RIP HUMAN CODERS! CREATE ANY APP With AI AGENTS in 1 PROMPT!
HTML-код
- Опубликовано: 6 сен 2024
- Say goodbye to manual coding and hello to automation with DEVIKA, the groundbreaking open-source AI agent that transforms your instructions into real apps right on your local machine! In this video, I'll walk you through how to install DEVIKA on your computer and I will then show you how DEVIKA can build an entire app in mere seconds. Imagine the future: No more manual coding, just your ideas coming to life effortlessly! This is what the future of local AGI agents will look like!
What do you think of Devika? Let me know in the comments!
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
SOCIAL MEDIA LINKS!
✨ Support my work on Patreon: / aitrepreneur
⚔️ Join the Discord server: / discord
🧠 My Second Channel THE MAKER LAIR: bit.ly/themake...
📧 Business Contact: theaitrepreneur@gmail.com
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
✨ PATREON LINK: / aitrepreneur
Devika: github.com/sti...
Devika Ollama fix: github.com/hja...
Ollama: ollama.com
Search Engine : developers.goo...
Search Engine Id: programmablese...
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
►► My PC & Favorite Gear:
i9-12900K: amzn.to/3L03tLG
RTX 3090 Gigabyte Vision OC : amzn.to/40ANaue
SAMSUNG 980 PRO SSD 2TB PCIe NVMe: amzn.to/3oBR0WO
Kingston FURY Beast 64GB 3200MHz DDR4 : amzn.to/3osdZ6z
iCUE 4000X - White: amzn.to/40y9BAk
ASRock Z690 DDR4 : amzn.to/3Amcxph
Corsair RM850 - White : amzn.to/3NbXlm2
Corsair iCUE SP120 : amzn.to/43WR9nW
Noctua NH-D15 chromax.Black : amzn.to/3H7qQSa
EDUP PCIe WiFi 6E Card Bluetooth : amzn.to/40t5Lsk
Recording Gear:
Rode PodMic : amzn.to/43ZvYlm
Rode AI-1 USB Audio Interface : amzn.to/3N6ybFk
Rode WS2 Microphone Pop Filter : amzn.to/3oIo9Qw
Elgato Wave Mic Arm : amzn.to/3LosH7D
Stagg XLR Cable - Black - 6M : amzn.to/3L5Fuue
FetHead Microphone Preamp : amzn.to/41TWQ4o
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
Special thanks to Royal Emperor:
- Totoro
- TNSEE
- RG
- Gluthoric
- Peter Bernaiche
Thank you so much for your support on Patreon! You are truly a glory to behold! Your generosity is immense, and it means the world to me. Thank you for helping me keep the lights on and the content flowing. Thank you very much!
#devika #aiagents #textgeneration
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
WATCH MY MOST POPULAR VIDEOS:
RECOMMENDED WATCHING - All LLM & ChatGPT Video:
►► • CHATGPT
RECOMMENDED WATCHING - My "Tutorial" Playlist:
►► bit.ly/TuTPlay...
Disclosure: Bear in mind that some of the links in this post are affiliate links and if you go through them to make a purchase I will earn a commission. Keep in mind that I link these companies and their products because of their quality and not because of the commission I receive from your purchases. The decision is yours, and whether or not you decide to buy something is completely up to you.
HELLO HUMANS! Thank you for watching & do NOT forget to LIKE and SUBSCRIBE For More Ai Updates. Thx
Is your discord link down? I have a question about devika
"Make me a better version of you.... GO!!!"
We are very close to scary times.
"print little pieces on 3d printer that act as mini-ai's, ask them to create new ai's, and the new ai's to create new ai's"
That argument is a little outdated since we know how reliant AI is on data.
@@Ausstein Its only a matter of time. Exponencial development incoming.
I’m getting kinda tired of Snake Game Demos. LLMs can build these direct from the chat bots. Even a custom GPT was able to achieve this.
Same pong tetris tank pacman, I'd like to see complex snes style games like a remake of wwf royal rumble
Snes types of games require another type of A.I to design and animate characters.
@infj5196 nahh man, animations work well in python. Generating consistent sprites is mainly the issue. The more time you put in you will achieve 💯
npm install has an error , that it can't find the " package.json " file,
I found it in the devika folder under the name " package-lock.json " , renamed it and the installation completed successfully .
I was just coming to the comments to find this issue..... I asked Claude about renaming it and it told me I would have a broken pakage.json, as they are completely different contents. It seems in this case, Claude was wrong and the github repo has been published to with files in a poor state.
I kinda miss the old reviews of models, however great video as alway.
I am honestly salivating for both an automatic model picker, where instead of being limited to a single AI model.
We can have loads of them each hyper specialized in certain tasks that can be loaded and unloaded as needed.
Since despite the fact that even if the RTX 5090 can't handle 70 Billion Models, The RTX 6090 most definitely will.
Most people won't want to spend that kind of cash on a GPU so smaller more specialized AIs will allow more people to enjoy the benefits of this technology.
And Addons for programs like Unreal or Godot where we ask the AI to do something and the AI searchs the Net on how to do it if need be and then does it.
Similar to what BuildBox 4 is prototyping.
AutoGen 2 can sort of do that. It allows you to define different agents that might have more specific tasks and can use separate models per agent.
Takes more planning because you're responsible for defining the purpose and tools the agents will use. From there, each agent can spin up and down while tasks are delegated across the team you define
RTX 6090 won't be able to do it either. It's all about VRAM and Nvidia hates giving lot of VRAM for gaming GPU's
Not to mention the speed. If we had many light specialized models, they would run like beasts on better GPUs
@@touma-san91 You're assuming that the RTX 6090 will also have 24 Gigs of VRAM And that we won't get any more efficient at running models in the next 2 to 3 years.
I'm glad you listened to your audience in the last poll
Anyone who works in the industry will tell you that this is still far from usable for real projects.
Why’s that? Creativity?
@@stresseddude Requirement definition!
When making a real project, there will be a lot back and forth communication between the devs and the client. It will never be one prompt
@@tengun I don't think that's how many people who will use AI in coding and are will rely on just one prompt. The person who posted the video and their title is a bit hyperbolic.
WHOA!!! Llama 3 is available for personal use? Can you create a video showing us the latest/best way to use LLMs and where to get Llama 3?
Could you have not made better examples? I can an in one line get llama 3 8b to make a better snake game with score, game over screen. And it would not have to search the internet or install software. GPT 3 could make a Conway game of life game in one prompt too.
Its because these people are non-coders. They all use the same example and that example only shows that these LLM's already have this code in it. Nothing Special.
Except for needing to rename a file during the NPM installation process, these instructions were successful for me. Thank you, sir.
same for me, idk how he missed this
What file did you rename?
@@zacharymaupin3626 package-lock.json to package.json i guess
@@zacharymaupin3626 package-lock.json to package.json
Excited to see where this project goes. Thanks for the upload and encourage others to join his Patrion, totally worth it.
AI being a midwit trap leads me to believe this has a ton of limitations/issues not showcased...
"b-but it'll get better, the project just started.."
Sam Altman has been claiming OpenAI has or is "fast approaching" AGI for over 2 years now. And GPT only hallucinates more and more, while now struggling with basic tasks; wake up.
Who pissed in your coffee this morning?
How do i know you're poor without telling me you're poor.
you not being able to use AI doesn't make AI bad, also you literally take someone elses words that is beholdent to shareholders and use them for yourself as if they were your own. talk about irony and midwitts LOL.
The lobotomised version the public gets vs OpenAIs unreleased uncensored tech? Cmon man. Rumors are they already have AGI. Also Altmans a charming liar.
@@piplupsuper0 How does being poor have to do with AI not being as smart as Dumb Main Stream Media say it is? Its not that smart... if you understand how it works. 99.99% of the world has no clue how to works and thinks its magic. lol
Amazing, I will install it tonight
Holy crap. This is nuts!
Devika is alrey capaby of doing more than devin and it takes very little time. 😂
Good work, AItrepreneur👏
bun install not work
same here... npm didn't work neither
Apparently it the Devika folder needs to be on C: for the bun install to work
@@andremanzato4718 npm install requires you to rename a package-lock.json in Devika folder to package.json, it's looking for that file.
Finally, something that will save me the effort of cut/paste from Meta AI into VS Code. Seriously though, am I missing something here?
What the value over LM Studio + VS Code + Meta AI (or the model I load into LM Studio)?
Nice but what about C# support?
I would love to test some ideas with C# and even Unity.
That would be great and a real game changer.
That would be amazing.
thx buddy!
🙌🏻
Very interesting video!
This AI coding only works for project like ping-pong, flappy bird, snake game, calculators, etc...
Real projects can't be done with AI now.
This is crazy❤
it's not full app. Rest api with login and database. Then you have to launch the app in VM. its still a lot work. setup the vm part all the commandline. more like toy program
Please tell me that AGI isn't being misused now because none of these LLMs qualify as AGI
I have a question, could you make a video on how to uninstall what you show us to install ?
i followed the tutorial you did a while back on the webui for chat bot, but i would like to unsintall properly, not just delete the folder.
could you do a tutorial for that please ? that would be great !
what happened to the video where you train with 1 picture?
🔥👍Thnx!
Very cool concept and video, but could you put the things you type into the description? some of its hard to read in youtube.
😵💫🙌🏻
Personally I find CrewAI to be much better than Devika. It may not have the fancy ui but it doesn't matter since with it you can build one yourself.
im getting SOCKET info MESSAGE: {'type': 'warning', 'message': 'Invalid response from the model, trying again...'}... there is no other branch on devika project anymore
Hello, could you please advise on how to install requirements for each project separately? I really don't want to affect my system each time I run tests. Thank you!
ok so... if you have a newer python version then the 10.6 that he shows in the video, you can get a lot of troble installing this tool, because of Torch and Xformers and etc. I've been trying to fix several problems the whole day, and i just figured it out. I'm installing my phyton with this older version to see if my theorys are correct.
Update: yep, python above the 10 version is indeed a problem
Can it write python code to do calculus questions?
someone know if they merge the fix for local LLM in the main project?
Eh its not bad for a small app. but it doesnt perform well if you need to expand it out further than the basics
I wonder if I can train this tech to make some features in Godot.
anyone knows if it can be adapted to work with Unity (based in C#)?
maybe for your next video you can make about open-devin to it setup locally, please
You should contribute your installation files to Devika afyer a few days.
im total analfabet in coding and it was pretty easy straightforward for me to follow github install notes....
Is python a requirement? Whenever I install it, I just get massive instability, blue screens etc. Which all vanishes with 0 problems after I Uninstaller python. Then the same problems if I install it.
the bot runs into a whoo there partner error cuase redit blocked it from searching
wonder if it can program in unreal engine!
You have showed your Google Engine ID :)
Hello,👋🏼 I was wondering if anyone can help with a concern and worry I’m dealing with? My worry is mobile phones and mobile apps won’t be around or available for my lifetime I’m 31 years old? I’m just concerned that’s all. Just need some reassurance that they’ll be available or around for my generation or lifetime at least. 👍🏼 Appreciate the input and help from everyone!! 🙏🏼
I prefer the term AIG or GAI. AGI is a confusing cluster mess.
To clarify:
AIG = AI Generation (or AI-Generated when referring to content).
GAI = Generative AI.
These terms are more universally adaptable than AGI or "Artificial Generative Intelligence". First time reading that abbreviation (which was just a few months ago in 2024), I could not understand at all what it meant. When I was 1st introduced to AI generation, the first term known to me was GAI. When it blew up in popularity this decade, the next known term was AIG, commonly used to describe AIG art and AIG prompts.
I believe AGI stands for Artificial General Intelligence. It's supposed to be the next step for generative AI.
@@draill No it is not "the next step for generative AI". It is a broad term that covers everything AI technology related and LONG existed before GAI became a thing. It is too broad a term to be used to represent GAI. Even Wikipedia (and the sources it derived from) will confirm this.
Using the term GAI is already sufficient, and AIG as an adverb when describing content made using GAI technology.
@@OnyeNacho Well Wikipedia says:
Artificial general intelligence (AGI) is a type of artificial intelligence (AI) that matches or surpasses human capabilities across a wide range of cognitive tasks.[1] This is in contrast to narrow AI, which is designed for specific tasks.[2] AGI is considered one of various definitions of strong AI.
Creating AGI is a primary goal of AI research and of companies such as OpenAI[3] and Meta.[4] A 2020 survey identified 72 active AGI R&D projects spread across 37 countries.[5]
The timeline for achieving AGI remains a subject of ongoing debate among researchers and experts. As of 2023, some argue that it may be possible in years or decades; others maintain it might take a century or longer; and a minority believe it may never be achieved.[6] There is debate on the exact definition of AGI, and regarding whether modern large language models (LLMs) such as GPT-4 are early, incomplete forms of AGI.[7] AGI is a common topic in science fiction and futures studies.
Contention exists over the potential for AGI to pose a threat to humanity;[8] for example, OpenAI claims to treat it as an existential risk, while others find the development of AGI to be too remote to present a risk.
I believe I was right... In my understanding, there is machine learning > deep learning > generative AI, and then AGI as a super human form of artificial intelligence.
@@draill Alright, I think I see what you mean now. But still, it is very broad a topic to use just to describe Generative AI. The Wikipedia page even states "Not to be confused with Generative artificial intelligence." in their disclaimer before the rest of the page.
That was my problem with this video. The uploader used the wrong term to describe just one particular technology. I didn't come to this video to learn about machine learning or deep learning, I wanted to learn about Generative AI and how to create content with it.
@@OnyeNacho Yeah me too, and I must say, Devika is very impressive ! Did you manage to create what you wanted with it?
how much storage space does this program take up and can I use a pre-existing python enviroment or connect it to a local ai ui like oogabooga or faraday?
ERROR: Could not find a version that satisfies the requirement xformers (from versions: none)
ERROR: No matching distribution found for xformers
you need to install a previous version off python (10.11 or lower)
RIP human. Dont worry we can lying in the bad and wait diabetes coming
gyus, can not run locally, get CORS error. do u know how to fix? using windows
what are the specs of the local machine to run all locally?, can ollama work on cellphones?
Make a video about using llama 3 for autodev devika.
no api keys if you run the llm locally
Manual instructions fail with the 'NPM install' part. Why is this the case?
(env) C:\Users\myusernameishere\devika>npm install
npm error code ENOENT
npm error syscall open
npm error path C:\Users\myusernameishere\devika\package.json
npm error errno -4058
npm error enoent Could not read package.json: Error: ENOENT: no such file or directory, open 'C:\Users\myusernameishere\devika\package.json'
npm error enoent This is related to npm not being able to find a file.
npm error enoent
Hi. Could you recommend a good video enhancer? Thanks
I've tried with both Edge and Chrome, and it will not let me add the API keys in the settings. I can't type in the boxes there, nor can I delete the text already in them.
did you click the edit button at the bottom left, below netlify?
You have to click the "Edit button" at the bottom, then after you change the inputs, you need to click save to commit your changes.
What would be the best model for C# programming? Likely godot?
Godot is a games engine, not a model. Or did you mean "What is the best model for programming c# in Godot?"? In that case it might be Starcoder. Unfortunately i havent really found many models specializing in C#
the fix is still not merge as of today - 2.5.2024
Skynet comments inbound.
5 hours in...still at install process...
DEVIKA vs Crew AI?
I installed it right away hoping it could replace me, so I could scale away my coding work. I told it to build a simple bookstore library SaaS platform using .NET core framework with EF code-first entities. I wanted the architecture to be in CQRS pattern with event sourcing in place. It needed to be in Clean Architecture code using Separated of Concerns as highest regard. It was sure to say it couldn't even create the Database entities correctly.
While coding AI agents will one day going to be reality. We are still very very far from replacing human coders and saying RIP just because AI can code up a classic game from 1980. I do expect that with larger context-windows we might be to do more advanced stuff. But coding is quite easy but programming and building a full-stack system will always be a little bit difficult I assume.
if the trainingdata are there, even that will by todays models be no problem
Good video
This is still unacceptable..
Let's wait for better versions.
For now, I'll focus on mastering proper programming, and such tools will be beneficial in the future
You realize that "better versions" don't just fall out of the sky right? This is like saying I'm not going to practice the piano now because I'm bad, I'm going to wait until I'm better.
@@generichuman_ No dude.. piano missing half buttons.
What computer specs do you need to run it with Llama3
24 GB DDR5 RAM and at least 16 GB of VRAM (and of course a fast internet connection 1 Gbps+ up and down)
omg finally wow
I’ve thought about it a lot and it’s actually good for human coders. Here comes a tsunami of shitty code someone has to maintain or propose a rewrite😅
It’s Indian Devs phase all over again.
cool cool cool
Anybody has tried it? Is it really any good?
I can see you in the future "Hello humans, are you tired of trying to conquer the world, but is too boring do all the to bureaucracy? have you wish an AI could do it for you? NOW IT CAN!!
how can i reach you i need to talk you
Welcome AI overlords, welcome ! 😅
SWE-Agent seems to be better. Tutorial on him next? 👀
neat
Thanks!
Nah.. not yet...
Did You upload one click ... yet :)?
Salut, j'aime ton accent, tu as une chaine en français ?
Uh, I built this a month ago.
Uh, who the hell cares?
is she indian
I really would like to get a kick in the ass by the AI and end to flipping burgers at McDonal for the rest of my days but unfortunately I must say that right now those tools are quite useless in real dev context...hopefully in few years my dream will come true
Devika is useless. Fails to complete the easiest of tasks.
🙌🏻
😵💫🙌🏻