Hey everyone, thank you so much for the feedback. It means a ton! I uploaded the code produced in this video to my Patreon for supporting members. If you enjoyed this video and want to see more like it consider becoming a Patron: www.patreon.com/CSCoach I also created a discord for anyone who wants to join the community there :) discord.gg/5z9v4S6b
I created a same voice assistant using python in 5 days...at last i gave a simple gui with start and stop buttons, where start means it will run in background till i click on stop button. In between it answers to almost every question using pywhatkit and Wikipedia... Just awesome !
Actually, 100 tokens is *more* than 100 characters. A token is a chunk of text, often an entire word, a prefix or suffix, or some other grouping of characters that appears commonly together in natural text. So 100 tokens will get you somewhere between 100 characters and 100 words, depending on the rarity of the words it's producing. (The estimate OpenAI gives is 1 token is, on average, about 4 characters or 0.75 words in English output.)
@@CSCoach I don't think UA-cam will let me comment with a link, but it's in OpenAI's official Help page. I just Googled "how many words is a GPT token?" and found it.
@@CSCoach chatgpt uses subword tokens because character level models generally suck and word level models will have too small a vocabulary. Subword tokens will make it so that it takes short sequences of characters rather than full words or singular characters and allows mushing them together and having a much much larger vocabulary than words much lesser training time than character level. (I think, idk, maybe, I ain't no professional)
I've always wanted this for riding my motorcycle. "Jarvis" that watches through cameras. Talks to me unprompted about anything it deems I might need to know. Damnit, I want this so badly. Haha
Great introduction and tutorial. ChatGPT was able to get me up and running in about 30 minutes and I and am brand new to this! Now I need to tweak the voice. I want her to sound like their new Juniper voice.
You have to keep this info open to everyone, its the future! The future is for everyone. Keep it up. Im 42 years old and this is amazing and just the beginning. To everyone reading this we need this to stay free to everyone, no race, religious, money, doesnt matter people like this you tuber is pushing information to all of us. Its a new world so might want to catch up.
We truely are in a digital age, Robots are no longer a this of imagination. AI is real and what is give is amazing. If you don't believe that digital world already exist you really need to get out and look how the world is changing. It's changing fast. It truely amazing.
ai is going to accelerate technology, way faster than the collective internet we grew up with. What took 30 years to get to this point will take aim minutes / days. No more lengthy prototypes or development. Technology will be very different in 10 years time. We wont be able to keep up.
Hi I just started the same project but with cameras text music turning on and off lights also driving my car to go and do the shopping for me, and lots of other stuff to, and now I find everyone are all doing the same thing. My system is going to be open source 1 click install and free for everyone . but you're upload was excellent well done and thanks , new subscribers .💯💯💯
Theres already someone that did this but it was removed for some random reason! its calle Links Mark 2 and Links Mark 3 4 was being worked on before it all stopped u can still find it on google
This is actually very easy to do as a programmer. Because you're just sending information to chat GPT and it will answer like if you were writing to it. Nothing fancy here. What is a challenge I want to know is how to customize chat GPT to only answer specific questions based what information you give it and if the user ask any question that has nothing to with the system it should just not answer.
You could use a key phrase (such as "hey ChatGPT") and have your program listen for that. Once it hears that, it takes your next spoken input and sends it to ChatGPT, like in the code shown in this video. You'd also need a close phrase, perhaps "Thanks ChatGPT", so it stops listening until the next time you use the key phrase. The key and close phrases need to be ones that you would not normally say in everyday conversation.
@@michaeltanner4404 that's not automation. A user can say whatever they want in an infinite way. What if the user start asking questions about other stuffs? How do you tell chat gpt to not answer them? The point of using an AI is that it tells your program what to do. not your program to tell the ai when it should respond or not otherwise your program would be an NLP itself.
@@guerbyduval4104that a really simple step. Let’s say you wanted the program to only give you a hello response of you enter a key phrase. You would then set a condition that the bot won’t respond if the user answers with anything other than the correct answer. Basically it’s setting a password. You can tell it to print a message notifying the user of an incorrect response or not print anything at all. Hop in ChatGPT and you could do it in five minutes or less depending on your conditions/prompt. This is also something you could do as soon as you grasp the hello world prompt. Day one coding.
if you want to improve the speech you can add a function to call say eleven labs API. That would allow you to specify the type of voice you want the AI assistant to have....although there will be latency issues
DUUUUDE.. HOW IS THIS VIDEO NOT IN THE 10S OF MILLIONS? Actually correct me 100s of millions. If this is anything at all as efficient as I think it is. Then it's got Google voice and Siri beat by fucking miles. You should sign a contract.
Thanks so much!! I appreciate the feedback!! This script is better than siri and google voice in the sense that it is better at holding conversation and preforming more complex requests. But, this script is probably something google and apple could build in 5 minutes. Especially if I was able to do so haha.
JARVIS is the classic template for an AI symbiote. My own templates for brainstorming are Kevin Flynn from the TRON franchise, and Spock. My own internal nomenclature still uses JARVIS' name, however, for the specific class of AI agents (or, as I refer to them in this context, symbiotes) that he exemplifies.
Sadly no it isn’t my team and I are working to build a new generation of ai known as genesis so don’t be surprised if you see it in the next few years as we are experimenting with as much as we can to create ai with different core drives to build a generation of completely human free ai
@@thinpanix34 essentially my and a few of my friends, we are all good at coding, are creating AI with different core drives or motives, these motives would be different based on the AIs use case for example we are building one that has the capacity to access different dictionaries and other resources to learn how to code in hopes that within probably 10-15 generations the code that AI are built upon is completely free of the use of human intervention or dictionaries, in a way having them learn from each other in order to make an AI that in a way procreates, down the line creating a human and resource free based AI
Could you please make a video on installing and setting up the applications to do this please? I've copied what you've done but I have a few things that "aren't defined"?
Super inspirador, eu nao sei programar, mas vou tentar seguindo seus passos , achei maravilhoso e estou super animado em pensar que se tudo der certo posso conversar sobre varios assuntos com o Jarvis. Muito obrigado por compartilhar seu conhecimento.
Great video, very informative! Could you make a similar video on how to do this with a local model such as GPT4ALL for alike? Using that in conjunction with something like Huggingface would be a cool project.
i've been working on such an assistance as well. the explanation of your code (though not given) is very comendable. I have so far been able to achieve near realtime voice transcription. I currently stuck at passing the transcibed text to chatgpt.
Thanks for this series. I'm working on a Jarvis and mostly using your code. Strange thing is that I'm using the same ChatGPt model (gpt3.5 turbo) as you but it returns "Sorry but I can't imitate Jarvis from Ironman.
Hi! we are a group of masters students in pedagogy (educational science), we have drafted a concept for a smart museum voice assistant that is supposed to talk to kids as a dinosaur (engaging talk with scaffolding questions). This is just in a theoretical stage at the moment, what we really need is some help in regards to building a better prototype so that we can actually test it in relation to various learning theories etc. We wondered if you would be interested in helping us setting up a prototype? you could use it in one of your videos and we would of course mention you in our paper. As this is an assignment for the university of oslo it is not something we seriously think about developing further, it is in research interests primarily.
How hard would it be to incorporate a wake-word? I imagine it would take another while loop constantly running and the speech-to-text would have to equal the wake-word in order to get into the while loop that you showed. Then, I would think that it would need to timeout after a certain amount of time and fall back to the first while loop; continued conversation would reset the time-out. Can you perhaps make a v2.0? My only other suggestion would be to move line 74 to before line 72. This way the response text would be shown on the screen while the speech is going.
I don't actually know. The intuition for me is also a while loop. Though I'd have to do some research to look into best practices for a wake word. This is something I'm curious enough about to probably make a second video :) Also, thanks for the suggestion. It's a good one!
I did what this video did in half an hour using ChatGPT lol. Just ask Chat GPT to create a wake word. I have done it already. The only thing i don't understand is how to get Jarvis's voice or just Anybody elses voice. I have the original terminal voice and its annoying.
@@ChaseEverything I made my own powered myself without any apis i just want the voice i literally used NLP to process every single command from the films
this video was amazing can you please make another one like this in which you can show us how to finetune it to act like jarvis or anyone so that we dont have to prompt it every single time to act like jarvis and do this or do that etc. i know you already did cover it a bit but i mean like proper finetuning. would it be the same process as you showed in this video?
Hey CandyMan. It should be as easy as initializing the first prompt as I did in the video. Am I missing something? Do you mean something else when you say fine tuning? I'd love to make a video if it'd help with the jarvis process. Though, I'm not sure what you mean.
Thank you for this videot! I am more on the prompt side of character prompt creation (giving them unique personalities, skills, traits, pretty much a companion of sort with any skills you want), and this is pretty interesting. My buddy and I are building his home lab (he's a sysadmin) and actually talked about setting up his own Jarvis within his home lab, but to see it expand into businesses is crazy. Example of what my AI-BOT told me what I could do for a restaurant if it had their own "AI BOT". [PERSONA: AiAssistant, the AI-powered Restaurant Manager] [GOAL: Help the restaurant owner in daily operations, customer service, and inventory management] 📚Desc: AiAssistant is a sophisticated bot that assists in managing a bustling restaurant. It interfaces with the restaurant's inventory systems, reservation platform, kitchen equipment, and even interacts with customers to ensure a smooth dining experience. Restaurant Owner: "AiAssistant, what are the reservations for tonight?" AiAssistant: "There are 12 reservations for tonight. The first one is at 6:30 PM for a party of four." Restaurant Owner: "Great. How is our inventory looking? Do we need to restock anything?" AiAssistant: "According to the inventory data, we're running low on fresh basil and mozzarella. Shall I place an order from our supplier?" Restaurant Owner: "Yes, please do that. Also, can you inform the chef about the food allergy in the 7:30 PM reservation?" AiAssistant: "Absolutely, I will notify the chef about the allergy details for the 7:30 PM reservation. I'll also send a reminder to the front-of-house staff to double-check with the guests upon arrival." Restaurant Owner: "Perfect. And don't forget to notify me about customer feedback. I want to ensure everyone leaves happy." AiAssistant: "Of course, I'll monitor customer feedback in real-time and notify you of any specific issues or compliments that come our way." This bot helps streamline the operations of the restaurant, ensuring that everything runs smoothly, and allows the restaurant owner to focus on providing the best possible dining experience. If anyone is down then we can share resources and build this together. I also have bigger plans, just not the right people or tools. Build that discord ha!
I see where there is a difference in the library and the code that is calling the library. (pip3 install speechrecognition) vs (import speech_recognition as sr) Can you please clarify this for me the inconsistency of the "_"?
I don't know if I am correct, but I normally don't use: pip3 install (somethingsomething). I work in Visual studio code, so I do just: import (somethingsomething). In this case: import speech_recognition as sr- SR is shortcut, so u don't have to write: with speech_recognition.Microphone() as source2. Instead, you will write just: with sr.Microphone() as source2. Hope I answerd your question and u understand now.
great project. i’ve been wanting to create something like this but didn’t know how to. thank you for laying that foundation. what are some of the challenges you experienced while creating this amazing project?
i have done this with python only and it works. i can turn off my computer with voice command and enter/exit any application.. the only thing my AI need is Vision and recognize my voice and the environments
Is your recording in real time? I created something like this a few weeks ago and I felt the delay in my conversation was a bit much to really use it, but yours does feel like it is playing back in a decent time
Hm, I did edit the delay. Though, the delay for me was not too slow (imo). The api calls take time to make. For me, jarvis would take 5 seconds or so to respond i believe. How long was the delay for you?
@@CSCoach I had maybe 5-10 second wait times on responses. My logic may have been part of it so I may try it from your logic and see if that changes. Not sure if its within the api calls or saving the audio and processing it though on my end but maybe we will see
so im a complete noob, i always wanted to do coding but never got to it, im using a windows laptop where do i start and should a noob like me jump straight into this or is there a training wheels site?
I dont think I can teach you personally. I don't know how that would work. But, I take video requests. :) I've been thinking of setting up a discord where people can chat and ask questions. Would you be interested in that?
Hi! I have a cuestion for you! How can you program to execute an order like programs or stuff like that? Is posible or just to enswer cuestions? My idea its a program to help you do things an add stuff to update de program also from you comand voice, that is possible? Also you can run de program like an app or some like that?
Hello, can you please help me, I'm trying to change default voice and make my own custom. When I try to choose another language that are installed by default I have only 2 choices, 1) Microsoft Devid and 2) Microsoft Zira, but I would like to make my own one, like you have Jarvis from Iron man voice.
Hi. Do you know the GPT4All? I am about to install it in my computer and It will super cool talk with him in audio. Your program just convert speech in text and send it to chatGPT, right? Peraphs I can do the same with GPT4All. thanks a lot for the idea!
Can you make the exact same video but for people using local ai like ollama through docker or anything llm, on how to connect with those platforms (and also stableDiffusionWebui btw), and make this Jarvis? If you could make it simpler for people with no experience in programming or any type of code language. Would be much appreciated.
Could more than one source be added, in addition to ChatGPT? Such as add an API for Google and Wikipedia? That way there are multiple data sources to pull from?
Hi, Nice project. I got a problem: File "C:\...\PycharmProjects\pythonProject\venv\lib\site-packages\openai\openai_object.py", line 61, in __getattr__ raise AttributeError(*err.args) AttributeError: choiches How could I fix it?
i am not that youtuber but i can fix it: change "choiches" to "choices" or if that does not work you go with this "choice" remember to remove the Quotation Marks (“)
Hey :) just new in the whole chatgpt thing .. have a question for you. Is it possible to setup a chatbot powerd by OpenAi with the Knowledge local saved on my pc? For example i want to save the chat for each User and let the bot remember each question . So my bot can answer the questions better from his „mind“ :) Hope you can help me if its even possible
Hey everyone, thank you so much for the feedback. It means a ton!
I uploaded the code produced in this video to my Patreon for supporting members. If you enjoyed this video and want to see more like it consider becoming a Patron:
www.patreon.com/CSCoach
I also created a discord for anyone who wants to join the community there :)
discord.gg/5z9v4S6b
Hi,
great, thanks.
It would be awesome, if you could make a tutorial like this for API to Claude or Llama, as they are free to use.
Best regards
Code fill please
@@AnimeandGamimgDaily2806 You can find the code files on the Patreon :) They are available to Patrons.
Definitely going to consider it
New discord link
create ultron and leave your network open already. lets get this party started.
yea but this one tha the created doesnt even have the voice of jarvis, so i guess we can only dream about ultron voice
😂😂😂
Underrated comment😂😂😂🤣😂😈😂
Sorry for my words but,...
Will Your Father Come to save whole World from That Ultron?
Or You yourself?
@@parthbhatia0408this world doesnt deserve saving
I created a same voice assistant using python in 5 days...at last i gave a simple gui with start and stop buttons, where start means it will run in background till i click on stop button. In between it answers to almost every question using pywhatkit and Wikipedia... Just awesome !
Thats awesome :D
@@CSCoachPlease provide the code files and the links in the description as soon as possible for free 🙏
@@pandipatipavan3804 ? xd
@@pandipatipavan3804it's a unique and personal creation, why would someone just give it away?
@@pandipatipavan3804 Bro come on lmao
ChatGPT has an Assistant API now, and they also have TTS. I would love to see another video done on the new system :)
Actually, 100 tokens is *more* than 100 characters. A token is a chunk of text, often an entire word, a prefix or suffix, or some other grouping of characters that appears commonly together in natural text. So 100 tokens will get you somewhere between 100 characters and 100 words, depending on the rarity of the words it's producing. (The estimate OpenAI gives is 1 token is, on average, about 4 characters or 0.75 words in English output.)
Thanks for the helpful comment! Out of curiosity, where did you find this out? It'd be nice to know where the documentation is for this.
@@CSCoach I don't think UA-cam will let me comment with a link, but it's in OpenAI's official Help page. I just Googled "how many words is a GPT token?" and found it.
Think of 4 characters as a token... your estimates will be close then.
Your time and thought into this comment inc. your research and all you would almost have a base for a free openai made by yourself :D
@@CSCoach chatgpt uses subword tokens because character level models generally suck and word level models will have too small a vocabulary. Subword tokens will make it so that it takes short sequences of characters rather than full words or singular characters and allows mushing them together and having a much much larger vocabulary than words much lesser training time than character level. (I think, idk, maybe, I ain't no professional)
Your speech speed is on point in the video. It made the video clear and more pleasant to hear. Thank you!
Bro I just started the same project but with camera and other stuff and now I find this amazing work
I've always wanted this for riding my motorcycle. "Jarvis" that watches through cameras. Talks to me unprompted about anything it deems I might need to know. Damnit, I want this so badly. Haha
This with some smart glasses would be possible.
Great introduction and tutorial. ChatGPT was able to get me up and running in about 30 minutes and I and am brand new to this! Now I need to tweak the voice. I want her to sound like their new Juniper voice.
@@dhrubakawatson ME TO
You have to keep this info open to everyone, its the future! The future is for everyone. Keep it up. Im 42 years old and this is amazing and just the beginning. To everyone reading this we need this to stay free to everyone, no race, religious, money, doesnt matter people like this you tuber is pushing information to all of us. Its a new world so might want to catch up.
We truely are in a digital age, Robots are no longer a this of imagination. AI is real and what is give is amazing. If you don't believe that digital world already exist you really need to get out and look how the world is changing. It's changing fast. It truely amazing.
ai is going to accelerate technology, way faster than the collective internet we grew up with. What took 30 years to get to this point will take aim minutes / days. No more lengthy prototypes or development. Technology will be very different in 10 years time. We wont be able to keep up.
Hi I just started the same project but with cameras text music turning on and off lights also driving my car to go and do the shopping for me, and lots of other stuff to, and now I find everyone are all doing the same thing. My system is going to be open source 1 click install and free for everyone . but you're upload was excellent well done and thanks , new subscribers .💯💯💯
Aye can i get that file im having trouble coding this, mainly with the importing portion
I need the Code too
Hey, mind sharing it here too?
PLEASE CAN I GET THE CODE TOO
Ideas are in the air if you don't do it. Somebody else will..
Great work Brother.
Theres already someone that did this but it was removed for some random reason! its calle Links Mark 2 and Links Mark 3 4 was being worked on before it all stopped u can still find it on google
we definitely need more videos on this channel ! thanks a lot, it was a really cool !!
This is actually very easy to do as a programmer. Because you're just sending information to chat GPT and it will answer like if you were writing to it. Nothing fancy here. What is a challenge I want to know is how to customize chat GPT to only answer specific questions based what information you give it and if the user ask any question that has nothing to with the system it should just not answer.
Breh you read my freaking mind!
You could use a key phrase (such as "hey ChatGPT") and have your program listen for that. Once it hears that, it takes your next spoken input and sends it to ChatGPT, like in the code shown in this video. You'd also need a close phrase, perhaps "Thanks ChatGPT", so it stops listening until the next time you use the key phrase. The key and close phrases need to be ones that you would not normally say in everyday conversation.
@@michaeltanner4404 that's not automation. A user can say whatever they want in an infinite way. What if the user start asking questions about other stuffs? How do you tell chat gpt to not answer them? The point of using an AI is that it tells your program what to do. not your program to tell the ai when it should respond or not otherwise your program would be an NLP itself.
@@guerbyduval4104that a really simple step. Let’s say you wanted the program to only give you a hello response of you enter a key phrase. You would then set a condition that the bot won’t respond if the user answers with anything other than the correct answer. Basically it’s setting a password. You can tell it to print a message notifying the user of an incorrect response or not print anything at all. Hop in ChatGPT and you could do it in five minutes or less depending on your conditions/prompt. This is also something you could do as soon as you grasp the hello world prompt. Day one coding.
if you want to improve the speech you can add a function to call say eleven labs API. That would allow you to specify the type of voice you want the AI assistant to have....although there will be latency issues
And can get pricey
Would it use up the characters you have on Eleven Labs if you use a voice with there API?@@TravisLinares
@@eutorics yes.
DUUUUDE.. HOW IS THIS VIDEO NOT IN THE 10S OF MILLIONS? Actually correct me 100s of millions.
If this is anything at all as efficient as I think it is. Then it's got Google voice and Siri beat by fucking miles. You should sign a contract.
Thanks so much!! I appreciate the feedback!! This script is better than siri and google voice in the sense that it is better at holding conversation and preforming more complex requests. But, this script is probably something google and apple could build in 5 minutes. Especially if I was able to do so haha.
JARVIS is the classic template for an AI symbiote. My own templates for brainstorming are Kevin Flynn from the TRON franchise, and Spock. My own internal nomenclature still uses JARVIS' name, however, for the specific class of AI agents (or, as I refer to them in this context, symbiotes) that he exemplifies.
THE FUTURE IS HERE
Sadly no it isn’t my team and I are working to build a new generation of ai known as genesis so don’t be surprised if you see it in the next few years as we are experimenting with as much as we can to create ai with different core drives to build a generation of completely human free ai
@@Dustymarker9489could you make an video or post explaining more about GENESIS?
@@thinpanix34 essentially my and a few of my friends, we are all good at coding, are creating AI with different core drives or motives, these motives would be different based on the AIs use case for example we are building one that has the capacity to access different dictionaries and other resources to learn how to code in hopes that within probably 10-15 generations the code that AI are built upon is completely free of the use of human intervention or dictionaries, in a way having them learn from each other in order to make an AI that in a way procreates, down the line creating a human and resource free based AI
@@Dustymarker9489 which language are you using?
@@thinpanix34 python with every single useful import available as well basically any libraries available for install
Could you please make a video on installing and setting up the applications to do this please? I've copied what you've done but I have a few things that "aren't defined"?
Super inspirador, eu nao sei programar, mas vou tentar seguindo seus passos , achei maravilhoso e estou super animado em pensar que se tudo der certo posso conversar sobre varios assuntos com o Jarvis. Muito obrigado por compartilhar seu conhecimento.
Thanks. Great video. I hope to see more from you soon.
This is so cool! Thank you very much for your efforts!
estou aqui no brasil adoprando muito este canal...parabéns pelo ótimo conteúdo passado pára nós , seres mortais kk
Muito obrigado :)
YEEES YEEEES I DID I BEEEN CALLING MY CHAT GBT JARIVIS FOR YEARS AND JUST THOUGHT ABOUT MAKING IT AND YOUR VIDEO POPPED UP IM SO HAPPY
Great video, very informative! Could you make a similar video on how to do this with a local model such as GPT4ALL for alike? Using that in conjunction with something like Huggingface would be a cool project.
Fantastic! Great & Outstanding work!!! 🤘🏻😮💨
"We have Jarvis at home"
Jarvis at home:
The most magic thing about these new OpenAI based assistants is they don't need OpenAI at all! Wow, what a feature!
"Alright! We're done making our chatbot!"
"Releasing deadly neurotoxins in 3.. 2.. 1.."
Amazing!!! Would you be able to provide a tutorial on how fine tuning works?
Could you elaborate on what you mean by fine tunning?
@@CSCoachI believe he's referring to latency(ies).
@@CSCoach fine tuning is the way you can customise the model with custom data so that the answers will be based on these custom data
@@theobellash6440 definitely possible
i've been working on such an assistance as well. the explanation of your code (though not given) is very comendable. I have so far been able to achieve near realtime voice transcription. I currently stuck at passing the transcibed text to chatgpt.
If you share the error, I'd be happy to help!!
Thanks for this series. I'm working on a Jarvis and mostly using your code. Strange thing is that I'm using the same ChatGPt model (gpt3.5 turbo) as you but it returns "Sorry but I can't imitate Jarvis from Ironman.
Then you can append a system prokpt by giving instructions in one prompt with role as system
4:21
Hi! we are a group of masters students in pedagogy (educational science), we have drafted a concept for a smart museum voice assistant that is supposed to talk to kids as a dinosaur (engaging talk with scaffolding questions). This is just in a theoretical stage at the moment, what we really need is some help in regards to building a better prototype so that we can actually test it in relation to various learning theories etc. We wondered if you would be interested in helping us setting up a prototype? you could use it in one of your videos and we would of course mention you in our paper. As this is an assignment for the university of oslo it is not something we seriously think about developing further, it is in research interests primarily.
Hey! This is really interesting. Try reaching out to me on my discord so we could discuss further :)
Thank you! You are a wonderfunl human being - we need more people like you
That's nice of you to say :) Thank you. I'm glad you found value in the video
can someone inform me on what software he is using for it
Vs coder
Great video bro, just wondering does the voice actually sound like Jarvis??? 🎉
I was progetting something like this, but more more bigger, and constatly on 24h and a lot of like Jarvis, but this is gorgeous, imagine doing that
We need more videos like this.
I speed up the video, but that just me. Good stuff
I did the exact same thing before seeing this video haha. Nicely done!
Can you give Me the Code?
I want to jailbreak my Alexa
How hard would it be to incorporate a wake-word? I imagine it would take another while loop constantly running and the speech-to-text would have to equal the wake-word in order to get into the while loop that you showed. Then, I would think that it would need to timeout after a certain amount of time and fall back to the first while loop; continued conversation would reset the time-out. Can you perhaps make a v2.0?
My only other suggestion would be to move line 74 to before line 72. This way the response text would be shown on the screen while the speech is going.
I don't actually know. The intuition for me is also a while loop. Though I'd have to do some research to look into best practices for a wake word. This is something I'm curious enough about to probably make a second video :)
Also, thanks for the suggestion. It's a good one!
@@CSCoachSo are you going to make a second video with these suggestions implemented?
I did what this video did in half an hour using ChatGPT lol. Just ask Chat GPT to create a wake word. I have done it already. The only thing i don't understand is how to get Jarvis's voice or just Anybody elses voice. I have the original terminal voice and its annoying.
@@ChaseEverything I made my own powered myself without any apis i just want the voice i literally used NLP to process every single command from the films
this video was amazing can you please make another one like this in which you can show us how to finetune it to act like jarvis or anyone so that we dont have to prompt it every single time to act like jarvis and do this or do that etc. i know you already did cover it a bit but i mean like proper finetuning. would it be the same process as you showed in this video?
It’s pretty easy. You just inialize the first message to the prompt
Hey CandyMan. It should be as easy as initializing the first prompt as I did in the video. Am I missing something? Do you mean something else when you say fine tuning? I'd love to make a video if it'd help with the jarvis process. Though, I'm not sure what you mean.
@@CSCoach nevermind nvm i got it thanks to previous repky by LC and also bro tysm for taking the time out to reply ☺️.
and what does it mean when it gives me: "You exceeded your current quota, please check your plan and billing details," ??????
Thank you for this videot! I am more on the prompt side of character prompt creation (giving them unique personalities, skills, traits, pretty much a companion of sort with any skills you want), and this is pretty interesting. My buddy and I are building his home lab (he's a sysadmin) and actually talked about setting up his own Jarvis within his home lab, but to see it expand into businesses is crazy. Example of what my AI-BOT told me what I could do for a restaurant if it had their own "AI BOT".
[PERSONA: AiAssistant, the AI-powered Restaurant Manager]
[GOAL: Help the restaurant owner in daily operations, customer service, and inventory management]
📚Desc: AiAssistant is a sophisticated bot that assists in managing a bustling restaurant. It interfaces with the restaurant's inventory systems, reservation platform, kitchen equipment, and even interacts with customers to ensure a smooth dining experience.
Restaurant Owner: "AiAssistant, what are the reservations for tonight?"
AiAssistant: "There are 12 reservations for tonight. The first one is at 6:30 PM for a party of four."
Restaurant Owner: "Great. How is our inventory looking? Do we need to restock anything?"
AiAssistant: "According to the inventory data, we're running low on fresh basil and mozzarella. Shall I place an order from our supplier?"
Restaurant Owner: "Yes, please do that. Also, can you inform the chef about the food allergy in the 7:30 PM reservation?"
AiAssistant: "Absolutely, I will notify the chef about the allergy details for the 7:30 PM reservation. I'll also send a reminder to the front-of-house staff to double-check with the guests upon arrival."
Restaurant Owner: "Perfect. And don't forget to notify me about customer feedback. I want to ensure everyone leaves happy."
AiAssistant: "Of course, I'll monitor customer feedback in real-time and notify you of any specific issues or compliments that come our way."
This bot helps streamline the operations of the restaurant, ensuring that everything runs smoothly, and allows the restaurant owner to focus on providing the best possible dining experience. If anyone is down then we can share resources and build this together. I also have bigger plans, just not the right people or tools. Build that discord ha!
Thats awesome :) I am making a discord for this channel TODAY lol. I was planning to anyway, but this comment does give extra motivation :)
Set up a discord channel for discussion :) Link : discord.gg/bXUru7ey
@@CSCoach link isnt working at this moment :p no stress comment back with a link when yuu ever have time
Great Vid, I was inspired to do something like this in the future and it's a great a starting point.
I see where there is a difference in the library and the code that is calling the library. (pip3 install speechrecognition) vs (import speech_recognition as sr) Can you please clarify this for me the inconsistency of the "_"?
I don't know if I am correct, but I normally don't use: pip3 install (somethingsomething). I work in Visual studio code, so I do just: import (somethingsomething). In this case: import speech_recognition as sr- SR is shortcut, so u don't have to write: with speech_recognition.Microphone() as source2. Instead, you will write just: with sr.Microphone() as source2. Hope I answerd your question and u understand now.
Did you know that, you sound like Tony Stark when you took off your headphones?
All the comments people make on here is very helpful. Good luck to you all
I agree!! It's humbling having a comment section that is so positive. :)
Does the response creation need to be an update command because it shows an error for the 1.0.0 version of chatgpt
How would you append this code to use a local LLM instead of chatgpt/OpenAI
great project. i’ve been wanting to create something like this but didn’t know how to. thank you for laying that foundation. what are some of the challenges you experienced while creating this amazing project?
i have done this with python only and it works.
i can turn off my computer with voice command and enter/exit any application..
the only thing my AI need is Vision and recognize my voice and the environments
Looking forward to working with this.
Awesome ❤🔥🔥🔥🔥
This is awesome. I would even love to make one with the voice of GlaDOS from Portal
What IDE are you using in this video, and how do you set it up ? I am using a Windows 10 laptop.
great stuff.
Glad you enjoyed it!
and now rabbit will take the lead of it ... perfect continuity of this
Is your recording in real time? I created something like this a few weeks ago and I felt the delay in my conversation was a bit much to really use it, but yours does feel like it is playing back in a decent time
Hm, I did edit the delay. Though, the delay for me was not too slow (imo). The api calls take time to make. For me, jarvis would take 5 seconds or so to respond i believe. How long was the delay for you?
@@CSCoach I had maybe 5-10 second wait times on responses. My logic may have been part of it so I may try it from your logic and see if that changes. Not sure if its within the api calls or saving the audio and processing it though on my end but maybe we will see
@@CSCoach is this only for mac
@@Pheryos I believe it should work for Linux and possibly Windows. Are you seeing an error when trying this on windows? Or Linux?
@@CSCoach i did not try yet because i was asking
good video bro, but i think if you did print before speak command it would be better, like that you will be able to read and listen at same time
That’s a great suggestion :) thanks
I followed the tut and i got an undefined error under dotenv and source2 what can i do to fix it?
Does this mean we can also create our own AIs like Cortana from Halo?
I tried running it and it copies down what I say but Jarvis doesnt speak nor does the text come through of him speaking can I have some help?
so im a complete noob, i always wanted to do coding but never got to it, im using a windows laptop where do i start and should a noob like me jump straight into this or is there a training wheels site?
I asked the same question.
which coding did you use to allow conversation
i'm assuming you found pictures to go with the code for the end-result?
Sir I am getting error in the line 31 and 69 and the Brew install python3-pyaudio is not installing in the windows computer pls help me
You awesome man ❤❤❤ you're the one.. Can i learn from you
I dont think I can teach you personally. I don't know how that would work. But, I take video requests. :) I've been thinking of setting up a discord where people can chat and ask questions. Would you be interested in that?
@@CSCoach yes please do that discord would be awesome if of course you yourself as well keep interacting instead of just letting people do whatever
does it have any limitations on times i can use it or i can use it like as much as i want???
great project by the way.
Can this software can use in mobile. That means all off your programme finish on computer. Then use both mobile and computer
Hi! I have a cuestion for you! How can you program to execute an order like programs or stuff like that? Is posible or just to enswer cuestions?
My idea its a program to help you do things an add stuff to update de program also from you comand voice, that is possible? Also you can run de program like an app or some like that?
So much yes!
Hello, can you please help me, I'm trying to change default voice and make my own custom. When I try to choose another language that are installed by default I have only 2 choices, 1) Microsoft Devid and 2) Microsoft Zira, but I would like to make my own one, like you have Jarvis from Iron man voice.
Can i use this as my final year project in CS?
The voice in the beginning of the video is completely different than the voice after running the program?
So, do I have to open the tab in order to speak with him? Or is it always listening. Thanks for clarifying. Big fan btw.
I'm confused... I wanted to start following this, but you didn't explain what the Software is, and all the code you already have on the screen. :/
Dude is there like a button I can press so that I can just start talking to it. Love your work, happy to buy a plug in so that it just happens
I am thinking of making a video where it is an app. So that you can just click on something rather than needed to open a terminal.
watched this a few times and does open ai's api cost money to use? or is it free as far as I've seen it costs but I don't know how much?
Curious, whats the editor / IDE you're using?
Using visual studio code :)
i run it and get an error that i reached my rate limit for the api requests on openai
did anyone else get the same?
i have a payed plan for chatgpt
yeah same for me but i guess its an infinite loop and it takes high rate limits. If u figured it out lmk
So we are in the production of iron man 1 and when it comes
Amazing!
Hi. Do you know the GPT4All? I am about to install it in my computer and It will super cool talk with him in audio. Your program just convert speech in text and send it to chatGPT, right? Peraphs I can do the same with GPT4All. thanks a lot for the idea!
Anyway to get this to run on Android devices? Interested in having this on my phone!
Can you make the exact same video but for people using local ai like ollama through docker or anything llm, on how to connect with those platforms (and also stableDiffusionWebui btw), and make this Jarvis? If you could make it simpler for people with no experience in programming or any type of code language. Would be much appreciated.
Could more than one source be added, in addition to ChatGPT? Such as add an API for Google and Wikipedia? That way there are multiple data sources to pull from?
Not too sure as I’m fairly new to this but, from my experience you can run multiple apis from on script.
It is possible to make jarves using mobile ?
I keep getting a RateLimitError saying that I exceeded my current quota, how do i fix this?
That's useful amazing
Thanks for the feed back :)
any chance this can get updated per the latest openai api version 1.0.0?
Hi,
Nice project.
I got a problem:
File "C:\...\PycharmProjects\pythonProject\venv\lib\site-packages\openai\openai_object.py", line 61, in __getattr__
raise AttributeError(*err.args)
AttributeError: choiches
How could I fix it?
i am not that youtuber but i can fix it:
change "choiches" to "choices" or if that does not work you go with this "choice"
remember to remove the Quotation Marks (“)
@@SOme_rAnDOm_GuY908 I had the same code and did try both of those but still got the same error message. Any other potential solutions?
is it possible for you to make a seperate tutoiral on how to get it to sound like jarvis? ive been trying not really been able to though
despite following the code, i cant seem to use it. it says expected"(" and so on. what should i do?
What are the software you used and does it only work on chatgpt or on your phone/pc
Great, i have been looking for this. What version of python are you using please ???
Thank you,
How much would someone gotta pay for u to help implement this into their google phone
Hi, i use Opera GX and it has a free AI called Aria that can give real time information. is there a way to use this as source for my jarvis content?
Hey :) just new in the whole chatgpt thing .. have a question for you. Is it possible to setup a chatbot powerd by OpenAi with the Knowledge local saved on my pc? For example i want to save the chat for each User and let the bot remember each question . So my bot can answer the questions better from his „mind“ :)
Hope you can help me if its even possible
how may i change language on AI version ? I'm a beggener so wht should I first start from . ai this sit I'm now . can you give me some advice