Subscribe
Share
Share
Embed
AI First with Adam and Andy: Inspiring Business Leaders to Make AI First Moves is a dynamic podcast focused on the unprecedented potential of AI and how business leaders can harness it to transform their companies. Each episode dives into real-world examples of AI deployments, the "holy shit" moments where AI changes everything, and the steps leaders need to take to stay ahead. It’s bold, actionable, and emphasizes the exponential acceleration of AI, inspiring CEOs to make AI-first moves before they fall behind.
Andy Sack (00:00)
but I'll do three things at once. And then I'm just managing co-work as it's going through its process, thinking through, doing the work that I tell it to do. And the fact that you can do three things at once, like radically increases productivity. And basically you end up in a much more managerial role of these virtual agents. And I think that is the future of work.
This is AI First with Adam and Andy, the show that takes you straight to the front lines of AI innovation and business. I'm Andy Sack alongside my cohost, Adam Brotman. Each episode, we bring you candid conversations with business leaders, transforming their businesses with AI. No fluff, just real talk and actionable use cases.
Today's a mini episode on agents and the future of work. And I think what you're going to get is each Adam and I are going to share anecdotally how our work has changed as a result of Claude's co-work. And what we think that means for the future of work for executives at companies across the globe.
So with that, Adam, good morning, good afternoon, good evening, whichever the case may be.
Adam Brotman (01:24)
Good morning.
Andy Sack (01:24)
let's get into it. Like co-work out there, you're using it. I'm using it. It's blowing our minds. What do you have to say about claude co-work and how like it would be good for you to just share. Like what your experience has been in one specific anecdote, which you would like to highlight for the audience.
Adam Brotman (01:43)
it's so interesting working with co-work because I'm not a programmer. In other words, I'm not used to using an AI tool until I started using co-work, which I now pretty much only use, which is weird. Now, it's not totally true, but it's mostly true. I wasn't used to.
the power of what happens when you have the AI on your desktop, on your computer. I was used to sort of a web browser based either on my phone or on my desktop and doing prompt and response, sometimes uploading documents, asking the AI, whether it be ChatGPT or Gemini or Claude to do something,
There's something pretty amazing that happens when the AI is sort of sitting on your computer and has a more agentic personality, and also has complete use of my computer, including my browser. Like that step to it's not just prompt and response on a chat interface.
and sort of is on my computer, more agentic, has tool uses, et cetera, has changed the way I use it. And the anecdote would be, for example, I asked it to take a diagram I had for something that you and I are working on at work. it was kind of an architectural diagram for something we're going to ironically, using agents. But it was like,
Andy Sack (02:48)
Thank you.
Adam Brotman (03:06)
a business plan and thought. And I didn't like the diagram. It needed to be improved a little bit. I went to Claude Cowork, and I said, here's the diagram. And I want you to sort of create this new diagram for me. so what did it do differently than I would have norma? I could have done that on ChatGPT or on Claude, the normal Claude. So what was different about doing it on Cowork? Well, first of all, when I uploaded
attach the diagram and explain what I wanted. It turned around and created a series of action plans for me. was like, OK, I'm going to do this in steps, and I'm going to do this and do that. And it was talking to me while I checking things off the list of what it was going to do to update this diagram. And then when it was done, it said, here, here's the diagram. And I just clicked on it, and it put it right into my browser so I could see it beautifully on my browser.
Andy Sack (03:39)
.
Adam Brotman (03:54)
And from there, I could have easily downloaded that to my downloads folder, sent it to you, which I actually did this morning.
That's different. It felt more like I was working with a coworker, hence the name, than it felt like I was just talking to a chat bot. And that's my example. And that's a good example of how cowork has been interesting. And we can talk about what that means for the future of work. But I'd love to hear your example and what your thoughts
Andy Sack (04:18)
Yeah, so I think I used Co-Work right when it came out. I think it was January 9th. And so 2026 clearly is the year of the agents. And I did a holy shit episode.
early January 10th. I continue to have that holy shit moment with. Cowork in particular. We're super excited about its business implications for Forum three and what. We're going to be doing in 2026. I'll tell two short stories. One is I've now.
started to try to make purchases. So I actually booked a hotel using cowork. And, you know, giving cowork my credit card, it wouldn't do that. So there's guardrails on that. There are going to be agents where you can give your credit card. And that's a game changer. The other story that I'll tell is that very early on in in the use of cowork,
I created a custom GPT on ChatGPT called Cowork Prompt Engineer. and that creates prompts that I then paste into Cowork. And the prompts are much better than ones that I can make. And so I told Cowork to review
all last year's letters to investors and the board meetings of the last 12 or 15 months synthesize that down into two documents. One document was a company overview and one was a company strategy document and those two documents I put into the custom GPT that I just told you about called Claude, cowork prompt engineer and.
two comments. Claude doesn't have memory. So that brings context about who forum three is, who Adam and I are about this podcast, about what we're doing. And it's filtered in automatically into all my prompts. That's thing one. thing two is it does much more detailed prompts. Thing three that I would mention is that,
It's notion of like review the past year review the last two years and do this mechanical thing over and over and over again. the agents are able to do things that humans just aren't that good at where like we could go review all that stuff, but the synthesis of that span of work over time is amazing. So.
Adam Brotman (06:29)
Right. Yeah,
so for our audience listening to this, it might be a little confusing. So let's step back. Let's just talk about Claude, because we're talking about Claude Cowork. So obviously, to use Claude Code or Claude Cowork, you have to download this little application on your computer.
And that's different than using Claude on your web browser, on your Claude app, different than using ChatGPT on a web browser, on the ChatGPT app. You're now working off of an application you've downloaded to your computer. because it's on your computer, it sort of has like tentacles and can use your computer. And it can use your files, and it can obviously use your browser.
Andy Sack (06:59)
.
Adam Brotman (07:12)
Right away, that's giving it more power than the regular Claude web interface. And
so when we're talking about using agents and Claude Cowork is kind of a form of an agent for sure, it's got this additional power that we just described. Now, a lot of the things that we've been talking, you and I have been talking about so far on this podcast episode, you could just kind of do with
like regular old Claude, Opus 4.6 on the web interface or whatever. I have another example I want to bring up. I was working on something, and I was like, oh, there was a Google Drive that had about 15 different documents and PowerPoints and spreadsheets in it. And I could have, in theory, uploaded those all to Claude and done a prompt and then gotten a response. But it was.
actually way easier and more interesting to give direct access to Claude Cowork to that Google Drive and say, read through everything here. And then do x, y, z. Save me the step of having to figure out what to upload. And that's what's so interesting is Claude Cowork has the intelligence to even think about what I just gave it access to and say, oh, I don't need to read every single thing in this drive. I'm going to parse through it and decide.
some of this is relevant, some of it's not. And I'll just kind of sample things and that sounds like a person going in. So it saved me this step the mental, friction of having to think about what to upload and did I upload the right things and did I label them because Co work has access to my browser. It sort of did that work for me. So this is where we're talking about on the spectrum of agents from
chat bots all the way up to like open-claw agements, which we'll talk about in a second. Like, co-work is a step forward from your normal prompt and response chat bot, but it's not like a full-blown autonomous agent with memory and it wakes itself up and it's just like working when you're not asking it to do stuff. It's sort of an in-between step with computer use and really is a, and you know, and because of that, we're using it all the time.
and you're like hacking memory into it and stuff like that. In the future, people are going to be using agents that are Claude Cowork on steroids where they have, you don't have to hack memory, they're going to have memory. You don't have to like tell it what to do. It's going to start doing stuff on its own and like reaching out to you. That's the next step after Claude Cowork, which is a true autonomous agent assistant, which is what this open-claw
Andy Sack (09:28)
.
Adam Brotman (09:46)
you know, phenomenon is all about. But I'm psyched that we, you and I went into the world of co-work because it was a step into that world of
full agents.
Andy Sack (09:56)
I have two comments, one because also listening to you to just make make sure it's clear to the audience. For those of you that don't know. Cowork was developed by Anthropic. It's a and for a long time in traffic had a product called Claude code and then they used a couple of engineers there use Claude code for about 7 to 10 days and built the application. Cowork.
It was released, believe, January 9th. I could be wrong on the date. And that's the product that Adam and I are using. It has transformed the way I work. Not a little bit, not a lot, totally. And if you're listening to this episode and you've gotten this far, do not pass go immediately. Go to to entropic download, co-work and try it. It will blow your mind. You'll have a holy shit moment in a good way.
using it has really given I think both Adam and I insight into the future of work. in addition to that span of time piece, I find myself giving tasks, multiple tasks, usually three. I don't do five and I don't do one.
but I'll do three things at once. And then I'm just managing co-work as it's going through its process, thinking through, doing the work that I tell it to do. And the fact that you can do three things at once, like well, like radically increases productivity. And basically you end up in a much more managerial role of these virtual agents. And I think that is the future of work.
Adam Brotman (11:27)
I think that's so... Yeah, final comment is what you just said is such an important point to hover over and highlight. You use the word managing or orchestrating or delegating. You didn't use the word prompting or just having a conversation. So all of our experiences up until the co-work step of agents as we go into this world of future of work agents has been...
Andy Sack (11:27)
Adam, final comments from you? Bring us home.
Adam Brotman (11:53)
conversing effectively. Yeah, you might be uploading a spreadsheet. You might be asking it to do research. So there's been somewhat kind of feels like giving the chat bot, excuse me, of Claude or ChatGPT, like a task. But it always felt like a conversation. With co-work, what you just said is so important. And I agree with you. I'm the same boat. Because it's so much more powerful, because it has computer use, it's got
Andy Sack (12:11)
Thanks.
Adam Brotman (12:16)
it sort of thinks in steps and thinks like a digital employee or real assistant, you as the user feel like you're managing it as opposed to just conversing with it. And that is an abstraction up from conversation to managing or delegating that I think is the step of how everyone's going to work with agents in the future. So more on that on a future episode, but I think that was a really important point you just made.
Andy Sack (12:39)
Yeah, in addition, I'll say we probably will do another mini episode on what is all the hubbub about. Claude bought open claw, etc. So there's a next level activity that got birthed right after co-work, but for this episode, we'll bring it home. Thank you all for tuning in to AI first with Adam and Andy.
For more resources on how to become AI First, you can visit our website, form3.com, download case studies, research briefings, executive summaries, and join our email list. And of course, Adam and I invite you to connect with our AI First community, a curated hub and network for business leaders turning AI hype into action.
We truly believe you can't over invest in your AI learning. Onward.