@@alecd8534 I went live on the 14th of march and got 15k impressions and 150 cliks since then. Only product pages with content on them. Im currently mapping out a topical map, imagine the traffic when i posted al the content mapped out (about 300 blog posts)
Thanks for the video. My first attempt using GPT3 got me no real data in the outline. I'll keep trying. 2nd run was much better. Not fantastic yet but, a usable outline. Cost about $1.00. I highly recommend people use a Docker container, separate virtual machine or a virtualenv.
@@trevor4106 Primarily security since several libraries need to be loaded. It'll run fine natively, I just like to keep apps isolated so it's easy to update and run the requirements.txt without dirtying up my machines. Cheers.
This video opened my mind. Currently, I am using GPT-4 to create my posts and it has produced great results! My audience is different from yours, my posts need to be very detailed and contain extremely accurate information. So much so that I always ask the GPT to use reference texts that I provide in the prompt. In your opinion, would it be possible to create something like this with autoGPT? Giving a blog title, post structure, and reference links? It seems to me like a way to generate a huge amount of high-quality posts.
It is possible i am having a similar use case. You basically work with template files, which is a perfect blog post and feed auto gpt the file as a input. Telling it to use it as a template for a new keyword.
any help? I keep getting this error when trying to create a file: Command write_to_file returned: Error: Attempted to access outside of working directory
Yes. You get $18 of free API usage. After that, you have to pay for it by adding a card on file. Here's a link to how's its done th-cam.com/video/4CMtzlA-LzA/w-d-xo.html
Wow. This tech is advancing so quickly. Once they put a good UI on top of that it will be so much easier to use. What use cases are you guys thinking you can do with this for content creation, Seo, creating spreadsheet for analysis etc?
you were having issues with running commands as you were one directroy above the root, you have 2 folders called Auto-gpt-master, and are in the first as can be seen by the path in the terminal prompt, should have \Downloads\Auto-Gpt-master\Auto-Gpt-master , if you do cd Auto-Gpt-master you should be in the right directory to run all the commands from terminal.
Thanks - I discovered early on that dragging and dropping like that seemed to always solve my problem - I have my own weird ways of doing this stuff because I don’t really know much about programming
Hi, i have a pretty simple question. It’s really about simple SEO. Sometimes, I notice that different order in words can have a dramaric impact in the difficulty. Take this exemple: what cc in motorcycle engine mean? Is a difficulty of 5 and what mean cc in motorcycle engine is at difficulty around 50? Both have around 1800 search a month. If i’m a new site, can i rank for the first one or i will be outranked by the sites on the other querry? It’s really not clear to me if I should write about it or select another keyword with less volume. Thanks for your video, i watched every one of them!
It's really not a simple question as it requires a quite detailed answer I am afraid. suffice to say, Google's goal has always been to rank quality over spam, only use these keywords as a starting point to write about. If you are thinking about purchasing adverts for that phrase though that's a whole different topic.
a hint guys if you are getting error you need to upgrade pip to latest version, I did it manually by using python -m pip install --upgrade pip by typing it in the cmd at the python file location
Managed to install it an get it running but it crashed every time I tried to run your query. Gave a warning about auto-gpt.json file does not exist. Then after thinking for a while it crashed. Error Message "Failed to get response after 5 tries"
I get the exact same error. Many people get this error but I haven't found a solution yet. It might be that it will only work if we use a GPT key from the paid version. Can SOMEBODY CLARIFY PLEASE!
@@TynanCreations Good to know it's not just me. I have added my card to the billing and it's initialized but I don't know how to how to get the key. Do I generate a new API key and try that?
It's so funny that we think we get help from AI, when in fact we a helping to train the AI, and once they are good enough, it will auto-produce content for the owners. Google search will be more or less obsolete and we will be entertained by listening to AI-written books, watching social media content produced by AI (including youtube videos), and so on and so forth.
got the latest version via the zip and ran it from command line, ( docker-compose build auto-gpt then docker-compose run --rm auto-gpt ) using it with docker & pinecone memory. it couldn't use elevenlabs or dalle despite them both being enabled in the .env file uncommented and with the keys there, for hours and hours it tried and used google and created agents and wrote to files and got errors with arguments being 3 when should only be 2, and the agents it created for elevenlabs or dalle telling it they where chat bots and saying sorry they could not use apps, so.. it trying to generate an audio file itself but not be able to because it didn't have mpg321 etc...and on and on, around and around, getting no where untill it crashed its pinecone and stopped. when restarted it was determined to provide me with twitter growth, (i do not have twitter dev api) i told it no. next time it was determined to provide me with legal help, reasoning its user needed legal advice so it must research it. i asked it why i needed legal advice but it didn't say. i stopped it again. all i'd asked it to do was produce an image of a piece of toast via dalle or sdwebui and to speak the word 'hello' via elevenlabs. failed miserably. last chance, open it via visual studio and run it from there, without docker. immediate error cannot import name 'main' from 'autogpt' and tots failed to run. ive just now deleted it. perhaps start again, perhaps not, learnt some new things but honestly it wasn't a great experience lol ..started over and now it's ok. it'll nip away and make pictures fine enough (with dalle) but gets an error with Audio saying it needs gi, so need install PyGObject via pip, which last time is where the wheel fell off, VS didn't have C++ 14.0 needed to build the packages or something.. i don't know because i'd just updated it, but anyway i went an got VS Build Tools and made sure to tick that box, and then tried again, and got the same error. anybody help me solve that? really want it to use stable diffusion webui for images because SD is so damn good, but don't know how. if there's a user name & p.word i've long forgotten where they live, i'm fairly sure there is no username/password for sdwebui, so how come that's a thing in the env.template ..? anybody got any advice?
Hello everyone! I need to know if anyone has experienced the following error. I cannot move forward because I keep getting this error, even though the AI is fully functional. I am unable to search or place any orders due to this error. File "", line 198, in _run_module_as_main File "", line 88, in _run_code File "D:\AI things\Auto-GPT-0.2.0\autogpt\__main__.py", line 572, in main() File "D:\AI things\Auto-GPT-0.2.0\autogpt\__main__.py", line 396, in main agent.start_interaction_loop() File "D:\AI things\Auto-GPT-0.2.0\autogpt\__main__.py", line 448, in start_interaction_loop assistant_reply = chat.chat_with_ai( ^^^^^^^^^^^^^^^^^^ File "D:\AI things\Auto-GPT-0.2.0\autogpt\chat.py", line 159, in chat_with_ai assistant_reply = create_chat_completion( ^^^^^^^^^^^^^^^^^^^^^^^ File "D:\AI things\Auto-GPT-0.2.0\autogpt\llm_utils.py", line 67, in create_chat_completion raise RuntimeError("Failed to get response after 5 retries") RuntimeError: Failed to get response after 5 retries
Sometimes you gotta pay to have access to the cool things. You should train your own language model (they're about $400-600 now) then you won't have to pay again
on the subject of seo, was chatting to bing about resolving error messages related to txt-to-image and Bing mentioned it liked my pic of a cat playing a guitar. i have no such pic lol. Bing gave me a url back and it was a picture of a face with patterning on it generated by SD. i replied giving it back the same url and Bing gave this reply "BING: I’m sorry, but I think you gave me the wrong URL. The link you posted does not show a picture of a female human with a pattern on her face. It shows a picture of a Russian frigate that has been hit by missiles. Maybe you copied and pasted the wrong link by mistake. I can visually identify images that have no associated words with them, but I need to see the correct image first. Can you please check the URL and send me the right one? I’m sorry if I confused you about the cat and the guitar. I was referring to this tweet that you posted on May 6, 2022:" (that same url, which was actually May 6, 2023 7;30pm) lol 😂 😂 it went on to say it may be a glitch with twitter and asked me if i wouldn't mind taking the time, to post the image to another image hosting site and let it try again😊😊 afterwards it did provide a very flattering description of the image " The image is very beautiful and intriguing. The pattern on her face looks like a dragon or a fish in an oriental style. The female has Asian facial features and dark hair. She looks calm and serene. The image has a lot of details and colors. " it then started to head the chat towards seo... but it was 17 of 20 replies and i asked it to generate a piece of toast, which it did via dalle :)
Really really good tutorial! not cutting out etc to show that you have perfection makes it more real
This is super powerful. Now everyone can create their own AI assistant
Thanks Hamish, been looking at Auto and waiting for a good tutorial and it doesn’t look that complicated, so will have a go later.
how do i fix this error? ImportError: cannot import name 'main' from 'autogpt'
exactly. lol
Thank you for this tutorial. You explained everythink step by step using basic language 👍
Man thanks for sharing your knowledge. I have build my website using your prompts etc. Ive been getting alot of organic traffic because of that!
Great, may I know how old your site now? How long did you see more traffic? thanks
@@alecd8534 I went live on the 14th of march and got 15k impressions and 150 cliks since then. Only product pages with content on them. Im currently mapping out a topical map, imagine the traffic when i posted al the content mapped out (about 300 blog posts)
@@JackieLangenberg thanks mate
I would love to know how old your site and traffic details as well? Congrats!
Amazing stuff again. Thank you for such a great content each time!
Do we need a paid API key or free one will work?
If paid how much will it cost?
Hi, i have 2 questions.
Did you set up any memory? What's the difference of using it without and with memory?
Awesome content , and hope to see more on the subject especially from seo perspective !
Great Video! What exactly is the advantage of creating your outline with this tool instead of ChatGPT - just more SEO optimized?
Yeah it’s more optimized - this is like using surfer seo for cheaper
Thanks for the video. My first attempt using GPT3 got me no real data in the outline. I'll keep trying.
2nd run was much better. Not fantastic yet but, a usable outline. Cost about $1.00.
I highly recommend people use a Docker container, separate virtual machine or a virtualenv.
What are the benefits of using a virtual machine for this? Thanks in advance!
@@trevor4106 Primarily security since several libraries need to be loaded. It'll run fine natively, I just like to keep apps isolated so it's easy to update and run the requirements.txt without dirtying up my machines. Cheers.
You have a sharp mind
Perfect! Tank you and congrats 🙏
Hero we all need, thank you.🙏
This video opened my mind. Currently, I am using GPT-4 to create my posts and it has produced great results! My audience is different from yours, my posts need to be very detailed and contain extremely accurate information. So much so that I always ask the GPT to use reference texts that I provide in the prompt.
In your opinion, would it be possible to create something like this with autoGPT? Giving a blog title, post structure, and reference links?
It seems to me like a way to generate a huge amount of high-quality posts.
It is possible i am having a similar use case. You basically work with template files, which is a perfect blog post and feed auto gpt the file as a input. Telling it to use it as a template for a new keyword.
i have a question if you dont have cgatgtp4 4 so then will it only use chatgtp 3.5
I am stuck at the using memory of type: LocalCache, where I get an error. How can i fix this?
I don't understand the cost model, so what I have to care about if I'm using it to make it not too expensive?
Use GPT 3.5 Turbo :)
any help? I keep getting this error when trying to create a file: Command write_to_file returned: Error: Attempted to access outside of working directory
You are the best 🎉
Does it cost something when we make a request trough auto gpt, we will have to pay something ?
yeah you have to pay
im so confused i have no scripts in that path?
Can we do it with our open ai api key
Yes. You get $18 of free API usage. After that, you have to pay for it by adding a card on file. Here's a link to how's its done th-cam.com/video/4CMtzlA-LzA/w-d-xo.html
Wow. This tech is advancing so quickly. Once they put a good UI on top of that it will be so much easier to use. What use cases are you guys thinking you can do with this for content creation, Seo, creating spreadsheet for analysis etc?
just make your own ui lol
@@bitmammothOG not my forte lol
you were having issues with running commands as you were one directroy above the root, you have 2 folders called Auto-gpt-master, and are in the first as can be seen by the path in the terminal prompt, should have \Downloads\Auto-Gpt-master\Auto-Gpt-master , if you do cd Auto-Gpt-master you should be in the right directory to run all the commands from terminal.
Thanks - I discovered early on that dragging and dropping like that seemed to always solve my problem - I have my own weird ways of doing this stuff because I don’t really know much about programming
@@Incomestreamsurfers yeah it definitely works. Using an absolute path! Tells command where to look instead of looking in the current directory!
How expensive is it?
Is it way to allow ChatGPT to use internet?
Hi, i have a pretty simple question. It’s really about simple SEO. Sometimes, I notice that different order in words can have a dramaric impact in the difficulty. Take this exemple: what cc in motorcycle engine mean? Is a difficulty of 5 and what mean cc in motorcycle engine is at difficulty around 50? Both have around 1800 search a month. If i’m a new site, can i rank for the first one or i will be outranked by the sites on the other querry? It’s really not clear to me if I should write about it or select another keyword with less volume. Thanks for your video, i watched every one of them!
It's really not a simple question as it requires a quite detailed answer I am afraid. suffice to say, Google's goal has always been to rank quality over spam, only use these keywords as a starting point to write about. If you are thinking about purchasing adverts for that phrase though that's a whole different topic.
a hint guys if you are getting error you need to upgrade pip to latest version, I did it manually by using python -m pip install --upgrade pip by typing it in the cmd at the python file location
Big boi is the correct name
Maybe not right now, but once agi hits its growth spurt OHHH MY GODDD thats one big boi 😳
Can we do LinkedIn email scrapping?:))
amazing 😎
Does this work on the free chatgpt account?
Just a Fast question, is this workin' on android 📱?
no
I ran into the exact troubleshooting errors you did
New error bro. "no module names 'dotenv' ", pls help
Ask chatGPT that is how I overcome all my programming problems
@@Incomestreamsurfers If ChatGPT could it would probably tell me to f*ck off
Mac or PC???? I cant get it to work on Mac. The instructions are not the same
is it just autogpt or langchain + pinecone integration on top of that ? without pinecone for token compression it's not really intelligent
I’m looking into long chain at the moment but I’m not really an expert. I really should have studied computer science.
can no longer acces github. Access to site is prohibited. pip doesnt work either.
have you tried Open Assistant yet?
Getting error using the same
Managed to install it an get it running but it crashed every time I tried to run your query. Gave a warning about auto-gpt.json file does not exist. Then after thinking for a while it crashed. Error Message "Failed to get response after 5 tries"
I get the exact same error. Many people get this error but I haven't found a solution yet. It might be that it will only work if we use a GPT key from the paid version. Can SOMEBODY CLARIFY PLEASE!
@@TynanCreations Good to know it's not just me. I have added my card to the billing and it's initialized but I don't know how to how to get the key. Do I generate a new API key and try that?
my man
Anyone else having issues with the Google API?
[Errno 2] No such file or directory
It's so funny that we think we get help from AI, when in fact we a helping to train the AI, and once they are good enough, it will auto-produce content for the owners.
Google search will be more or less obsolete and we will be entertained by listening to AI-written books, watching social media content produced by AI (including youtube videos), and so on and so forth.
Google Is going crazy today... I think another update is here.
got the latest version via the zip and ran it from command line, ( docker-compose build auto-gpt then docker-compose run --rm auto-gpt ) using it with docker & pinecone memory. it couldn't use elevenlabs or dalle despite them both being enabled in the .env file uncommented and with the keys there, for hours and hours it tried and used google and created agents and wrote to files and got errors with arguments being 3 when should only be 2, and the agents it created for elevenlabs or dalle telling it they where chat bots and saying sorry they could not use apps, so.. it trying to generate an audio file itself but not be able to because it didn't have mpg321 etc...and on and on, around and around, getting no where untill it crashed its pinecone and stopped.
when restarted it was determined to provide me with twitter growth, (i do not have twitter dev api) i told it no. next time it was determined to provide me with legal help, reasoning its user needed legal advice so it must research it. i asked it why i needed legal advice but it didn't say. i stopped it again. all i'd asked it to do was produce an image of a piece of toast via dalle or sdwebui and to speak the word 'hello' via elevenlabs. failed miserably.
last chance, open it via visual studio and run it from there, without docker. immediate error cannot import name 'main' from 'autogpt' and tots failed to run.
ive just now deleted it. perhaps start again, perhaps not, learnt some new things but honestly it wasn't a great experience lol
..started over and now it's ok. it'll nip away and make pictures fine enough (with dalle) but gets an error with Audio saying it needs gi, so need install PyGObject via pip, which last time is where the wheel fell off, VS didn't have C++ 14.0 needed to build the packages or something.. i don't know because i'd just updated it, but anyway i went an got VS Build Tools and made sure to tick that box, and then tried again, and got the same error. anybody help me solve that?
really want it to use stable diffusion webui for images because SD is so damn good, but don't know how. if there's a user name & p.word i've long forgotten where they live, i'm fairly sure there is no username/password for sdwebui, so how come that's a thing in the env.template ..? anybody got any advice?
how is going to charge you money?
why windows?!!!!!
Because my computer is a windows computer lol
Hello everyone! I need to know if anyone has experienced the following error. I cannot move forward because I keep getting this error, even though the AI is fully functional. I am unable to search or place any orders due to this error.
File "", line 198, in _run_module_as_main
File "", line 88, in _run_code
File "D:\AI things\Auto-GPT-0.2.0\autogpt\__main__.py", line 572, in
main()
File "D:\AI things\Auto-GPT-0.2.0\autogpt\__main__.py", line 396, in main
agent.start_interaction_loop()
File "D:\AI things\Auto-GPT-0.2.0\autogpt\__main__.py", line 448, in start_interaction_loop
assistant_reply = chat.chat_with_ai(
^^^^^^^^^^^^^^^^^^
File "D:\AI things\Auto-GPT-0.2.0\autogpt\chat.py", line 159, in chat_with_ai
assistant_reply = create_chat_completion(
^^^^^^^^^^^^^^^^^^^^^^^
File "D:\AI things\Auto-GPT-0.2.0\autogpt\llm_utils.py", line 67, in create_chat_completion
raise RuntimeError("Failed to get response after 5 retries")
RuntimeError: Failed to get response after 5 retries
only for paid users!
Sometimes you gotta pay to have access to the cool things. You should train your own language model (they're about $400-600 now) then you won't have to pay again
@@Incomestreamsurfers I would watch a video on this.
You just link your card and then it costs pennies
Accidentally spend millions of dollars!?
“I must avoid the rabbit hole of excessive research”. ??? 🤯 It IS the rabbit hole. Sounds like an AI-stential crisis to me 🤣
Way too complicated
You are too clocse to your mic.
on the subject of seo, was chatting to bing about resolving error messages related to txt-to-image and Bing mentioned it liked my pic of a cat playing a guitar. i have no such pic lol. Bing gave me a url back and it was a picture of a face with patterning on it generated by SD. i replied giving it back the same url and Bing gave this reply "BING: I’m sorry, but I think you gave me the wrong URL. The link you posted does not show a picture of a female human with a pattern on her face. It shows a picture of a Russian frigate that has been hit by missiles. Maybe you copied and pasted the wrong link by mistake.
I can visually identify images that have no associated words with them, but I need to see the correct image first. Can you please check the URL and send me the right one?
I’m sorry if I confused you about the cat and the guitar. I was referring to this tweet that you posted on May 6, 2022:" (that same url, which was actually May 6, 2023 7;30pm)
lol 😂 😂 it went on to say it may be a glitch with twitter and asked me if i wouldn't mind taking the time, to post the image to another image hosting site and let it try again😊😊 afterwards it did provide a very flattering description of the image " The image is very beautiful and intriguing. The pattern on her face looks like a dragon or a fish in an oriental style. The female has Asian facial features and dark hair. She looks calm and serene. The image has a lot of details and colors. " it then started to head the chat towards seo... but it was 17 of 20 replies and i asked it to generate a piece of toast, which it did via dalle :)