Episode Categories:
Resources:
Join the discussion: Got something to say? Let us know on LinkedIn and network with other AI leaders
Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineup
Connect with Jordan Wilson: LinkedIn Profile
Start Here Series in our Inner Circle Community: Join for free access
Claude Cowork: The New Standard for AI Productivity on the Desktop
In 2026, AI productivity discussions have zeroed in on a tool quietly setting new expectations for workplace automation: Claude Cowork by Anthropic. While AI coding assistants have saturated tech circles, Claude Cowork targets the broader business community—offering practical, immediate gains in productivity for nontechnical professionals. This detailed summary unpacks exactly how Claude Cowork operates, its current limitations, and, most importantly, the actionable ways it can change how companies approach file management, information retrieval, and workflow automation.
Desktop AI Interface: Direct File Access for Business Tasks
The value of Claude Cowork lies in its desktop-native approach—enabling direct interaction with local files and folders, not just cloud-based services. Unlike standard AI assistants that rely on web-based connectors with limited file access, Claude Cowork grants Claude controlled access to designated folders on a Mac OS device. Once set up, this functionality allows the AI to read, edit, and create files as easily as a human user, driving productivity for countless administrative and operational workflows.
Critically, the system does not randomly scan the machine; access is always user-consented and limited to selected folders, allowing organizations to maintain data security protocols while experimenting with hands-on automation.
Claude Cowork vs. Other AI Desktops: Key Differentiators
Where other AI-powered desktop tools like the desktop ChatGPT fall short, Claude Cowork stands out by integrating three operational domains:
Access to all local files within the chosen directory
Full terminal (command line) interaction for advanced batch processing
Controlled web browser automation through the Claude Code Chrome extension
Unlike Google Gemini or classic cloud-based Claude, Cowork’s use of the local terminal and browser simultaneously enables workflow automation that bridges the “duct tape” gap—where a person was typically required to copy, paste, download, or upload files to move data between applications.
Security and Privacy Considerations for Enterprises
Business leaders will note that this power comes with new security and compliance implications. Claude Cowork’s initial setup requires explicit folder permissions, and users are warned of privacy risks, especially on company-owned devices. For organizations considering deployment, IT buy-in and robust onboarding around local-access privileges become essential. Anthropic’s consistent messaging on potential risks reflects an understanding of enterprise security demands.
Workflow Automation in Practice: Hands-on Use Cases
The episode demonstrates several highly practical workflows that connect day-to-day business needs directly with Claude Cowork’s capabilities.
1. Automated App Creation:
A request for a simple, Kanban-style CRM app that runs locally can be fulfilled in minutes. The user interacts via a conversational interface, selecting pipeline stages and data storage options. Claude Cowork codes, tests, and provides a ready-to-use application—no manual file or code manipulation needed. For small teams, this translates into immediate deployment of micro-tools tailored to unique internal processes without developer bottlenecks.
2. Local Folder Organization:
The system scans large, disorganized folders—like a busy downloads directory with over 180 files—and, after user approval, creates suggested subfolders (e.g., Installers, Videos, Documents). The AI is capable of moving thousands of files into the correct location with a single instruction. For organizations handling sensitive or high-volume client data, such an operation can reclaim hours of manual labor and reduce organizational risk by standardizing file structures.
3. Browser and Document Handling:
By controlling the Chrome browser, Claude Cowork can review social media accounts, analyze tweet and reply history, and even model content style. This opens up automated social media engagement, market research, or competitor analysis—all managed from the desktop, and with the option to incorporate proprietary documents or context.
Subscription Model and Platform Availability
Currently, Claude Cowork is gated behind Anthropic’s Claude Max subscription ($100–$200/month) and limited to Mac OS. This means IT decision-makers need to weigh the cost against the dramatic time savings demonstrated—especially for employees who otherwise spend hours on repetitive administrative or research tasks. Future releases are planned for additional platforms and lower subscription tiers, hinting at wider enterprise adoption on the horizon.
Evaluating Claude Cowork for Business Value
Claude Cowork is not simply another AI chatbot. Its real differentiator is in practical workflow integration: empowering knowledge workers to automate tasks across files, browser, and terminal—all from one unified, beginner-friendly interface.
Key adoption insights:
Productivity gains are immediate: Building, modifying, and analyzing documents, apps, or online activity takes minutes instead of hours.
Risk and compliance trade-offs must be managed, particularly for sensitive data and regulated environments—but folder-level permissions provide meaningful control.
Early bugs and feature gaps are present, but rapid updates and engagement from Anthropic’s product team indicate strong product momentum.
For business owners and decision-makers seeking high-leverage, practical automation without hiring dedicated developers, Claude Cowork demonstrates a clear, near-term path to higher workplace efficiency. Experimentation with this tool now may provide first-mover advantages as AI-native interfaces evolve from novelty to necessity in the modern enterprise
Topics Covered in This Episode:
- Claude Code vs. Claude Cowork Overview
- Claude Cowork's Nontechnical User Features
- Hands-On Claude Cowork Live Demo
- Claude Cowork Local File Access Explained
- Claude Cowork vs. ChatGPT Desktop Comparison
- Claude Cowork Browser and Terminal Control
- Claude Cowork Security & Privacy Considerations
- Practical Claude Cowork Use Case Examples
- Claude Cowork Mac-Only and Max Plan Access
- Claude Cowork Workflow Automation for Non-Developers
- Claude Cowork User Interface Walkthrough
- Claude Cowork Real-Time Task Management
Episode Transcript
Midroll [00:00:00]:
This is the Everyday AI Show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life.
Jordan Wilson [00:00:16]:
If you follow any of the chatter online, all of the talk so far in AI world in 2026 has centered around Claude Code. Not really sure what it is. I think maybe it was the late twenty twenty five release of Opus 4.5 combined with Claude Code, which had previously been available, that kind of created this explosion of interest and hype around Claude Code. And for maybe some of our listeners that resonated with you, and Claude Code is maybe a AI product that you use all the time from Anthropic. But I think for many of us, us nontechnical people who maybe aren't, coding all the time, Maybe it didn't hit, but out of nowhere, Anthropic did release a compliment to Claude code that I think is for the rest of us. And that is claw called Claude Cowork. And this is essentially the nontechnical version of Claude Code for everyone else. And if you're wondering, well, what the heck would I use it for? I think this is actually a preview on the future of AI interfaces.
Jordan Wilson [00:01:29]:
Because what Claude Cowork allows you to do is really everything that a human can do right now. Interfacing between your local computer and terminal, as well as being able to upload, download files, and use a browser. Kind of the things that humans do. So on today's show, we're gonna be going hands on with Claude Cowork, showing you what it is, what's the difference between that and Claude Code, and also hopefully giving you some practical use cases of how you could use it today. Alright. I hope you're excited for today's show. I am too. Welcome.
Jordan Wilson [00:02:02]:
And what's going on y'all? My name is Jordan Wilson. Welcome to Everyday AI. If you're new here, this is a daily livestream podcast and free daily newsletter helping everyday business leaders like you and me keep up with the nonstop AI developments. Just kind of cut the noise and take what's important to grow our companies and our careers. It starts here with the unedited, unscripted livestream podcast. But if you want the cheat code, that's our website, youreverydayai.com. Go sign up for our free daily newsletter because we're gonna be recapping the highlights of today's show. So if you miss something or you're like, I wanna watch the video version of this.
Jordan Wilson [00:02:35]:
Yeah. We're gonna be linking to that in our newsletter. So make sure you go check that out for that and all of the other AI news. So let's get a little hands on here in a minute. But if I I think first, we have to do a little bit of explaining. And, this is our AI at work on Wednesday series. Alright? So most Wednesdays, we go hands on, with a certain AI tool or update from usually one of the big four from either Microsoft Copilot, OpenAI's, ChatGPT, Google's Gemini, or INTROPICS Claude. So like I said, so far, in late twenty twenty five and early twenty twenty six, Claude Code has dominated, the conversation.
Jordan Wilson [00:03:17]:
And now with this very new product, very similar, Claude Cowork, I think it's worth diving into. So on today's show, we're gonna explore how Claude CoWork simplifies working with files and tasks on your computer. We're gonna demonstrate how to set up CoWork and Glib give Claude controlled access to folders and show some real examples of Claude automating everyday tasks for non developers. I mean, we'll see how that goes because let me just get this one out of the way right now. It's a very Claude Claude Cowork is an impressive, piece of software when it works. It's brand new. Right? It hasn't it's been out for, not even a week yet. And it's a little buggy if I'm being honest.
Jordan Wilson [00:03:56]:
But when it does work, it's been very impressive. And something else is very impressive, which I think is maybe why we even have to pay more attention to what's coming out of anthropic. According to the Claude team, Claude Cowork was built in a week and a half, and they use Claude. Yeah. A 100% of Claude Cowork was built with Claude Code. So, essentially, you know, the engineers were and I won't call this a a a vibe coding project, but they did say on Twitter that, yeah, they were just orchestrating, you know, Claude CoWork agents as it built Claude CoWork. Alright. So it is a little buggy for now.
Jordan Wilson [00:04:34]:
I'm sure it's gonna get a little, a little bit better. The anthropic team I've been following, they were rolling out updates over the weekend. So I do assume it is gonna get a little bit better. So FYI, if you jump in there and you're like, this thing stinks. Alright. It's hit or miss. It's been hit or miss for me. We're gonna do a live demo.
Jordan Wilson [00:04:50]:
Hopefully, it works because just ten minutes before this, I was trying it and it wasn't all working. So just keep that in mind. Alright. But let's talk about co work. And, you know, this is what Infropix says is Claude code for the rest of your work. So this is essentially a tool that allows Claude to access and work with your local folders, and that's pretty big. And there's a difference between Claude Cowork working with your local files and how you would normally work with your local files because for the most part, there's one or two ways. So even if you're working with a desk top version like of ChatGPT as an example, it can connect to some of your files.
Jordan Wilson [00:05:29]:
So you might be saying, what's the big difference? Right? It's limited. Right. So the, desktop version of ChatGPT as an example as maybe one of the closer competitors, number one, it can't access or control the browser. Number two, it can't access everything on your computer. It can access certain files that are open from certain programs, and you can upload files individually. So what Claude Cowork does is you essentially set a, a folder to work with when you start a new task. And you can essentially make it your home folder so it can access literally everything on your computer. That is the big difference.
Jordan Wilson [00:06:06]:
Right? Whereas, you know, if you're using, Google Gemini on the web, you know, ChatGPT on the web, even Claude on the web, yes, it has these connectors, but these are just your cloud services that are already synced. So this is more of you have to be thinking about personal productivity, all the things that live both on your local machine, right, but also in those other places that Claude can access as well. One other thing, I did find this out, like, halfway through, investigating. So Claude also has connectors, and it says that Claude Cowork can access them, but for whatever reason, if you are on a, a Gmail account, you're good. But if you're on a Google Workspace account, Claude Cowork cannot connect to your connectors. A little confusing because online, if you go to claw.ai, you can use their connectors to connect to your, you know, Google Drive, Gmail, Google Calendar, etcetera, but not on Claude Cowork, the desktop program. And that was I literally was reading through a lot of people's early reviews, and everyone was confused on why it didn't work. Everyone's like, wait.
Jordan Wilson [00:07:09]:
It doesn't work. Yeah. So when you connect your connectors in Claude Cowork, it's not gonna say, oh, this doesn't work. It's just gonna not work. So, hopefully, the anthropic team, can fix that. I did kind of, set a DM to a person or two to let them know that they should maybe clean that up because it does cause a lot of confusion. So, keep that in mind. Right? There's there's some, you know, some some fine print here.
Jordan Wilson [00:07:33]:
But, so it also Claude CoWork enables Claude to read, edit, and create files on your computer, and it's built for non developers to easily collaborate with Claude on task. So, yeah, this is essentially you get all the good stuff of Claude code, but with the ease of just using a more, beginner friendly interface. So this is how Anthropic kind of describes it. So I'm gonna go ahead and share my screen here. Live to your audience, let me know if you can see. But they said when they released Claude Code, they expected developers to use it for coding. And they did, but it quickly began to be using for almost everything else. And that kind of led them to build Claude Cowork, which they say is a simpler way for anyone, especially non developers, to work with Claude in the same way.
Jordan Wilson [00:08:21]:
So, essentially, this is a response from Anthropic. A lot of the early stories that kind of went viral online of how people were using cloud code was not just for coding applications. Yeah. So there's great things if you are on the software development side, you know, engineering side. You know, you can push things directly to to GitHub. Right? You can do all those cool developer things from quad code. It can spin up sub agents to break, these complex tax, complex tasks into, more manageable chunks. But the big thing here is it's bringing a graphical, interface to essentially harness or better harness the power of cloud code.
Jordan Wilson [00:09:02]:
Because when you're using cloud code, you're essentially using a a terminal, a command line interface, right, which, for nontechnical peoples, you're just staring at code jumping across the screen, and you can't really do anything else. It's not super, user friendly, we'll say, if you're not used to working in a command line terminal. Right? So this quad code is essentially it can accomplish a lot of the same or sorry. Quad co work can accomplish a lot of the same things quad co can, but the big thing is it has a more, you know, friendly interface. So that's what Claude Cowork is and, well, what's the big deal? But like I said, I think this is a preview of what's to come for future AI interfaces. And this is according to my knowledge, which I covered every single day. Right? But this is the first mainstream AI native interface that allows you to easily control what humans do right now. Right? That's your local computer and everything that it entails.
Jordan Wilson [00:09:58]:
Your terminal, which, you know, there's reasons why using a terminal is advantageous for an AI, you know, tool. Right? So even if you're a non developer, you don't have to worry about that too much. But, the ability to access anything on your local computer and use your local computer's terminal is huge. But not just that, well, you're actually, you know, using the cloud platform. So bringing in all of your, you know, your cloud memory, your cloud skills, that's huge as well. But then also the ability to, control a browser. You can control the Chrome browser via the clogged for code extension. So I'll kinda show you, how that's set up here in a minute.
Jordan Wilson [00:10:36]:
So that's like I said, this, even though it's a little buggy, I do expect it to get a lot better. But this is essentially Claude Cowork is just that. It is a true coworker. When it works, it is filling the role of what a human would normally be doing to get the most out of these AI tools. Because usually right? And this sounds bad. Right now, aside from aside from orchestrating, agents, humans are kind of like the duct tape. Right? I feel like duct tape all the time, especially if you really know what you're doing using these different AI tools. Right? Aside from orchestrating and, you know, you're the one commanding.
Jordan Wilson [00:11:12]:
But from there, it's like, okay. Well, I have to copy and paste this. Right? I have to save this to my computer. Then I have to download it. Then I have to upload it. Right? And then I have to go, you you know, use a couple things in the browser because, you know, chat gvt.com or gemini.com can't use the browser. Right? So all of these small little intermediate tasks, right, the duct tape, the connectors that us humans are playing right now, That's what Claude Co worker could do. And I do think that the other, big companies are going to follow suit, but that's why this is such a big deal.
Jordan Wilson [00:11:42]:
So who has access to Claude Co worker? Sorry. Not everyone. So right now, unfortunately, you have to be a Claude Max subscriber, which is a $100 a month. So, yeah, I kind of begrudgingly bit the bullet, I don't know, about a month ago and upgraded my, Claude subscription just because on the $20, $20 a month limit, plan. Sorry. I know I'd say about this a lot and I bellyache about it, but a quad has the absolute worst limits. For $20, it is the absolute worst value. However, I will say this.
Jordan Wilson [00:12:12]:
At the $100, yeah, it's a little more expensive, But the limits on that are absolutely bonkers. Right? Especially if you're normally paying for Infropix models on the API side, which are historically, way more expensive. Claude Max, whether you're on the $100 a month plan or the $200 month plan, the limits are actually very generous if you were comparatively just using it on the API side. So it's it's kind of a a polarizing, you know, point of view when it comes to what value Anthropic provides. But, especially if you've been using Claude on the API side, the Claude Max subscription is worth it, especially now since you get access to Claude Co Works. So right now, that's the only way that you can get it, and it is only for now available on the Mac OS app. So you have to download it. So they said in the future, this will be rolling out, to other, paid plans for, so you have to sign up on a wait list, and they said that Windows maybe eventually.
Jordan Wilson [00:13:06]:
So right now, unfortunately, to get access, you do have to be on that 100 or $200 a month Claude Max plan and on a Mac. So I've already talked about this, but let me just quickly break this down. The difference between Claude cowork and ClaudeCode. So cowork is just designed for broader, simpler non coding tasks. Both rights now though share the same foundation, but cowork just has a more approachable use. Right? It just has a pretty, you know, clawed graphical user interface. Right? The gooey, you know, it it's very easy to work with. And I've actually seen a lot of the people that, you know, I've been paying attention to over the last few months that are really, kind of been leaders in educating others on the cloud code space.
Jordan Wilson [00:13:46]:
They've even said for themselves, even though they don't necessarily need the advantages of Claude CoWorks, some of these graphical interfaces. People said that they're just using it just because it's a little easier and a little, you know, easier on the eyes, so to speak. Right? So, CoWork also adds skills for documents, presentations, and browser tasks. So here's how it works. So when you set it up, like I said, this part for me has been extremely buggy. And when I do a demo here in a minute, maybe you'll see it, maybe we won't. We'll see. You know what? I'm, I was thinking about, like, maybe I'll restart it.
Jordan Wilson [00:14:21]:
I have it open on my computer. I don't want to, because I wanna show you guys this live. And, if it doesn't work, that's okay. I have, versions of these prompts that I just did that I can show you and walk you through it. But, essentially, you grant Claude access to a specific folder on your device. There's a lot of security and privacy concerns that come along with this, especially if you are on a work computer. Right? It's a little different for me. I am the business owner.
Jordan Wilson [00:14:50]:
This is my computer. I've understood and have signed off on the data security and privacy risk, and hats off to Anthropic. They always warn you about that. So keep that in mind. There's obviously especially when you're connecting your local computer, the terminal and its ability to control the web. If you don't know what you're doing, and if you are a little loose and goosey with your data, especially if it's a company, company computer, company account, I would probably advise against this unless you really go through the proper steps of getting your team on board. Alright. So essentially, you grant access to a specific folder on your device.
Jordan Wilson [00:15:23]:
You log in with your normal cloud account. So a lot of your skills and other chats, they carry over in a different section. And then, Claude makes a plan. And then sometimes it will ask you questions, you know, kind of like on a deep research how it would, but the cool thing is is you can just click through it with the keyboard. Very nice where normally, you know, if you're on Claude code, it's all in the terminal. It's a little hard to navigate. So it's a nice little interface. So it makes a plan.
Jordan Wilson [00:15:47]:
You can look at the progress, the artifacts, the context that it's using, and then you can queue tasks and give feedback without waiting for completion. Right? So if you are giving it kind of a multi step task, right, something that involves accessing documents on your computer, something that involves, obviously, you know, writing code or using something, online that you would normally, you know, do on on GitHub or something like that, navigating multiple websites, you can send Claude, co work messages in between. You don't have to click stop. Right? And it's not gonna, you you know, pump the brakes on the current task. So that's nice. They can see messages, and you can give it feedback as it's going on, as you observe what it's doing right or maybe not right. Alright. So here's what it unlocks, and then we're gonna get into a live demo.
Jordan Wilson [00:16:29]:
Well, it just simplifies things. Right? I hate that this is, the the the example that kinda went viral with Claude Cowork, you know, organizing things on your desktop. I don't think that's, very helpful, but I'll show you what that looks like. But also creating reports and managing data easily. Then it allows you multitasking with parallel task processing like a human coworker. Right? You can, you know, collaborate on a project with Claude Cowork as a coworker. You know, dump a bunch of context, give it a bunch of tasks, whether they're non nontechnical, you know, document creation, research, you know, using your skills that you have saved in Claude, or just, you know, going in doing more technical work. And this also does empower nontechnical users to automate complex workflows safely and flexibly.
Jordan Wilson [00:17:13]:
Alright. Let's look live. To tell you the truth, I'm, like, 5% nervous about this one because I feel it's gonna break. So, we'll see how it goes. Live stream audience, do me a favor here. Alright? I'm gonna share my screen and podcast audience as always. You can go to youreverydayai.com and watch the video version. So, what I have now, hopefully, it's it's coming up here.
Jordan Wilson [00:17:40]:
Oh, I just got rid of it. But I am sharing my entire, computer screen. Alright? And the reason is is because, I'm sharing and showing a desktop application, but, if this works, it's gonna take over the browser as well. So hopefully, that won't screw anything up because I'm also using the browser, to run this little, livestream here. So let's see how it goes, what could go wrong. So we're gonna try some very simple things first, but I'm actually gonna show you, around the interface. Alright. So what you'll see on the and this is the normal Claude desktop program.
Jordan Wilson [00:18:16]:
So I've always used Claude desktop. I actually really like it. But now you'll see in the upper left hand corner, there's now three different options. It's a very clean interface. So your default is chat, and that's just gonna be like you're using Claude online. Then you have the new co work section, which is in the middle, and then you have your Claude code. Alright. So I'm just going into the co work section.
Jordan Wilson [00:18:38]:
So one thing that you'll see as I click around, it's important to know. Your Claude co work chats are going to stay in cowork. So you can't access those in the normal kind of chat view of Claude. So keep that in mind. That's important. Alright. So let me kind of show you around. So Claude cowork gives you these kind of starter prompts that you can click, you know, and and, you know, create a file, crunch data, make a prototype.
Jordan Wilson [00:19:06]:
All these are these are starter prompts, and they give you little placeholders to fill them in. Not super useful, if if you're, if if I'm being, honest. So then over here on the right hand side of the panel, you have something called progress. So, if you give, Claude Co work something that's a little complicated or multiple steps, you'll see it kind of make a plan, and you can see it go through the progress. You can watch it, you know, leave, come back, check-in, leave feedback, etcetera. Then you have your artifacts. So anything that is created. Right? So let's say you have it go do a bunch of research, maybe you're gonna have it scrape a bunch of websites, put together, an Excel sheet, put together a PowerPoint, anything that it creates, any files that it creates along the way, you can always go and download them there.
Jordan Wilson [00:19:51]:
That's gonna be in the artifacts. And then context. So if I upload anything, to share with Claude Co worker, so a bunch of files or if it accesses certain files in the folder that I give it access to, that shows up with the context. So that's on the right hand side. So essentially, left hand side, that's where you navigate between your chat, your co work, your Claude code, and your different Claude, co work, chats. So you can go back and reference them in Claude co work, but not on the web. In the middle is where you talk with Claude, co work. And on the right hand side, you have your progress, your artifacts, and your content.
Jordan Wilson [00:20:24]:
Alright. Let's try something simple. Let's see if this works, and this maybe might be an example. So I'm saying something simple. I'm using Opus 4.5. I'm saying, can you, code me a simple yet sleek Kanban style CRM that I can run locally. Right? So this is just, this is just an example, right, of what you could do. One thing I talked about last year is that I think that, you know, everyday people were gonna vibe code certain software just to get things done.
Jordan Wilson [00:20:58]:
Alright. So now it's asking me some questions. So the cool thing is, I can just hit on my keyboard. I don't have to type anything. It's asking me what contact deal information do you want to track for each card. So basic, is number one. Sales focus, number two, full CRM, number three. Alright.
Jordan Wilson [00:21:16]:
So for, for speed, I'm just gonna click two on my keyboard and then click next or enter on my keyboard. Alright. Then it says, what stages column do you want in your pipeline? I'm just gonna go with the basic one, which is lead contact proposal one loss. Alright. So I'm gonna click one on my keyboard, click enter. And then it said, how should data be stored? For now, when Claude does Claude CoWork recommend something, it says recommended, but it still, you you know, takes you through all these different options. So I love this interface. I've said for a long time, I love how OpenAI's deep research kind of ask you a couple questions.
Jordan Wilson [00:21:48]:
So similarly, Claude Cowork goes through a series of questions. And, obviously, the answer to maybe your first question might, impact the other questions that it asks you along the way. So I'm just gonna go for browser storage. So now it's gonna create this for me. And, and you might be wondering like, okay, Jordan. What are you doing here? Right? How many times as a nontechnical person have you been stuck on something? Or maybe you have to use a certain piece of software, And you have to, you know, hit, you you know, hit 20 different, you know, mouse clicks. You gotta go into 10 different interfaces all for something simple. Right? So this is something that I'm gonna probably start doing more and more is creating a simple distraction free kind of Kanban style CRM or Kanban style, you know, project management tool that's disposable.
Jordan Wilson [00:22:40]:
That's just for my task to get done for the day. Because right now, sometimes, I'll just write down, hey. Here's my five important tasks that I need to get done on a little, you know, text edit notepad, and then they just kinda sit there and die. So, you know, why not spend literally probably a minute or two? This is actually taking longer than the first one to just use quad code to build myself something because I can personalize it and customize it to my needs. And then you saw when it did ask me, the different questions. Right? I could have, you know, made an actual app with authentication, right, a login. And you can do this for the most part for free because there's free tools out there that do this. So you could even create a very simple version of a CRM that you log into that saves all that information right as long as you're using the same computer or if you wanna get a little technical, you can do it as well.
Jordan Wilson [00:23:32]:
So in the upper right hand corner, there's a progress bar. Right? So essentially, there's two different tasks. So it's already completed the first task, which was create the Kanban CRM react application. And then number two, it's testing and verifying the application. And then under artifacts, this is what it's created. So right now, it's just an HTML file. And then under context, even though it didn't need anything, I did give it access to my downloads folders. Alright.
Jordan Wilson [00:23:56]:
So let's scroll down. Alright. And it's done. Let's see how it works. So I'm gonna click open in Google Chrome. Are you still running in circles trying to figure out how to actually grow your business with AI? Maybe your company has been tinkering with large language models for a year or more, but can't really get traction to find ROI on Gen AI. Hey. This is Jordan Wilson, host of this very podcast.
Jordan Wilson [00:24:24]:
Companies like Adobe, Microsoft, and NVIDIA have partnered with us because they trust our expertise in educating the masses around generative AI to get ahead. And some of the most innovative companies in the country hire us to help with their AI strategy and to train hundreds of their employees on how to use GenAI. So whether you're looking for chat g p t training for thousands or just need help building your front end AI strategy, you can partner with us too, just like some of the biggest companies in the world do. Go to your everydayai.com/partner to get in contact with our team, or you can just click on the partner section of our website. We'll help you stop running in those AI circles and help get your team ahead and build a straight path to ROI on GenAI. Alright. And then let me go ahead and drag this onto the screen here. And livestream audience, what do you think? So I'll try to describe what this is, but it gave me a nice looking, and this works, just locally here.
Jordan Wilson [00:25:28]:
So give me a nice looking Kanban style CRM, right, with leads, contact, proposal. So I can click new deal. So I'm just gonna put test here. Right? Test test, you know, test@test.com. Phone number, there we go. Deal value, a thousand dollars. There's an expected close date. Right? So this is very much like something that you would see out of a, you know, simple, you know, simple CRN.
Jordan Wilson [00:25:57]:
Right? I can't get the date right here, but, that's okay. Let me try this again. So 2027. Alright. And I'm gonna click create deal, and there you go. It just created a card. It put it in the lead stage. I can drag things around.
Jordan Wilson [00:26:12]:
It put some placeholder things that, placeholder cards that are already there that I can go in and edit. Right? So very simple. Right? So if you're someone that is constantly, you know, creating your own little systems to do work better, so not only could I make this, you know, obviously customizable, but I just did this in one minute. Right? If you need to connect your real data, you can. Right? If you need to save this to a database, you can without knowing anything. You can just work with Claude Cowork, and it'll do it. However, I will say this is more of an example of what you would do if you were using Claude Code. I just wanted to show you, hey, how you can, you know, go vibe code something fairly easily.
Jordan Wilson [00:26:48]:
Alright. Let's do another quick task. Let's try. Alright. So I'm going to say even though I don't like this example, we'll see how it works. So I said, can you tell me the best way to organize my download folders? Then create those additional folders inside downloads, but don't move anything yet? Alright. So it is running the command. Alright.
Jordan Wilson [00:27:08]:
So it ran some commands there. It saw all of the folders. It's already done. So it said, I have a clear picture of your downloads folder. You have a lot of content here, around a 180 items. I actually have a new computer, so there's not a lot in there. So it has some installers. So this is actually I might actually do this.
Jordan Wilson [00:27:25]:
Right? Even though I knocked on it for, like, why would you, you know, use it to reorganize things on your desktop? I should probably be doing this. I'm very unorganized sometimes when it comes to my local files, and I should do better because I sync all of those for use on AI anyways. So went through and it said, hey. According to these 180 folders that you have, you know, you should create an installer's folder, a videos folder, images, documents, all these things. Oh, and it looks like it already created them even though I did tell it not to. Right? So there you go. Maybe going a little, a little off the rails here. Alright.
Jordan Wilson [00:27:56]:
So let me see. I wanna make sure I'm not showing anything too, too too private here on my computer. I'm not. So look. I mean, it did it just right there. It just created, all these new folders. Right? Everything it said. Captions, audio, data, presentation, documents, etcetera.
Jordan Wilson [00:28:15]:
Now if I wanted to, I could tell it go move all those folders in the corresponding, in the corresponding, or all those files in the corresponding folders, and it would do it. Right? So this is a task that, especially on my old computer, it probably would have taken me, honestly, ten hours because I had more than 10,000 files on there. Right? Right now, like, I actually sync my downloads folder straight to my Google Drive, so I don't know if I wanna do this yet. You know, I might wanna spend a little bit more time. But there you go. A simple, example of using the power of a large language model that has access, to your local, folder. So pretty cool. I could have also if I wanted to, I could have said, go look at all those files.
Jordan Wilson [00:28:57]:
Right? And maybe give me ideas or tell me which projects aren't done yet. Right? That's a huge unlock right there. Because if I went through and open all these files, there's probably some missing some some opportunities, some things that are undone that I should get done, etcetera. Alright. Last but not least, let's see if we can do one more. So surprisingly, alright, Claude CoWork, even though it was a little, finicky right before I got started, it's actually doing okay right now. Alright. So next thing I'm gonna do is, I'm saying go through my last 20 tweets and replies.
Jordan Wilson [00:29:27]:
Alright. Let's see what it's gonna do here. And I'm telling it to essentially go through my Twitter account, see how I tweet. Alright? Let me bring up the folder here. So it's already launched a browser. I might have to give it access to do it. Nope. It's already in there.
Jordan Wilson [00:29:46]:
Alright. So, I mean, this is pretty cool. Right? So here's what happened. I essentially sent a tweet or sorry, not a tweet. A message to Claude Cowork. I had our I had previously already given it access to, number one, access my Chrome browser, and you do have to have the Claude code extension in Chrome, and then Claude Cowork can then control your browser. Right? But you can go tell it to do things, and it can use its big smart, large language model brains. What I said is go look at everything that I tweeted, both my tweets and replies to tweets and tell kind of tell me your style.
Jordan Wilson [00:30:23]:
Right? Or if I wanted to and I did test this, by the way. Right? I said go, you know, go find my tweet style and just go go reply to someone. But what I did first is I uploaded a bunch of documents. You know, I said, hey. Here's how I feel, about AI. Here's kind of my truths. Here are some of my hot takes. Go analyze my style and go, you know, go reply to some tweets.
Jordan Wilson [00:30:46]:
Right? And, you you know, I and I said in there, go quote tweet some of my own tweets in reply to other people. And, yeah, I can't say x. How would I say reax my x's? That sounds sounds weird. Yeah. So I said, go quote tweet, some of my tweets in reply to other people's tweets. Is that confusing? It's confusing as I said it, but in my mind, it made sense. So let's see. Alright.
Jordan Wilson [00:31:09]:
It's already going through. And, again, I can see exactly, both in the Claude CoWork interface. I can see, what it's looking at. Right? So it's using my computer. First, it's navigating around, and then I can also watch live. So you will know, what it's doing because you can watch. So right now, I have this instance of Chroma. It has a nice little orange, kind of glowing box around it.
Jordan Wilson [00:31:32]:
So at that point, I know that Claude is using my browser, and there's also a notice at the top that says Claude started debugging this browser. So I know that it's working. One thing I will say, though, and I did reach out to the anthropic team, but I haven't heard anything back. I'm actually like, this is so annoying, but it's not technically an an anthropic problem. It's just this is how browsers work because I've run-in the same the same thing, you know, with Manus as an example. Right? One of those, you know, super agents can control your browser. When it goes in and out, so it might go look at the browser for a minute, and then it goes back into clogged co work, and it thinks about something. So then that little bar goes away.
Jordan Wilson [00:32:09]:
Right? So the whole display kind of keeps jumping up and down. So that's not a clogged co work thing or an anthropic thing. It's actually just how it works, when something takes over, the browser in Chrome. So it is a little annoying if you're trying to do work while Claude cowork, does it. So it's I guess, maybe it is human. It's just kind of like a little annoyance that one of your coworker has. So it's it's hard to keep doing work in Chrome. So if you do this, you might have to launch, you know, Atlas or, Comet or, you know, Edge or a different browser to do your normal, day to day work, especially if Claude Cowork is gonna be doing something, that takes a lot of time.
Jordan Wilson [00:32:50]:
Alright. So I'm not gonna wait to, you know, have you go through this because it might take a couple of minutes. I just wanted to do, some very quick demos, get hands on on this AI at work Wednesday. So here's the recap y'all. You just saw what claw work what clawed cowork is now is not the same at what it as what it represents. Let me tell you what I mean by that. This is the worst it's ever gonna be. And, actually, I was expecting a lot worse for the live demo, but you saw it.
Jordan Wilson [00:33:22]:
It actually did fairly well. Right? It created me a little piece of software that I can use. It took about three minutes. Right? It helped me with, maybe an organizational task that I needed to do do with my old computer. Right? If I would have done that with my old computer, maybe I would have just shut it down and retired it. Right? Maybe it would have been a little bit easier for me to use it, but it was just chaos. Right? So, and then I just had to go through and probably do a task that I will end up doing, or I might have it go through and, you know, pull out some insights, in my podcast. Right? Because there's certain things in the podcast when I export all my data, it doesn't give me all the data that I need.
Jordan Wilson [00:33:57]:
I have to go in and, you know, click honestly to get all the data I need for a single episode. It takes, like, fifteen minutes because you have to go into multiple interfaces, and it takes a long time. So, right, I could just give that to Claude Cowork, you know, check-in every once in a while, make sure it's still doing a good job. So, that is our quick AI at work on Wednesday, hands on with Claude coworking. Like I said, what it is now is not what it represents. This is, I think, the next interface for AI native work. Right? It is, having a large language model connect the dots with what us humans are currently doing. Because unfortunately, aside from orchestrating on the front end and checking on the back end, what us humans do? Duct tape.
Jordan Wilson [00:34:42]:
Right? So I love this move from Anthropic bringing the immense power of Claude Code in an easy to use interface in Claude Cowork, and it's worth checking out. Right? So if you're not on the Claude Max plan, or if you're on Windows, maybe if you want to go sign up for the wait list, because it is something I think, you know, I think as business leaders, we always have to think about what's next. We always have to explore what's next. We can't wait until it's commonplace. Because at that point, if you're in a competitive landscape, it's too late. So, yeah, even when things are a little buggy, you have to be scoping things out. Because if you don't, you're gonna be left behind. And if you're a listener of this show, I want you ahead of the pack, not looking, you know, looking, you know, at everyone else passing you.
Jordan Wilson [00:35:30]:
I want you to be in the lead. Alright. So to be in the lead, what you gotta do is make sure to go to youreverydayai.com. We're gonna be recapping the highlights from today's show in the free daily newsletter. So thank you for tuning in. Hope to see you back tomorrow and every day for more everyday AI. Thanks, y'all.
Midroll [00:35:46]:
And that's a wrap for today's edition of everyday AI. Thanks for joining us. If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going. For a little more AI magic, visit your everydayai.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers, and we'll see you next time.