Time to Talk About AI

Sam. Good morning, LinkedIn friends, Value First Nation. Welcome to another episode of Value First AI Daily, your collaborative intelligence report. It is Wednesday, February 11, 2026. George, Nico, how we doing today?

I mean, I was doing okay until about two minutes ago in the green room, and you guys told me that Claude Cowork came out for PC. So now I'm sitting here watching it, setting up my Claude workspace because I downloaded it real quick and I feel like you might have just wrecked of the rest of my week and potentially my weekend, but I'm doing great other than that. Chris, Nico and Chris, how are you guys doing?

I'm doing the same. Downloaded it last night on my laptop. Watching it on my desktop right now. You know, heartfelt sympathies go out to Chris and all the rest of you that haven't quite gotten it yet. I'm sure it's coming today.

So do I have to go to somewhere else?

You have to. Actually, it's really kind of lame. You have to go redownload it and then like click to install. And then at least for me, I had to like click install, then close Claude and then reopen it. And then it's just there.

Yeah, yeah. And literally I. I searched when you guys were talking to the green room, I searched Claude code, got to the website, found the download PC, hit the PC. No hate for you Mac people. I know you've been using it for weeks now. Anyway, hit the PC, downloaded it, and yes, Chris, it shut my desk cloud desktop down, did the install, then popped it back up, which now I have chat co work in code at the very top of the desktop app. And I'm watching a bar saying set up Claude's workspace. It's at about 96, 97% right now. So, yeah, that's the process to get it.

Well, I can't exactly sit here and you guys like it so far ahead.

Hit the button in the middle, download Claude.

Yeah, just to remind everybody, now available on Windows, Claude Cowork can work with local files and tools. Available on all paid plans for Windows and Mac os. This is a research preview, and Agent Safety is still in development. Stay safe. All right, download Claude. And I did also just googled Claude Cowork to get here. This is right after last night. After our talk, George started kind of refactoring my local AI agentic workflow situation brick by brick. Seems to work as a phrase.

So should I unpack that a little bit?

Yeah.

So. So last night, Chris and I. Well, yesterday, I guess it wasn't Last night. But yesterday Chris and I had a conversation and we. I've been mentioning on the show for the last, I don't know, week, maybe week and a half, Echo and building out Echo. And one of the things mindset for me during this process has been this idea of bringing brick by brick. What is the piece that I need to put in place? What is the thing that I need to optimize, to create, to whatever and insert the words for your life, is what I told Chris. I said, but instead of trying to like, build the entire thing, just build the next brick. And then when you're building the next brick, ask the assistant, hey, I want to build this next brick. How do we build it in a way that it works with what we currently have? And so over time, as Chris got to see the. How do I want to say this? I don't want to say the back end of Echo, because that sounds terrible, but see the structure and the files and folders of Echo, you know, he was able to see how this kind of brick by brick mindset to building what could potentially be an ultimate AI assistant look like.

I think this is well, well on the way to that step, right? Like the. That methodology towards. Towards like building out workflows, towards building out like, anything. I think, like, this is the. The co. Work space is such a. A nice marriage to that methodology of like, hey, let's just put build this together first. Here's the context for this. We'll put that together next. And just building on top of what's already. What you've now, like, now is foundational. It's awesome that, like, you're. The code base is now your PC, right?

Yeah.

The entirety of your PC.

Yeah, yeah. It's been interesting to have like one dedicated machine to kind of be the thing. What's interesting too, because it looks like we may have lost Chris. Hopefully we get him back. But what's interesting too, Nico, now that I've downloaded Cowork, because we kind of started with that conversation, I'm seeing a couple things that interest me in this being a layer or brick with Echo, because I literally see this thing that as soon as you look at the bottom, it's like work in a folder. And because I know that Echo has been building lots of folders with information in it now to know that I can go into a certain directory or folder and get very granular with that. The second thing that I'm very intrigued to set up probably this weekend or later this week, maybe Friday, is this idea of customize with plugins because I also know the brick by brick mentality of Echo is that we use Zapier and mcps. So, like there were no API keys because we were paying attention to security. And so now it's like we've got the things that we can do with Echo, but then we've also got the plugins that we'll be able to use with Cowork. And so I'm, I'm sitting here talking to you this morning, just downloading this, going, man, what is the symbiotic? Totally said that. Kind of weird, but whatever. You know what word I'm trying to say, people, what's the relationship between Cowork and Echo, and plugins and MCPs and folders and directories? This could get quite interesting quite quickly, to be honest with you.

Yeah, I wondered that my. I think we were talking about that. I don't know if it was the green room yesterday or not, about how we each use our own personal AIs. And you have Echo, I've got Aries. I think Chris and Ryan have been working on V for Value first. And if you guys don't know what we're talking about, we've all kind of taken the time to put together our own personal AI agents that are ground up, personalized. They're not. It's not like, like we're obviously not creating our own foundational models or anything like that. But the Persona, I don't, I don't know how else. How to best put it. I've actually been talking to Aries about it all week at any rate. It's. It is that, that. I don't know. I'm having that debate with, with her all the time about like, okay, well, what is different between you and this new capability that came out? Or is there something that needs to get added that you're not capable of now that this thing is. But now there's this. There's like a. To your point, there's a symbiotic relationship now that's formed between the two because of the fact that Aries has computer use and has vision so she can go and use for desktop.

Right.

So she can actually leverage like any of the additional tools. So now it's this very meta. Yeah, that's just a really crazy meta working instance where, you know you're using your assistant who's using the software tools. Right?

Yeah.

And yeah, I mean, at some point it may actually, I'm sure it is now on a really, really, really minuscule scale. But one of the, one of the new terms, it's not, it's going to get popular quickly, but I'm not sure how long it's going to last because it's definitely going to get seen as derogatory very, very fast. Meat puppet is going to become a term, like pretty quickly because what I'm starting to see now is this full metacycle that goes from human to agent to software, back to human, only the other human is the end, like is the worker at the end of the, of the tree. Right. So in other words, agents are starting to hire humans to do tasks like real world tasks that the agents are being asked to do by other humans.

Yeah, you brought this up. You brought this up yesterday. So, like, is this a scenario of like. Because I'm trying to wrap my brain around how, what this looks like. It's like a human goes out to their assistant, says, I want to get this. Chris isn't here, so I can't make him mad, but let's just go with this route. Human goes out, says to its assistant, I want to build this fabulous PowerPoint assistant instead of creating the PowerPoint. Maybe this is a bad example, but assistant goes to Fiverr and hires a human to create the PowerPoint. Like, is it that simple or. What are you talking about when you say this?

Like, so it could be that simple. It's somewhat dependent upon like agent skill set as well, but because it's more extended to like, let's say. And again, these are like pretty simple scenarios, but they make perfect sense to me. Let's say it's the case where, you know, you left, you went on vacation and you, you don't, you don't really know, but you're pretty sure that maybe you left your, your door unlocks left, you know, something. Oh yeah, left something behind, whatever it is, left the stove on. Or it's a case where like, you need something either delivered to you or delivered to somebody else and you don't want to necessarily have to coordinate all the effort. That's where, you know, normally you would have to pick up the phone and call somebody and have to coordinate a service or something like that. And this is now the case where because agents can pick up the phone and call somebody, you can actually just ask an agent, like, hey, you know, look, I'm in a bind. I need you to handle, you know, making reservations for me at the same time I left something at the hotel and I need that to get brought over to the airport. You know, can you make sure that that happens? Whatever, blah Blah blah, move on with your day. And the agent is able to then pick up the phone, call the hotel, talk to the human, make the arrangements with the human. So the human goes and does what they need to to take the bag to the airport.

Yeah, yeah, it's interesting.

And, and there's definitely other case I heard of where someone needed repairs for their like home repair or whatever because like pipe had burst or something in their home and they weren't available, weren't around for it. And they're able to make use of a agentix service, handyman service that basically monitors their home and if something goes wrong, goes and looks up, finds the most reputable, nearest, best person to handle the job, contacts them and automatically sends them out, like has a conversation with them and everything. So it's, it's this really odd case where now and it's so it doesn't, it's not quite as it may have seemed on the surface, but essentially on the end user side, imagine that you were a, you know, essentially manual worker of some type. Doesn't even have to be manual labor. Could be whatever at a copy store or something. And you pick up the phone and you think you're talking to a person, but you're actually talking to an agent. Yeah, that's basically what's happening.

It's interesting because my mind goes in maybe what is a weird direction for this? And it's like for years I've been fascinated with these people who have really turned their homes into like smart homes. Like the lights are smart, the. And so like listening to you, I'm like, wow, maybe this exists and I'm just, you know, not wealthy enough or I just haven't researched it until it entered my brain this morning. But like who, who's going to be the person or who is the person that has the house that is so automated that the, you know, the flange in your toilet, all of a sudden that starts to go awry and occasionally you have to like hire a plumber. But the house is so optimized so you know, Internet of things, if you will, that it realizes there's like a 27% inefficiency in the flange of your toilet. And so all of a sudden the system just calls a plumber and a plumber then like shows up at your door to basically fix the flange. Because your house assistant, it's, it's directive if you will, is to make sure that the house stays optimized. Now that gets really kind of freaky but cool at the same Time to me to think of it in that manner.

You guys remember the days where it was best practice to make sure everything else was off as you installed new software and how far we've come where it's like, oh, it's fine. We can be on live stream and download stuff. Yeah, I got the invisible crash report that I could send along.

Oh, geez.

But yeah, we're back and I really want to click on this button so I'm going to share my screen again.

Yeah.

Hope I didn't miss too much.

Nah.

Unboxing software is like, yeah, yeah, yeah.

So we've got Cowork at the top here now and I'm gonna hide that. Click this button. Let's knock something off your list.

Yep. So two things that I'll tell you, Chris, that I mentioned that I thought were very interesting because it was the idea of Echo Cowork having this symbiotic relationship. The fact that I see a work in folder option and I could. I could select a folder that historically had been a folder of Echo and then work very granularly in that folder. So for me, I started to think about things like all the Hub Heroes transcripts or beyond. Your default transcripts now are in Echo. But then also Cowork could be doing work around or with said context or information. The second thing that I brought up was that customize with plugins that shows up in that box as well because there are things in there that are very interesting and just like legal and all sorts of different things. Yeah, you might need a reboot because things aren't showing up on your computer. Yeah, it's not showing up as. And maybe it's just shut that down and redo it because it's not showing up in the way that it's showing up on mine. But like, if I go to customize with plugins, it's like customer support, plugin, data support, enterprise search, legal marketing, product management. And so like one, one, when we usually think about plugins, we think about like integrations almost. Right. Like apps like hey ClickUp and HubSpot and. And this is like something totally different than that. So there's probably that somewhere. But yeah, again, probably if you do like marketplace from GitHub or marketplace by URL or upload a plugin. Like there's just a bunch of different things that I'm interested to see how the two start to work together or what I can do that I couldn't have done previously before go to work so much.

Yeah, like this, this falls into place with such perfect timing because I was just talking to Chris about wanting to revamp, go to market and how to revamp it for everyone else in a way where you're not, you're just buying the education to do it. You're not buying the service to have to hang over your head for the rest of your life in order to have it done. And this kind of just like kicked open the door to that because yeah, if, if you, if ever there was a reason to start working with Claude, now would, now would be the time. Cowork offers, especially for people that do GTM offers data plugins that allow you to do all of the seriously awesome, nifty stuff that I do all the time with just a simple plugin request. And you can do it against your data, you can do it with deep research, you can do it in pretty much any way that you can possibly think of slicing the data with these plugins. And we're talking about validation, exploration, content extraction, visualization, SQL queries, even doing stat analysis, creating visuals, doing and doing, what do you call it, dashboard building. So any, like, yeah, I don't, I don't know how else to put it, but like a lot of b, like a lot of what was in BI and a lot of what was go to market and you know, sales and marketing, it's all like coming together in one. That whole thing of, of what Sam Altman was saying a few years ago when he said that, you know, it's not going to be too long until there's an AI that can do what an entire, you know, marketing agency can do. And yeah, this is definitely a huge step in that direction. Just being able to, I don't know, being able to do all of the very high end, complex things, high level things that you would need several professionals in most cases to help you do. Now you can just do it in like the click of a button. Right?

So let me ask you about the gap on that though because, like, you can do it in the click of a button or like I was talking to Chris last night or yesterday too about like, yes, certain types of humans can do. Like, is there still the education gap for like, I'm a leader. I've historically not paid attention to xyz. I go to my assistant and I hit the button and I don't know if I got a gem or if I got a turd or if I gave it the right, like talk me through the, the gap though, because what you just said made it sound like anybody can go do it.

So that to me, that is the gap, right? Like, the only gap is the awareness. It's not even so much like the education, it's just the awareness. So, like, even, even people who might be interested in wanting to get into it and wanting to get into, like, go to market and let's say, like, okay, well, is the research that I'm doing the right type of research? You know, Chris and I talk about this all the time. It's like, okay, well, who are you asking? Like, aren't you. Why don't you just ask the same person that you're working with already who will. Who will then work with you to find out whether or not what you're doing is the right thing to do, right? Like, you have. That's. I tell people all the time, lean into it. Like, just lean into it hard, right? Like, the more you do, the easier it gets because you just fall into this perpetual thing of like, oh, man. Well, I'm not sure if what I'm doing is the right. Oh, well, okay, let's. Let's work that out. Let's work out whether or not the plan I'm doing is the right thing in the first place, right? I. It's. It's funny because I, I usually say take. Take it one step further with like, whatever idea or concept you're having. But to your point, when it comes with. To working, it's almost like, take one step back. So it's like, yeah, I want to do this thing. And it's like, okay, take one step back from that. What's. What does that look like? Oh, that looks like, how do I go about wanting to do this thing? Right? And then like, one step back from that is like, how do I go about figuring out how I should go about doing this thing? Right? So, like, you can definitely put all of that. That how do I do this into the model to be able to get the. The answer and the output that you need to get started with it, right? So even that. That's why I say, like, the educational layer is baked in. Like, the, the what you need to know to do it is there. You just have to take away the assumptions that, like, that that's not possible or that it can't explain it to you or that, you know, you. I, I really do. I think so many people have this block of like, well, it couldn't just be as easy as me and this machine.

Yeah, well, and I think it is kind of. Yeah. What I hear you saying. And, and it's funny because as you're talking about this, I'm like, yeah, of course. And I'm reflecting back to like my journey from non terminal user to terminal user. Right. And it was the power of asking smart questions, the power of not allowing myself to get frustrated or dare I say overly frustrated when trying to get the answer to the question and being able to ask the right question based on the journey that we're going through. So I think there's this very interesting. Is it possible? Yes. What is the gap? You asking the right questions with your assistant to get to the place that you're trying to arrive to?

Yeah, I think the questions are so important. Like critical thinking is usually the gap. And I think the issue with like you can get trapped by AI just as much as you can get trapped by the agencies that you don't want to work with anymore or you've lost trust in.

And.

I'll share the AI native page that talks all about that. But like if, if you were to ask HubSpot or if you were to talk to AI about using HubSpot, if, if you don't give it any other information, it'll drive you in the marketing and sales direction every time. And if you were to ask questions like, well how do I get like can I use my product data or can I do bi or like how do I fit that into the system that I'm working in? It's going to suggest other tools for that because the training data says it's not like HubSpot is a marketing and sales CRM. Same goes for how long should something take? Right. If you don't direct, like if you come in right now and you're like oh I, I hear you know, people building websites with AI like let me go talk to AI and then ask it about this website and how long it will take. It will give you a human based time. Yeah, it'll lie right. About how long it takes. Which you can salt. You can get on the right track with just one. Like if that doesn't. It's basically like if it doesn't feel right to you, like talk to it. Right. Talk to AI. Like can you have a conversation where this is why, like whether it's like if you can pick collaborative versus I'm going to use it as a tool and delegate and I just wanted to get something done or it knows everything and I don't know anything so I'm just going to give it over and it's just going to wow me. Like that's the balance that I think, you know, you have to give yourself permission to ask for help, but then know that you know something. Like you've learned some things as a human. That is valuable experience. Right. And a lot of, you know, especially before I started doing stuff on LinkedIn, it can be hard to respect that. No matter what your experience has been, it's different than everybody else is. And it can bring something to the table. Right. And as long as you bring that to the table with AI, you can get to any, any end result, in my opinion right now. And that's where a lot of, you know, we were talking yesterday, George, about website stuff and just the platform. Even if you're a marketer and you've just been buried in Leeds land. Like our human experiences of working with the Amazons of the world. And like all of our personal experience, we know we have an inherent knowledge of how software is supposed to work. Right. And just because you're not on that team at your job doesn't mean you don't know what an outcome should look like. Right?

Yeah.

And if you can just ask the question of like, no, yeah, that looks cool. But here's the screenshot of the site that I like. Right? Can we do that? Right. And as long as I think people can enable themselves into the conversation of. Again, we're back to learning like there's no. It can be an easy button and it's wonderful when it one shots and does exactly what you wanted it to do. But it doesn't happen as often as you think. But what does happen is when you can ask that second question where it couldn't have possibly read your mind in that first prompt. Right. That the way it can give you the result on the second one. Like Nico says go one step further. Like, that's when that the magic starts to happen, in my opinion. Yeah, we're. We're over time here. I gotta wrap up. Sorry for the. The cutout. Thank you to George and Nico for, for carrying on. And we'll be back tomorrow, hopefully with a working cloud cowork instance.

There you go.

That we can, we can show off. Happy Wednesday, fellas. We'll see you tomorrow.

See ya.

See you guys.