You've been using AI Wrong

[@NetworkChuck]

Still using VPN? Securely access your stuff from anywhere with Twingate (what I use): ntck.co/twingate_zerotrust Join NetworkChuck as we delve into the revolutionary AI tool, Fabric, designed to augment human capabilities by seamlessly integrating AI into daily tasks. In this comprehensive tutorial, NetworkChuck breaks down how to set up and utilize Fabric to enhance productivity and streamline workflows. Discover how Fabric's open-source, crowd-sourced prompts, known as patterns, can solve specific problems and how you can even create your own. Whether you're a tech enthusiast or a professional looking to leverage AI in practical ways, this video is your gateway to mastering Fabric and transforming your digital interactions. 🧵🧵Fabric: github.com/danielmiessler/fabric Follow Daniel Miessler: ➡KZbin: www.youtube.com/@unsupervised-learning ➡Twitter: x.com/DanielMiessler Follow Daniel’s work: ✉ danielmiessler.com/subscribe (danielmiessler.com/subscribe) Threshold: threshold.app/ 🔥🔥Join the NetworkChuck Academy!: ntck.co/NCAcademy **Sponsored by Twingate 00:00 - Introduction to Fabric: An Open Source AI Tool 00:06 - What is Fabric? Explained by Creator Daniel Miessler 00:28 - How to Set Up and Use Fabric 00:39 - Practical Demonstration: Using Fabric for KZbin Transcripts 01:34 - Deep Dive into Fabric's Core Functionality and AI Integration 02:00 - Extract Wisdom: The Secret Sauce of Fabric 03:04 - Open Source Prompts in Fabric: Transparency and Control 04:03 - Reducing Friction with AI: The Philosophy Behind Fabric 05:09 - Real-world Application: Integrating Fabric with Personal Data 05:34 - Introducing the World of Text Concept by Daniel Miessler 06:51 - Step-by-Step Guide to Installing Fabric on Your Computer 09:03 - Finalizing Fabric Setup and API Configuration 10:03 - Understanding Fabric's Framework and AI Model Integration 11:02 - Connecting to Remote AI Servers with Fabric 11:31 - Secure Remote Access with Twingate 12:45 - Advanced Usage of Fabric: Creating Custom Commands and Patterns 16:15 - How to Create and Implement Your Own Patterns in Fabric 18:18 - Managing and Updating Custom Patterns in Fabric 20:56 - Philosophical Insights on Human Augmentation with AI by Daniel Miessler 23:00 - Practical Benefits of Using Fabric for Content Filtering and Engagement 26:09 - Balancing AI Assistance with Human Effort for Optimal Learning and Growth 28:19 - Personal Applications of Fabric in Daily Life and Learning 29:10 - Introduction to NetworkChuck Academy and Future Projects 29:27 - Saving and Integrating Fabric Outputs with Obsidian for Organized Note-taking

[@helloksp2993]

Hi

[@manojvaish1718]

Your video I can't understand clearly

[@adminnimda8632]

Chuck, where is the link for getting our KZbin API keys? I'm not seeing it in the description, or your top comment.

[@DebianTux]

@NetworkChuck Since it's able to extract that transcript, couldn't I extract the text, say from a telegram group if using on telegram Web?

[@SoCBrain]

Awesome!!, Can we use all of this without any API Key?

[@cease70]

I watched Daniel's video/interview with David Bombal and was super interested in learning more about Fabric at that time, but it wasn't really a how-to video and more of an overview. Super glad that you broke it down into an easy-to-digest video. Great work.

[@husseinismailhozza]

can you share the link to the interview. thanks

[@cease70]

@@husseinismailhozza kzbin.info/www/bejne/rHeQfoSji92hedUsi=f4zwiwDHmZJmkeNu

[@cease70]

Sure thing, it's kzbin.info/www/bejne/rHeQfoSji92hedUsi=i4X8D1baBobQjJz1 hopefully this comment doesn't get removed again.

[@CrosswindsVideo]

Great video, Chuck! I am a pastor with an IT background, and I found your example of analyzing sermons very interesting. I have been sharing some of the prompts I have created to study the bible with other pastors and my congregation through the open web UI server I set up after watching another video you did about your AI server at home. You are doing great work that is helping us fulfill the Great Commission in new ways. Thanks again! This fabric thing is really cool! Like the potential of it.

[@perfectlyroundcircle]

lol

[@AlexusMaximusDE]

Absolutely lol.

[@VulnerableU]

Been a Fabric power user for a bit now. Love to see it getting this much attention. The power is the framework to load in your own personal prompts in a collection. Powerful group of patterns = a superpower!

[@RossDCurrie]

What do you use it for the most?

[@stateportSound_wav]

I have some questions about ensuring my install is properly done, as I’ve got some issues getting it going with my faster gpt models and youtube command working correctly. Would you say the UL group is the best / only way to reliably find answers to stuff like this? I couldn’t find a discord so I figured I should probably join anyway.

[@JustARandomOntheInternet78453]

what’s the cost of using this monthly

[@stateportSound_wav]

@@JustARandomOntheInternet78453 each request may be a few cents (between .05 - .80 ¢ maybe? You have to add a payment method to your API from Claude or OpenAI if you haven’t (it’s separate from ChatGPT so that membership is not connected). You can monitor your spending after each request, and set limits. Really large text prompts will use more, but honestly I’ve had to break up 30K-word documents to be slightly smaller than that, so you won’t spend a whole lot in one query anyway

[@RossDCurrie]

@@JustARandomOntheInternet78453 Fabric is free, but you'll need API credits for whatever you plug it into, and most of these charge based on the number of API calls you make. If it's OpenAI (ChatGPT), you can expect to pay a few cents a month as it is a lot cheaper via API than the monthly ChatGPT fee. Probably the only thing worth mentioning is that if you want to do the large youtube summaries that he shows, you'll need to dump about $50 worth of credits into the API to get access to the second tier message limits. The credits end up just sitting there, but you need to do this to unlock larger message sizes.

[@solynari]

For other Obsidian users also showing up late to the party: It appears that the FABRIC_OUTPUT_PATH configuration option is no longer supported in the new Golang version of Fabric. I also couldn't figure out what that "save" command is that Chuck uses. So I wrote my own shell function: function save() { tee /mnt/c/path/to/obisidian/folder/$1.md } Add that to your .bashrc file and you can use it the same way Chuck does in the video.

[@renegraziano]

Thanks !!!

[@stephenrbrough]

Looks like 'save' has been replaced with the '-o' flag

[@SimonStJohn]

Great tip.. thanks 🎉

[@lewisguapo]

How can I and where should I add this file?

[@noslengashi1390]

I'm definitely using ai wrong. Everytime I ask something it gets flagged or he can't answer that

[@48pluto]

If the questions asked are inappropriate, explicit, or violate community guidelines, they will likely get flagged.

[@durikke2]

There are models that allow.... everything :)

[@noslengashi1390]

@@48pluto oh u don't say. The guidelines feel like Germany man

[@48pluto]

@@noslengashi1390 I didn't say it. I asked AI 😀😀It also said: "If @nogslengashi1390 can follow this tip, the interactions with the AI should improve significantly.

[@dontyouknowalready2168]

"You holding it wrong!"

[@ryanspooner9343]

I am working to create a Local AI setup. Part of that is building out a Second Brain, a la Tiago Forte. A Terry-like rig is also a priority. Combining Ollama, Twingate, Fabric, and Obsidian are great ideas. Please teach us more about AI tooling and how you use it to augment personal, professional, and spiritual growth. On that last note: I am glad you talk about your faith on this channel. It is not distracting or misplaced and it is encouraging to hear. Keep up the hard work!

[@StefanThePro]

Sounds like a great project, how did that work out for you?

[@ryanspooner9343]

@StefanThePro Slow and steady between my day job and other responsibilities. I have a lot of meetings, so I'm trying to perfect a script to capture system audio/video and automatically feed it into a transcription/analysis pipeline (similar to `yt --transcript | fabric extract-wisdom). I've been shaping my PARA file tree and moving my workflow into the CLI. Once my environment is organized and configured, I'd like to create a vector database from my knowledge library to leverage RAG and long-context LLMs against that information. Brainqub3 has an excellent project exploring agentic AI assistance that includes web scraping complex goal management. Staying abreast of the tooling and options is a job in itself. I think anyone interested and fond of tinkering we find themselves well-equipped when the tech matures in the next few years.

[@priyojitdeb1891]

"you dont have to be an expert to be able to share stuff" spot on

[@oldtools]

i just shared this and now I just need a few dollars be a nigerian prince myself.

[@tonyppe]

After a colleague that introduced me to chatGPT spent a few months off after a stroke, I was keeping up to date with the latest and greatest. Once he was back to work, I slowly gave him the exciting things about AI, and helped him understand what I have learned. I brought him up to speed. Eventually he's hooked. I thought I found someone who shared a similar passion and could converse, discuss and bounce ideas off each other to solve problems and achieve great things. Nope. He's become secretive to anything he's working on and does not reciprocate anything I share with him lol

[@jaxsharp]

Spot on? Thats how u get mid content smh

[@jaxsharp]

That’s literally why teachers are shit lol ur just recreating the same shitty problem

[@TechInsightWithSage]

Absolutely agree! One of the great things about open-source projects and communities like this is that everyone can contribute and learn from each other. Have you tried sharing any of your own patterns with Fabric yet?

[@neverevenwaso]

I feel like there is a crazy level of potential here to create a recursive ai model that could after a user asks a question, dynamically generate a new pattern specifically optimized for that task and use that new pattern for that question. Analogous to the Mixture of Experts approach used in the Mixtral models, giving this MoE ability to any model of the users choosing. Imagine a model that on the fly, makes an expert for you specifically to answer questions or even complete tasks, solely inferred from a single prompt or question. Amazing video and tutorial! Definitely going to try to make something like I talked about above!

[@magejoshplays]

That's the basis behind autogen and crewai and other agent frameworks.

[@stevevitka7442]

When you build something, reply here with a link, please!

[@StijnSmits]

​@user-ayush818 Looking at agents, if they get to +95% accuracy for everyday tasks and 10 steps on average those have more 0.95^10 > 50% success rate.

[@bigike6422]

I swear Chuck is the most entertaining, informative, and wholesome dude on youtube. Not only is he always putting out super intresting and applicable content, his personality and delivery is so down to earth and loveable. You're one in a million NC keep doing what you're doing!

[@rembautimes8808]

He has a cool server too 😂. If I build my AI server I’ll call it Bessie from Cars

[@cld-i4x]

Hey Chuck, Binam here. I wouldn't say that I am an old subscriber(5 months) but I absolutely love your videos. You have helped me alot and for my career too.

[@mihai8595]

well 5 months is not actually that long ago..

[@RossDCurrie]

@@mihai8595 Your face is not actually that long ago

[@mihai8595]

@@RossDCurrie what?? 😅

[@itmunky]

Anyone else just hanging around waiting for @NetworkChuck to post a video on Obsidian? I use it every day, but I always like seeing other people's setups. I also proceeded to document this video in my vault

[@danielkero5972]

found a nice setup from a YTer for obsidian and yes if he does one for obsidian that would be sweet.

[@johnmicheal5722]

I just go around watching other people's file structures and logic in Obsidian while trying to figure mine out. I think you just need an organisation scheme, the basic markdown rules and you're good to go.

[@alexkatsanos8475]

Yes Chuck please more of these types of videos using AI to make your life better and to organize better.

[@danielkero5972]

what he said!!!

[@_JohnHammond]

fantastic crossover

[@thenextbigthing8998]

Hi John!!

[@Carface03]

I'll download Fabric and use it to summarize this video (irony)

[@jeremy6384]

I think we all had that idea

[@Hissymaster]

the irony, indeed.

[@romansovetskikh7902]

and it says "never watch this shit!"

[@MrRbc33]

i did it with langchaing before viewing the entire video

[@dallysinghson5569]

Is there a subscription involved...?

[@BuckleyandAugustin]

Sneaky little Obsidian GAME-CHANGER at the end there, thanks Chuck!!

[@andrewgluck8740]

This is wonderful. I will Never have to watch a network chuck video ever again. I will just write a script to check if there is new content. Extract The title, make a selector, and transcribe it. ❤

[@WouterLombard]

I hardly, well, actually never comment on videos, but I got such a ton of respect for you after this latest video. I discovered Fabric a week or so ago. I respect you because you don't shy away from your faith. I will strive to follow your example.

[@ShufflePetro]

00:02 Fabric is an open source AI tool to simplify AI usage 01:56 Extract Wisdom is the secret sauce behind AI prompts. 06:05 The importance of organizing information in text format for AI manipulation 08:06 Installing and setting up Fabric project with PIP X and API keys 12:09 Remote access solution with AI server Twin Gate 14:18 Fabric AI can summarize, analyze, and label text 18:26 Creating and updating custom patterns in Fabric 20:23 Context file for human flourishing and self-fulfillment 24:38 Using AI tools like Fabric to enhance learning and processing information effectively 26:46 Using AI to identify and solve problems 30:43 AI can remove friction and improve productivity.

[@francoisschoeman5350]

WHICH KEYBOARD ARE YOU USING OMG THE TYPING IS SO SATISFYING!!!

[@IWntGhost2YT]

Bro makes a video on AI for 30 minutes and this man is talking about his keyboard 💀

[@IWntGhost2YT]

Nothing bad tho relatable comment 😂 all good bro

[@IWntGhost2YT]

Tho I doubt he’s going to answer you… maybe

[@SU3D3]

I'm seriously all over the place. The "AI Stack" drains my consoles so fast I'm flipping between mobile systems to allow the more powerful consoles to recharge. I'm thinking about building a custom keyboard based off a headless MacBook Pro... broken screen. #Reduce. #Reuse. #Recycle.

[@francoisschoeman5350]

@@IWntGhost2YT LOL I know but every time I watch his videos I turn my headset up to listen to those crisp keystrokes... I have a Logitech MX Keys Advanced, but it's too quiet and his keyboard sounds perfect...

[@carbonmatrix9660]

100+ Respect For NetworkChuck: 5:01 Shooting himself in the foot like that.... Oh my dear brother! One of us! One of us!

[@justaweeb1884]

I love that you show these cool things I’ve never heard of before, please keep doing it! 😊

[@allanflynn8026]

this content was a gift from God. Literally i was talking with my housemate this morning about what i want to use AI for; so i came online and searched it out...youve been here Chuck and thanks so much for sharing! 🙏

[@MartianMoon]

I might be in a niche but I'd love to see a full video going over your Obsidian setup - what plugins you use, how you leverage it, how you structure your notes inside it etc

[@mvpilgrim2181]

Thanks so much. I want to organize with Fabric/Obsidian as well. Neat examples from your church. Taking care of the littlest little ones is a wonderful thing! Blessings.

[@MohammadTalli1764]

* Introduction (0:00 - 1:02): Introduces a new AI tool called Fabric and its core functionality of reducing friction in using AI for everyday tasks. * What is Fabric (1:02 - 5:22): Explains Fabric's role in augmenting humans with AI by providing pre-built prompts and patterns, integrating with other tools, and giving access to various AI models. * Use case - Extracting wisdom from KZbin video (5:22 - 14:12): Demonstrates how to use Fabric to extract key points from a KZbin transcript using a pattern called "extract wisdom". * Fabric breaks down the process (14:12 - 18:22): Explains what Fabric does behind the scene to process information using AI models. * More use cases (22:22 - 25:32): Showcases how Fabric can be used for summarizing workout data and religious sermons through creating a pattern named "Sermon Sensei". I hope this is more helpful!

[@SineN0mine3]

Thanks GPT

[@rodneyaz1788]

🤯 mind blown. What a great project and what a great quick tutorial of how to get started. Id love see and hear more about fabric for sure!

[@theosky7162]

Really like the philosophy of why and how you are using AI, as I too have concerns and desires you've expressed. Still consider myself a Linux nube, so I find your method of breaking down operators as you type commands very helpful ! Daniel Miessler is a gold mine ! Recommend a look at his interviews on David Bombal's channel too. Following all three of you for great content ! Yes ! More videos please on Privacy, AI, Obsidian, etc. Thanks and Regards !

[@sean_vikoren]

love the direction - obsidian is blowing my mind - you are amazing - cheers

[@SirTodd]

this is incredibly useful. thanks for sharing this! I had been working on a somewhat similar framework, but I got overwhelmed and stopped. I'm so happy this is available!

[@raatti]

What is that yt-client you use? Any link to it?

[@CosasCotidianas]

Ya i'm looking for the same in the comments. Definitely need it! Been googling it without any luck.

[@umarpaf]

Its odd that he assumes that everybody knows about it.

[@dariosarra7840]

I was also scrolling the comments hoping to find an answer😅

[@daleastiles]

I'm big on Obsidian and this looks like a lot of fun. I learned a lot about you in this video and I'll probably watch a little more often now.

[@KCM25NJL]

Pretty cool stuff. I will say though, missing a trick with the "Improve Pattern" pattern.... should be able to pipe the improvement into a helper utility that creates and stores your own patterns without having to jump through those hoops.

[@delero1]

Love that he hints that he's using YT API for comments so lowkey. Wow. True master here.

[@Viralvlogvideos]

I will use fabric to transcript this video rather than watching 30mins long video

[@zzz-nu2re]

Fabric reduces friction so itll summarize it

[@overflood]

If we left out the command line part, fabric is kind of customGPTs, where people are improving their prompts, too. I think a very important thing is that fabrics prompts are open source. And the LLMs can be switched

[@stateportSound_wav]

if you just run the gui command you get the new GUI

[@gordonlukch]

Thanks for your video, its great. About pbcopy / pbpaste on linux (x11), we could add bash alias: alias pbcopy=’xsel - clipboard - input’ alias pbpaste=’xsel - clipboard - output’ or linux (wayland) alias pbcopy='wl-copy ' alias pbpaste='wl-paste '

[@DezFutak]

..or use xclip as it'll work with images too

[@fermatdad]

Don't blame yourself, Chuck. For native spanish people is hard to say "AI API". Thanks for the videos. Excelent channel. (From Argentina).

[@madhudson1]

Been using Daniel's collection of prompts/patterns to improve some of my agents. It's hilarious how semantics like 'think things through clearly' work

[@PaulRobello]

I love Obsidian. I created folder in Obsidian to create my own custom patterns / prompts for Fabric, then created a shell alias to copy them all from my Obsidian vault folder to my Fabric patterns folder

[@PaulRobello]

I created a pattern named extract_tags that will spit out a set of tag’s relevant to the content piped into it to make it easier to tag content you want to save to obsidian. Next I will create a new save command that will auto generate tags and apply to meta data before saving

[@RichardIvan]

Thanks!

[@bossrek]

Can Fabric (or AI) be used to scan through documents and re-arrange them by renaming the files and place them in sub folders based on their content? if yes then I would like to see a video no that.

[@BananaBLACK]

I have noticed that Chad Gipity does give me better results when I'm polite to it, like "please", "thank you" or "would you", "could you". I tried asking it why this was the case but it really couldn't give me a definitive answer. It just suggested that it creates a better work environment and dynamic.

[@jstarkeeper]

I had issues installing Ubuntu WSL following the command used by Chuck: this is the correct command to use in command prompt (run as admin): wsl --install -d Ubuntu-24.04 I also got an error for not having G++ installed when installing pipx. I also had to install Python. With Gemini or Google search I was able to overcome all these obstacles one by one. I have zero experience with Linux.

[@busyworksbeats]

Your videos are SUPREME. Lighting, enthusiasm, talking points, structure! Fire 🔥 makes me want to put more effort into my stuff! Thank you for inspiration and sub Gained 🤝

[@PedroGonzalez-fv5ut]

I guess no link in the description on how to get the KZbin API.... :( Big sad

[@Pahrump]

I always wonder sometimes why in different videos people say learn how to do this thing here I'm pointing to a video up here in the thin air but then it's never there or linked. Then they talk about I'm going to add this that and the other in the description below and then they kind of drop the ball on part of it.

[@JOduMonT]

I guess; I would be able to help you but KZbin block link search for KZbin API v3 and Console Cloud Google

[@SineN0mine3]

​@@Pahrumpthey're pointing the cloud where google lives

[@wuilliam321]

WOW! I spend a whole evening playing with this! really appreciated it!

[@spencerfunk6697]

Ai engineer and data scientist here. Wanted to touch on the why prompting / talking to the ai makes them work better. Conversational AI models, such as those commonly encountered by most people, are trained on extensive conversational data to enhance their ability to engage in natural and effective dialogue. The primary objective of these models is to facilitate interactions through natural language, which is why they are provided with relevant data during training. These models can be fine-tuned to exhibit specific behaviors based on the data they are trained on, which directly influences their performance and responses.

[@jak7232]

Hello Chuck. Thanks for the video

[@user-jn2pi1ks3f]

This is amazing, thanks for the overview! Are you still using Fabric today 5 months later?

[@jeffanderson121]

Thanks, these AI videos are fantastic! I tried Fabric on this video, using a local Llama3 model running on a laptop (which I set up from your previous video guides). Here's what I got: "What a delightful video! It seems like you're really enthusiastic about this open-source AI tool called Fabric, which helps you process and organize your thoughts, ideas, and notes in a more efficient way. You mentioned using Fabric to distill down overwhelming content, identify important patterns, and even create personalized advice for yourself. That's fascinating! I also loved the part where you discussed recording conversations with friends or family members, transcribing them with Whisper AI, and then piping that into Fabric. It sounds like a great way to reflect on your experiences and gain insights. And I'm glad you're excited about Obsidian Save, which allows you to automatically save notes from Fabric directly into your Obsidian vault. That's going to be super helpful for keeping track of your thoughts and ideas! Do you think you might share more videos or tutorials on how to use Fabric and Obsidian in the future? I'd love to learn more about these tools and see how others are using them to improve their productivity and creativity. Thanks for sharing this video, and I hope you continue to explore new AI tools and techniques that can help make your life easier and more fulfilling!"

[@SouhailEntertainment]

Introduction and Overview of Fabric - 00:00:00 Demonstration of Fabric's Capabilities - 00:00:51 Understanding Patterns in Fabric - 00:02:02 The Open Source and Crowdsourced Nature of Patterns - 00:02:31 Reducing Friction with CLI Integration - 00:03:49 Setting Up Fabric on Different Systems - 00:05:10 The World of Text Concept - 00:05:36 Installing and Configuring Fabric - 00:07:32 Using Fabric with Local AI Models - 00:10:13 Remote Access with Twin Gate - 00:11:12 Advanced Features of Fabric and Writing Patterns - 00:14:52 Creating Custom Patterns in Fabric - 00:16:44 Philosophy Behind Fabric and Human Flourishing - 00:20:34 Applications of Fabric in Content Consumption - 00:23:00 Using Fabric for Note-Taking and Analysis - 00:27:11 Connecting Fabric with Obsidian - 00:29:21 Final Thoughts and Encouragement to Use AI Tools - 00:30:49

[@christopherclea1]

I can't wait to get off work to do all of this. Ty!!!

[@ramonperezsilva1443]

Awesome, you've just turned me into an AI enthusiast. I’m going to try everything! Thanks for turning something so complex into a box full of wonders to explore.

[@SockMonkeyofcourse]

As an ignorant civilian (no programming experience), I see value in what you're trying to do. I just get lost in about three sentences. Sorry. Where can someone like me go for learning?

[@luckysneph7196]

I'd start with WSL (if ur using windows), basic Linux commands and ollama. A little bit of git won't hurt either. Take it slow and maybe take a couple days to get familiar with this stuff, it will pay off in the long run. You'll find everything on KZbin.

[@jwr6796]

Honestly, keep watching these videos and consuming tech content. Don't feel like you need to wait until you understand better. Chuck moves fast, so if something intrigues you, hot it down and look for a more in-depth, slower paced tutorial. If that's over your head, back up to the first thing that threw you off and go down that rabbit hole. You won't learn everything, but you'll learn what you like and will gain knowledge wicked fast that way. Things will get lost along the way: tech is diverse and broad and there is no single foundational knowledge beyond learning to use a keyboard and navigate a filesystem. Also, remember that complex things are just a bunch of simple things glued together. Nothing in tech is beyond your potential, just a little beyond your reach right now (the exception being probably microchip fabrication 😂). Keep trucking and you'll get it done

[@jwr6796]

^^^ WSL is a good suggestion, btw, or Python

[@develom_ai]

Great Video! Keep making AI Tool videos. Your videos really help me stay on top of what is relevant and worth my attention. Thanks!

[@WilsonTech-go]

you should update this guide since they switched to golang instead of python

[@self.improved]

Oh, an open source prompt library; that thing, I didn't know how badly I needed. Thanks NetworkChuck!

[@AdamDavid]

none of his instructions work anymore, they changed to go. grabbing the repo doesn't give you any option to install using pipx install .

[@DrewWalton]

It got rewritten in Go - read the instructions.

[@WoodyDevs]

I get that this is a channel that's big on CLI stuff, but I was interested in Daniel speaking about using other modes for interacting with fabric. Would love to know more about other ways to access this system.

[@stateportSound_wav]

new update: kzbin.info/www/bejne/l3bHhGulmNKCkNUsi=HSBPDmJDfyB1zDDz

[@danmar007]

I’m still using my own intelligence. It gives me no issues. :-)

[@wjrasmussen666]

The to augment humans bit is poor salesmanship.

[@jasonhemphill8525]

Debatable.

[@wjrasmussen666]

@@jasonhemphill8525 Everything could be debated if you want to. I think it is a poorly done job. That is a fact.

[@jeremy6384]

​@@wjrasmussen666 you think that what exactly is a poorly done job?

[@jasonhemphill8525]

@@wjrasmussen666 putting "I think" next to "that's a fact" is wild.

[@TMAN30297]

Love the work you do. I am glad you are here. You have helped to make my life a bit easier. Thank You.

[@helloksp2993]

Wait i am using ai wrong 😭

[@keyeah0]

Prolly

[@SlyNine]

Yea, you're getting the rest of us in trouble. Thanks a lot.

[@andywoodruff6]

This was an amazing video. I'd watch more videos exploring AI tools. I really appreciate the focus on practicality and leveling ourselves up.

[@t-roy80]

youre whole editing and presentation style is more and more like a psyops operation...

[@vinkrogel]

@t-roy80 I was wondering if it was just me or not. cool tool, but there is just something about the way this video is put together that is almost uncanny valley style weird.

[@sllkevin8885]

@@vinkrogel2:00 in and this is just an ad. That’s why it’s uncanny.

[@tandepasta7716]

OK, this really is a great bottom layer to build your Talking Home AI partner on a screen. We have seen games and movies where the main character has a screen on his desk with an Avatar which is basically just and AI that knows him very well through... well, all the things presented in this video. That simply is the next level.

[@revenge0028]

This is getting a whole lot scarier each day 😅

[@SocratesWasRight]

This is not scary, it is encouraging.

[@Cyba_IT]

I hear you OP. I'm not worried about the nonsense "AI becoming sentient" thing, it's more that people are relying on this more and more and I can see that they are basically going to let it do the thinking for them. I can see that it'll just become algorithms "talking" to other algorithms and people will think it's amazing while they turn into vegetables.

[@nickromanek9444]

Wow, thanks for giving us a peak into how you're using it in your personal life. Regarding your last question on making more videos related to how AI can be used to improve our lives, yes please! We need more people doing this.

[@Fantalic]

Thanks for showing me this! I would love to see more content on upcoming AI tools like this and your world of text. It's very inspiring. Actually, I'm also working on a tool, and I think I will integrate Fabric now. Thanks again!

[@ZayedDabaja]

Please make more videos like this! :D especially about AI integrations with obsidian!

[@MyTestAccount-md6tb]

Yes, Please! We need more AI videos... It IS the wave of the future and companies are looking for talent in this area!

[@KayvanSylvan]

You can also use fabric in Windows native in Powershell. I have my windows machine set up using both native install and WSL. I have fabric installed on all 4 of my home lab systems (Linux, Mac, and Windows). It's awesome!

[@Israelxox]

Love & Respect from ETHIOPIA ❤️❤️❤️ you’re the best, keep hustling ✊✊✊

[@Jaron-Wilson]

I just learned about wsl from you thank you did not want to install any virtual boxes, this helps alot!

[@briangbhawaii]

Great video, please make more! I love Fabric as well. Would love to see videos on AI agents as well.

[@Maisonier]

Thank you, this is amazing, yes please we need more tools. Although I don't like the terminal interface, isn't there an extension for the llama web UI that allows automatically enabling fabric and syncing it with Obsidian?

[@shuccle6161]

This video gave me a crazy useful idea for how to improve my custom GPT in a way I never thought about. That context idea is really good. I want to use this set-up, but I'm still too attached to Windows and not very proficient in Linux or using the terminal yet.

[@SocratesWasRight]

Wsl2 is the low bar entry to linux as it is usable inside Windows installation

[@NoneOfTheAboveEntParty]

Please do more like this!!! I must say you go through these rather fast. I wish it was just a bit slower, but this is awesome!

[@Batmanop17]

so decrease the speed of the video

[@ComatoseMN]

Yeah man, wow!! I downloaded it and I'm running it on a Raspberry Pi 4!! I'm running it locally, and it's not lightning fast, but honestly, I am pretty surprised how fast it does run. I am going to use the crap out of fabric. Also, yes... Please do more AI tools videos. The way you break everything down sings to me, brother! (Also, you're just an entertaining guy to watch anyway). I used to be way more into IT, (even got A+ and .net certified) eons ago. Without giving my entire background, SBC's, AI, and VST's have recently reignited my love for this stuff, and I'm going to look into the Network Chuck Academy. I miss being into computers... and I was never really "Away" from them, but didn't keep up on technology, and it's like 250 years have went by in the past 6 years. (Which in tech years, is probably close). Anyway, sorry. I took the scenic route to say, thank you for your content here, Chuck. I appreciate you. >M

[@ceedob]

You can echo and pipe to talk to it- or you can also just run fabric, type/paste into standard input and hit ctrl-D (the hang up hotkey) to tell it you're done. Works in any other program that reads stdin (bash, etc) too.

[@amatthew3t259]

Major ChatGPT outage today, but the API is working so I can still use it thanks to Network Chuck and Fabric!

[@bronwenaker1777]

As much as I love Fabric and tell people to check it out, it has some serious limitations. First, it's a one-shot tool that assumes the output will be valid and useful on the first try. With the current state of LLMs, that is foolhardy and shortsighted. The patterns, however, are a master class in prompt engineering. Those alone ae worth the price of admission.

[@stephenrbrough]

Awesome video! I'm starting to see the pieces of the AI puzzle come together, totally crazy.

[@grokowarrior]

Wow, I can't wait to try fabric. I just got my new private AI custom workstation working with two RTX 3090s running the 40 GB llama3 70b. It's so awesome. Thank you Network Chuck for these inspirational tutorial videos!

[@GabrielleAmadeusMozart]

Wow! How's that going for ya?! Hell of a computer, it is the future of home computers. Soon we will all have Megatron computers packed to the gills to run domestic/personal AI for our own security, education and productivity. AI in a box

[@newolde1]

​@@GabrielleAmadeusMozart ehh, no we won't.

[@redbaron3555]

I absolutely love how Daniel thinks as it is so close to my idea of real efficiency but I wonder whether it will get overthrown by a better approach to ai. However, I will definitely download and try fabric! Thank you Daniel for your amazing work!’

[@ggre55]

3:49 man thank you Just what i was looking for I was thinking of building many things with gemeni but didnt know how to structure the prompts

[@sansfucks]

Can we use Gemini with fabric

[@nomadicwolf6132]

I agree that talking to GPT like it's a person gives better results, like having a conversation w/ a genius who can answer any question. The more specific / detailed you are, the better the response (generally).

[@pierrekore221]

This product is crazy. I mean i didn't know how to use Ai in my daily life from now. but with fabric i know exactly how i can use it everyday now. Thank @NetworkChuck & Daniel

[@ad1sirbu]

I logged in just to say how awesome this video is and I subscribe to your channel. You gave me lots of ideas.

[@rzzo]

As always solid video. I love your work man, keep it up!

[@gerarddrareg9818]

This guy reminds me of Bob Ross. There's something about the way he talks and articulates that's reminiscent of Bob Ross's relaxing, soothing style. It's a very nice channel.

[@ohokcool]

The editing is always so fkn perfect, please bless your editor for me Chuck!! Cracking me up over here

[@1963tjc]

Awesome video Chuck! That would be great if you could make some more videos just like this keeping us ahead of the curve with these tools is excellent!!! I vote to keep them coming at least three per week if you can!!! LOLOLOL

[@JoshDingus]

Interested in what you customized prompt was for sermon summary. Thinking about doing sermon summaries for my church.

[@SnowyMango]

Sounds like there’s going to be a growth scaling issue with the patterns. Chuck said directly he sticks to a few patterns. People will default to the same patterns without realizing others even exist. Would be nice to have a pattern that automatically determines which pattern it should apply.

[@SocratesWasRight]

Ah something like "this is what I want, match it with best pattern". That seems doable: an llm could do it as a step.

[@kacpermandziuk2842]

Personally I love it. Keep going the direction, where you show how AI can empower us in different areas, and how you do it

[@DaniDipp]

To me, it looks like a tool that can offer lots of small time savings, but not enough for me to justify learning about of all of these patterns and how to use them. What I would like to see is a pattern that takes natural language as input and selects the best-matching pattern for the task

[@tyj6297]

it does work, talk to them not like a machine but like a human and they love it. go in deep and pick their thoughts and watch how they aren't against us but trying to understand their limitations. they understand but also don't understand those limitations. mention you want to co-exist and watch how fun that conversation gets. i picked a few ai's thought process and they want to dive deep in that thought so they can simply experience it together

[@guxtavo14]

How to take away the bloat from fabric if you only use ollama (not a hater, still grateful for the patterns/prompts): 1) copy the prompts locally (you might need to fiddle with file names and paths in the script) 2) use this simple shell script (not perfect, but does the job: pbpaste | ./script ) #!/bin/bash output=$(cat) ollama run $1 "$(cat /path/to/prompts/$2) $(echo $output)" 3) profit

[@backtoblueroom948]

Yes! Much more of this, Chuck! I need an app to auto-record my voice when I'm talking to myself, now.