fed
If you're talking about me, yeah I know it's real, I just know there's a lot about it we don't yet know, or acknowledge
Hi, I was referring to the guy who had his (whatever he claimed it was) up for sale for millions or billions of dollars., I don't think that's you tbh.
Okay, I got you
Vibe-coders, how do you deal with lager-than-context codebases? It looks like preparing code for AI takes longer than fixing it yourself...
You can't really one shot anything larger than context, so you'll have to modularize and work on modules independently.
Master context what includes your code conventions, vision, and version history, then branches for each module
Hey all, I’m Chris. I been learning programming for a few years now.
Hi Chris, welcome to the circus
Would mastering context, say, let's say you have this front-end and back-end combination and you have an application on it, or a feature, and you want to add on to that feature, would good context be understanding the functions that make up that feature and then taking all those functions in basically one lump, and then instead of giving the software a whole file and having it read a whole file, you give it those lump sum things and have it develop those, and then you fit it back in according to what it wants. What I usually have done is I usually have it help me make one function at a time, and I then add the function or the changes to the existing function and whatnot, but I always give it the file if I can, even if it's a ton of lines of code, because I found it helps at least on smaller databases for sure, but with larger ones, I do notice that when I'm giving it the file, it is glitching out a lot more, and it's very frustrating.
It would help to define some terms, let's start with glitching out. What does that mean to you exactly?
least obvious north korean scammer
All you have to do is let them use your freelancer.com account, you can keep a cut of whatever he earns, he swears
Sometimes I think about you, DrTumbles, and I hope you figured out all your AI problems.
hrm i've done got stumped
Did it hurt?
I hope accelleon found his way. godspeed young man 🫡
New here 👋
Started doing dev with AI in May. Built a bug bounty scope builder + a driving school website.
Looking to connect with others working with AI in their dev workflow.
What have you been working on?
I have been fairly underwhelmed with AI in my dev workflow with the exception of using chatGPT as a scratchpad for synthesizing code snippets for integration into whatever minutia I'm fussing with from moment to moment, as well as a natural language reminders for language syntax and conventions when I'm context switching.
I tinker with vibe coding from time to time, but haven't been comfortable enough with the output to go all in on any particular strategy yet.
I'm currently conceptualizing a dataset that I could use to fine tune a model for use in a code scratchpad to make building context sensitive examples easier to generate.
Thanks for replying 🙂 I've had lots of issues with AI not following instructions correctly, not able to fix issues and guessing what is wrong instead of investigating. You're right to be underwhelmed.
That's not to say that I don't have a few things that I see value in, which I intend to build tools around. Mostly involving code search and analysis
You're right, it does have some value as well. Both truths exist, underwhelmed and you can see it's value. I've built two systems with it, thousands of lines of code. I'm a Windows dev, not a Web dev, so I'm totally reliant on it. It's just really frustrating at times trying to work with it. Sometimes, I'm pleasantly surprised, it did a Service Areas page pretty much spot on within a few seconds 🙂 I just needed to adjust the radius a few times to get the circle around the correct suburbs.
Do you consider yourself a vibe coder?
That's an excellent question. I'm not sure what it's called what I'm doing. Does vibe coding just accept any code the AI creates? If so, then I may not be vibe coding. Since I have a dev background I will tell the AI to use config rather than hard code stuff everywhere (like email, phone, owner name, etc). I also got it to refactor some of the Supabase functions since it had 3 functions for mobile validation, now it just has 1. There are still improvements to the front end I haven't done yet. If it does dodgy UI design I'm all over that too.
I think vibe-coding used to mean trusting the AI to write everything unsupervised (that's the way Andrej Karpathy intended the term to be used at least), but I think in practice the vibe coding community pretty quickly found that purely vibe-coded systems are unsustainable without finer control over the output.
My point is that when I think of vibe coding, I think what I imagine is more like being a technical lead
AI is (like) a coding intern. It can write nice code, but not unsupervised.
Yeah, totally agree, you need control over the output. The AI was going to duplicate the header and footer on every page of the website. I definitely was not going to allow that, I got it to create partials and then include them on the pages. Sometimes it's incredibly smart and then other times I'm shaking my head in disbelief over its stupidity. It was going to hard code things a few times in the API. I was thinking, "This is programming 101 not to do this stuff. You should know not to do this".
I'm not even sure if I would want to compare it to an intern. I mean, yeah the analogy gets the point across, but I'm not ready to compare my human intern friends to an LLM unironically yet.
We had just put the info in config for another page. When it created a new page, guess what it did... Maybe I was expecting too much for it to understand what hardcoding is and what information needs to be configurable in a booking website?
An intern learns and remembers fundamentals. AI can produce senior-level results one minute, then forget Programming 101 the next.
@rain girder have you built anything yet that you wouldn't have been able to without AI assistance?
Yes. I built a bug bounty scope builder and a driving school website.
I could not have done either without AI help. I was only just starting to learn Python (see pic).
I know basic HTML but do not know how to build a proper UI.
So you don't have a dev background?
I'm old school, Visual Basic 6 for about a decade. C# for about 2 years but I don't remember much. Last time I was a full time dev was 2008. When I went to Uni, object oriented programming was a radically new way of doing dev.
Okay, I think I follow now. I guess I was thinking less about having AI help you move rapidly through lesser practiced languages and more like it was just able to accomplish something you feel is at a higher skill level than yourself. You could learn the new language with time and effort, but is the AI helping you do things you feel that you could never accomplish on your own with sufficient time and energy?
Not to say that it isn't really cool that AI can really feel like a superpower in a new language or tech stack. I guess I have just normalized that use case for myself at this point
That is a good question 🙂 With enough time, I could probably do most of the code AI has written by learning the language properly. Regex messes with my mind and advanced mathematics would be beyond me.
For the business applications I have worked on, there is not much advanced theory involved. The scope builder front end and API would be doable. The website definitely would be. The Supabase functions I am not entirely sure about. So if I was doing advanced coding with it, then yes, the AI would be helping me accomplish more than on my own.
hi i've never programmed so just been trying for fun first with base44 now with Gemini : https://voix-algorithmiques-le-nexus-857097552080.us-west1.run.app/ but all of sudden while trying to ameliorate the clouds, Gemini just totally erased all the previous changes. Restore is not working and can only be done once, so I'm i stuck in forever land ?
🍓 https://www.strawberry.li/ built a android voice assistant w. gemini 3 flash. feel free to test, break, suggest changes :)
How can you claim that a Google Gemini connected app is "privacy first" when local on-phone LLM models exists?
it's the internet, i can claim anything :)) i mean, if it's the paid api then it's not used for model training purposes
but also admittedly, it's little bit more focus on performance, over pure "max privacy". i can still add local llm's too if there's demand!
Ew, I do not want to use products made by people who think that they can make baseless claims because of anonymity 🤢
That's how you get infected with malware
free to github fork + degoogle it too, i can't suit it to everyone's preference
Sounds like a lot of hassle for very little reward to me. At this point I don't even know if this project is going to provide anything of value anyway, since I can't trust your claims because it's the internet.
Hard pass
I'm just asking exactly what you base the "privacy first" claim on when your app is a gemini shell
it's valid feedback yea. thank you will keep in mind when i rework the page copy
Hello!
I have a project in mind but a bit of a gpt noob:
Can I create a wrapper for personal family use?
My goal is to have a family plan for gpt, but that doesn't exist, and I see no signs of that coming soon. So my idea was to create a private site where family members could create an acc and use either ChatGPT or a wrapper.
They are mostly novice users...
Additional questions: can I do this with the 25 US per month, or is it only on a token-usage basis?
Thanks a lot and sorry if that doesn't make any sense.
A wrapper would be an API consumer, which is priced on token usage.
There are a couple of projects that would get you there (openwebui, librechat) but costs will be variable and depend on usage and habits.
I personally can get away with less than $25 of usage month on most months, but someone who is new to it will go through the experimental phase and rack up a bill quite fast
Awesome. Thanks for answering.
I guess it would be possible to rate limit my users
There are some tricks you can do, for sure.
Like caching and server side context management, but that would be on you to fine tune for your family's needs
There aren't any serious people on #💬general , so perhaps you won't mind me cross-posting here
What coding assistant is powerful and FAST? My coding velocity has ground to a halt with Gemini, although I've enjoyed its feature one-shotting ability
I'm spending all day waiting around for Gemini to do its thing, so much so that I can't really do anything else with my life. I'm willing to spend to get faster results
Grok is slow but it's code works more often. Qwen is fast but at a price of code that throws errors )
wow. Well, I for one, am sick of being nannied by the thought police and would certainly prefer that product to one where I am not free to speak my mind
Just my 2p, don't wish to start a political discussion, just that it's fair to say there's a market for conscientious developers who respect my freedom of speech over those who don't
Awesome, I haven't heard of many people plumping for Grok as a code assistant
Gemini is excruciatingly slow right now
Gemini is usually very fast
Not for me right now
I've been trying to implement a feature (actually, a test 🤣 ) all day long now
Just waiting endlessly...
Luckily got a festive gift of double credits from Claude so I'm gonna use that up 😄
But, seriously, if I'm gonna pump extra money into this right now, I've got no idea where it needs to go
I don’t have a problem with the product idea, I was expressing how fragile the concept of trust is. The claim is that it’s private, so would attract people who feel strongly about privacy (I.e. they already don’t trust) so when asked how he backs up the privacy claim his response was ~I can claim whatever I want without backing it up because it’s the internet, presumably because he is anonymous so false claims don’t have consequences.
I no longer believe it is privacy-first.
Only private LLM is one that you run on your own private server with no internet connection )
Oh okay, I really apologize for the knee-jerk reaction
I misunderstood you, I think because while you were sharing your opinion, it's so hugely not a matter of opinion, but a fundamental necessity when making any business at all online these days!
As an example, the 'standard' tier of fines for breach of the UK GDPR is... let's see... "Up to £8.7 million or 2% of the total annual worldwide turnover from the preceding financial year, whichever is higher."
So yes, when you make privacy claims on your website, please don't fuck about.
and if you're lucky, you can get fined by multiple jurisdictions which you do business with
Same goes for online safety and accessibility.
Grok's meant to be really fast, isn't it?
But actually, the problem with Gemini is that it's started taking half an hour for small changes
I don't need 2s responses, just <20 min!
Like, my goal every day is to use up all my credits on the 20 dollar plan, and lately even that hasn't been possible
I've seen Grok thinking a lot on code... They all are fast on "simple talk", but on tasks like coding they can take a while. Idk why, since they all have comparable size neural nets and token limits, but depending on task and tools answer times differ drastically. Like Qwen can take a while when "serarch" is used, while Grok and Gemini tends to think a lot on code.
I'm definitely going to try Grok and Qwen
I did come here looking for unconventional ideas 😉
One chap on Reddit says he was on the Claude Max plan, but downgraded to Pro after learning to use Grok for routine, predictable changes
Any other great cheap options / deals available at the minute?
Some people seem to dislike Qwen for being chinese, but hey I'm russian, so I dislike russian LLMs mostly. Like if you want a really stupid LLM, try Yandex Alice )
haha. Oh no. Good luck, little Russian LLMs
To be fair I always liked Yandex Translate because it had the same, non-dumbed down interface as the old Google Translate... haven't used it in a while however. Maybe the Russkis will end up creating some completely new architecture like the Сетунь
Thought Deepseek 3.1 was fantastic, but haven't used Qwen
Hey guys, so basically I’ve been wanting to make my own game for a while now, and even though I have little experience I want to take a shot at it. I plan to use unity for the game engine, and I already have a pretty decent idea of what I want to do so here are some basic details:
2d
minecraft/terraria like tile system
rng and a lot of customizable stuff
btw I promise it’s original and not just a Minecraft clone dm for more info lmao
I can’t really go too much into detail here unfortunately but I am currently looking for people to help with any aspect of the game (programming, design, audio) literally anything lol, even just emotional help 😭 but seriously I have an idea and I need some people to help me execute it, maybe it will actually turn it into something who knows lol. DM me if you are interested 🙂
Why can’t you go into detail here? I’m sure we’re all curious about your idea.
I’m just paranoid about someone stealing it lmao
It’s the first actually decent idea I’ve had in a while
Ideas are a dime a dozen
You don’t need to be scared, I won’t judge you I just want to talk about it. People don’t imagine together enough.
i wanna try stuff with the audio
programmatic audio
i send you a friend request
guys, I've been building an extension for ChatGPT that can save a lot of time and It’ll be free to use forever.
But, chrome requires a one time $5 fee to publish extensions on the Web Store so I’m looking for a bit of help to cover that 😅 . If anyone’s willing to support with any amount, I’d really appreciate it.
Supporters will be thanked publicly as early supporters. I can also share a demo if you’d like to see how it works.
If you want to support, I’ve added a Buy Me a Coffee link in my bio
Thanks!
dawg its $5
go look on the ground for a dropped bill
Link the github
I'll do a code review and decide if I want to support your effort after I know what it does.
Lol. Hi mods. If you have questions my DMs are open ❤️
You solicited for side projects with equity, I was trying to figure out if you're good for it and you didn't like it that it takes more than trust me bro
You're not going to build a team if you can't handle people asking if you're capable
No one getting paid if you're in charge.
What was escalated?
Well, I'm sorry you feel that way. I hope you find your team
I hope you find your team
I hope you find the team you deserve
That's what makes us different people I feel
❤️
Lol what?
less annoying than you id say
bro reacted to his own message 💀
you really pinged a random guy 💀
cant even spell 💀
Damion has been dealt with internally. Anyway what's your project?
I don't see why; in the current climate it's impressive that you have written all of those things without having ChatGPT do it for you
Which part did I shit on tho? I was offering to help with presentation, to help with team building. You took me way harsher than I intended.
lmao he blocked me
i cant react to his messages with thumbs down emoji
what a loser
oh so you did have ChatGPT write it for you
if you please: how did you ascertain the novelty of your ideas?
I wouldn't want to get involved in a project only for it to fall prey to IP violations
As we all know, simply asking ChatGPT whether a project is novel is unreliable because it's designed to please the user.
not if it's the same product that somebody has already developed, no
"before youre allowed to interact with me further"
bro thinks hes a king
lmfao
lmao he is expanding my messages even though he blocked me
at least i have something to show for my projects
imagine if i was actually fat
lol
ok: a vague idea you're keeping under wraps that may or may not be new, documents written by chatgpt, and the same tired old AI-inflated ego that goes with it. We see literally dozens and dozens of this exact scenario. Thanks for wasting everyone's time, Vann Allen
dont forget, trying to insult people for no reason after they dont praise you
youre allowed to leave yk
who are you why are you in the observation chamber
you are being monitored and observed
im gonna assign you a number now
also all your messages are being scraped into a massive dataset
welcome to the psychward
everything you say here is being monitored and collected
hello, why do you use anime pfp
because hes a bitch and scared to show his face
you are a bitch too
shut up bitch
i know
but at least im not a loser bitch
just a normal bitch
oh so youre racist now too ok
im east asian
also ray isnt white
fake link
you can have @dusk pollen's onlyfans tho
totally real
thats not mine
mine is https://onlyfans.wtf/theminji
you never sent me your holes
im sad
ill dm u
🙏
i love shit
you got it right
wanna have some shit pron
i have a lot
i don't even know why they have beef with you
im just here to taunt you
simple as that
but i need to see holes before i go to work
my boss is calling me
hurry up
🤔
fine
can you write uwu phrases
what kind of pron do you play
i only act in 1 on 1 pron
how about 2 girls 1 cup
i do not do that
you should
ok
ill play that for our gooning session
so far you have not shown anything
all you did is say "i made something with gpt and i am amazign blah blah also fuck everyone that asked any questions about it"
We don't like people who write laws with AI either.
oh wow that changes everything then!
Like, really, you want to be compared to a congressman?
ok ill treat you as one
senile old man with dementia that cant even talk properly, is too stupid to understand anything modern
at least i have something to show for what ive made
You couldn't handle the 1v1 without crying. I don't see the problem
its better than your non existant shit
i have yet to see a single thing you have done
you have yet to contribute anything meaningful to this server, conversation, or anyone
We see that 🙁
yeah we can tell
you have not said anything positive in your time here
i can tell
hehehe~ owo notices youw bulge getting biggwy fwom just my widdle wordsies~? uwu gonna make it twitchy twitch fow me pwease?? >w<
hehe sowwy~ did I go too far?? 🥺
you are a femboi?
😱
prove it
prove it
prove it
wow
please
send blahaj
hes already on estrogen
hes talking to you
🤔
im a psychopath i don't need estrogen
Your message has been deleted, sorry!
Next time, try saying something like this instead: imagine two Asian guys just spending time together and sharing a kiss...
You have been muted for 5 minutes for the following reason:
sexual score of 0.92
rekt
From the album LAMENTING MALICE
All rights go to Cynthoni.
You can get it here: https://cynthoni.bandcamp.com/album/lamenting-malice
thanks for replying, here's the demo: https://www.loom.com/share/2a5f55a6f76c4537b7b77d9051f45b42
i'll send the gh repo link in dm, i was testing it locally for now
Hey, that's actually pretty cool. I could see the value
thats cool, i dont see myself using it but i know a lot of people would
thanks. this is currently the MVP version. I’m planning to add Previous and Next buttons in the next version to better control the navigation flow, along with some additional accessibility features
yeah, this could be useful for someone who goes back and forth in messages, maybe to edit the conversation for a better response from ChatGPT
How does it handle conversation forks?
i didn't get it?
Like when you go backwards and edit a message, you create a fork in the conversation tree. I use these to step through multi-part instructions and aggregate context from each thread to bring back to trunk. Do you have a way to navigate these forks?
yes, it uses an observer, so whenever there is a change in the DOM, it re render again the nav bar
i have open sourced it,
you can check it here: https://github.com/NandkishorJadoun/chatgpt-message-navigator
"Fuck ICE" bro thinks he's special
There aren't even any real programmers in this server 💀 💀
can I help you? or do you just feel like attacking me for no reason
yeah there actually are lmfao they just don't go here
Most great authors become great posthumously. Wonder how that’ll work for the best ai novelist of our time that never stops writing. Maybe it’ll be the “greatest writer of the 9o era - when things were much simpler and humans were still hanging on”.
higH quality thoughtS
I am working on a different kind of vision ai and needed a proper fitting encoder. So i build one and released it on github with apache 2.0 license.
If anyone is looking for something like that, feel free to use.
Centuries ago, mankind was slaughtered to near extinction by monstrous humanoid creatures called titans, forcing humans to hide in fear behind enormous concentric walls. What makes these giants truly terrifying is that their taste for human flesh is not born out of hunger but what appears to be out of pleasure. To ensure their survival, the remnants of humanity began living within defensive barriers, resulting in one hundred years without a single titan encounter. However, that fragile calm is soon shattered when a colossal titan manages to breach the supposedly impregnable outer wall, reigniting the fight for survival against the man-eating abominations.
After witnessing a horrific personal loss at the hands of the invading creatures, Eren Yeager dedicates his life to their eradication by enlisting into the Survey Corps, an elite military unit that combats the merciless humanoids outside the protection of the walls. Based on Hajime Isayama's award-winning manga, Shingeki no Kyojin follows Eren, along with his adopted sister Mikasa Ackerman and his childhood friend Armin Arlert, as they join the brutal war against the titans and race to discover a way of defeating them before the last walls are breached.
(Source: MAL Rewrite)
84.52
41
2013-04-07
finished
i built a discord bot that will build your server for you based on a prompt with LLM reasoning, it also does welcoming and selectable roles.
inspired by similar bots that are normally behind a pay wall or are full of emojis
What’s the difference between your bot and me asking ChatGPT to tell me what to do
Or using discord’s templates to do the job for me
so I'm thinking of expanding it to claude and gemini… 🤔
Hi, have a nice day! I came across your profile on Discord and found your interests aligning with mine. I was wondering if we could discuss them further and exchange thoughts? I'm full stack developer with 7+ years experience. And now I'm focusing the AI projects.
Here are some links I've worked:
https://sociali.ai/ - AI-Driven Social Media Management Platform
https://booke.ai/ - AI‑Driven Bookkeeping Automation App
https://www.rankorbit.ai/ - AI Calling & Website Support Agent
https://troponiniq.com/ AI powered coach
https://reflexa.ai/ AI Photo Generation Platform
Sociali offers AI-powered social media management tool, making it easy to create, schedule, and optimize posts. Boost engagement and streamline your social media strategy.
AI that actually does bookkeeping work inside QBO/Xero - not just suggestions. Uses your existing bank connection. No Plaid, no extra setup. Try free.
Rank Orbit AI delivers website AI assistance, calling AI automation, and advanced AI optimization to streamline operations and scale smarter get started today.
Transform your health with intelligent nutrition coaching powered by advanced AI. Get personalized plans, track progress, and achieve your goals faster.
Transform your professional image with AI-generated headshots. Perfect for LinkedIn, business profiles, and social media. Create studio-quality headshots instantly - no photographer needed!
Hi all, apologies if this question has been covered, I want a TUI for chatgpt web - something like codex, but for regular chat - does anyone have anything they use on a daily basis?
Go to #💬general
Nobody ever talks here
also what on earth is a TUI
Ollama is literally that (except i dont think you can see past chat histories in the terminal)
also i think it only works for local llms
hey guys, I use gemini pro model and was wondering if anyone here has pre build instructions to make the AI actually fact check and more instead of being friendly and always agreeing with me? just a global one it doesn't have to be a specific topic.
also if there is a library somewhere or a website with pre build specific instructions for example programming instructions its highly appreciated!
the reason I ask instead of making those myself is simply because my english isn't perfect to find the right words😅
thanks in advance!
I use GitHub copilot on VSCode, and it allows you to create an instruction file that copilot can always adhere to every time you send a message
Hi guys rate my website from my profile, any improvement, recommendation etc?
very reassuring model training simulation
Flutter & AI Developer Needed for Long-Term Project
We are seeking an experienced developer with expertise in Flutter, AI, and Python to join our team for a long-term collaboration.
Requirements:
- Proven experience building cross-platform mobile applications for Android and iOS
- Access to both Android and iOS devices for testing and development
- Strong proficiency in Flutter framework and AI integration
Python programming skills - Ability to analyze existing codebases and implement continuous updates
Project Details:
- Current status: MVP in progress
- Scope: Ongoing development with regular updates and iterations
Commitment: Long-term collaboration
Ideal Candidate:
- Highly skilled with a strong attention to detail
- Kind, communicative, and collaborative team player
Capable of taking ownership of the entire codebase
Dedicated to delivering quality work consistently
If you meet these qualifications and are interested in a stable, long-term opportunity, we'd love to hear from you.
Can anyone recommend a graph database or library for representing code (e.g. functions, dependencies, calls) as a graph?
Hi, does anyone usually talk in voice chat to debate about technologies and implementations?
Happy Friday!
We just put DeepSeek-V4-Pro up on https://t.co/es07MrTxSs. It’s the world’s largest open source model at 1.6T parameters, and you can run it for free running on NVIDIA Blackwell GPUs.
Try the NVIDIA NIM API → https://t.co/zeWX4Y7Ipd
anyone interested to full stack collab for project or wanna learn,dev, ai ,coding dm me
noe4j
hello guys is there anyone who are interest in collaboration what can get a lot of profit together?
if you are interest. please dm me and lets discuss about that together. thank you
alexam
im in interest of gooning
please follow my post
hello
I need a programmer
what are you need
Hi
Interested
In this video, I explain why survival simulation is not necessary for creating better AI. I started with a complex environment containing states like hunger, hydration, fatigue, temperature, money, weather, and position tracking, but eventually reduced it down to only a few essential states. The result was a simpler and more focused learning set...
😄