Aydin Mirzaee Fellow.app - AI Agents That 10x Your Productivity

Speaker A: We are getting to a place. I think one example of this was ChatGPT's pulse, which was effectively, use ChatGPT, use it the way you do, and then overnight it will say, oh, uh, Aiden was asking a lot about these sorts of things and let me go out and do some research and then in the morning just give him some things that are relevant based on the things that he's interested in. Like, I think this is the place that we're going to go where AI will become proactive. And so that is going to be the next stage.

Speaker B: Hi, my name is Dimitri Panici and I'm a content creator, agency owner and AI enthusiast. You're listening to the AI Agents Podcast, brought to you by Jotform and featuring our very own CEO and founder, Idekin Tank. This is the show where artificial intelligence meets innovation, productivity, and the tools shaping the future of work. Enjoy the show. Hello. Welcome back to another episode of the AI uh Agents Podcast. In this episode we have Aidan Mirzai, the CEO and founder of fellow AI. How are you doing today, Aidan?

Speaker A: Great, great. Thanks for having me on. Dimitri.

Speaker B: Uh, I love what you're doing. I love the whole space that you're in. AI note taking is one of my passions. Uh, all the different tools that are out there. I was. So for those of you that, uh, are unaware outside of this podcast, my YouTube channel that got me into all this content stuff for not only AI companies, but prior to that, more SaaS companies and productivity apps is kind of what got me interested in the whole world of tech and AI in general. So really excited to talk with someone who's more in like my realm of this thing make you more productive in the pre. I know it obviously is going to be super AI powered and all that good stuff, but, uh, this is more my wheelhouse. So I'm really excited to talk with you. Um, first and foremost, tell us a little bit about how you got into the space of, uh, note taking. I know you have your whole journey probably, uh, you could tell us, but if you tell us in a few minutes. How did you go from where you started to implementing all this cool AI stuff? Um, as a meeting company?

Speaker A: Yeah, yeah, great question. So, um, yeah, to shorten it, I engineer by background. I, um, but, but always an entrepreneur. So I think, uh, I started three companies now with, with my co founders, five different products. Our last company was in the online, um, online survey space. We had this product called Fluid Surveys. We sold it to SurveyMonkey. Was, uh, at SurveyMonkey for a few years leading up to the IPO. Uh, and then my co founders and I left to start Fellow. And so the original idea behind Fellow was we wanted to build a tool for, for managers. But as we started building it, we realized that if we want to help managers, we've got to tackle meetings because they spend all their time in meetings. And so that would be the highest surface area. So, but, and then we started getting into meetings. So the first part of Fellow's journey was all around how do we like, build this tool that when you implement it in your company, it's going to help, uh, your meetings become effective and help you spend less time in meetings. And as Covid, you know, happened and more, more companies started becoming remote, that was a big thing. Like meetings started to explode. There was so many of them. And so that was the original vision. And then of course, the ChatGPT moment happened, end of 2022. And we saw that and we thought, oh my God, uh, meetings are going to be completely revolutionized. Uh, and one of the best applications of AI is going to be in meetings. And I think every industry is being transformed, but meetings and that type of work specifically was one of the first areas. And so immediately we kind of thought, how could we reinvent our product? How could we reinvent what we're doing to deliver the same value to our users and customers, but using AI? And it turned out that we had to do a lot. So we reinvented, uh, the product and the company. And it's, uh, it's been amazing over the last two years since we pivoted the company. We went uh, from uh, zero to eight figures in AI revenue. Um, so, and you know, that's a big transformation for a company that had already raised, you know, 30 million in capital, was already doing millions of dollars in sales, um, was already doing all that stuff. It's kind of like jumping out the plane and building parachute on the way down. Um, but yeah, now the company is in the, the best place it's ever been, fastest it's ever grown. And yeah, there, there is a, it is a more mature market, but it's also changing very rapidly. And uh, and we talk about like, where, where we see us going and like what niche we're going to dominate. I think everybody's heard of AI note takers. Um, no. Yeah, so it's uh, Yes, I think most people know what it is, but you know, the question is, how are we winning? Even though.

Speaker B: Yeah, no, that is the question. Yeah, that just to kind of say from the get go, really cool story. You managed to change your product so quickly, make eight figures in AI revenue that quickly. I have a question about your product. Right. Because I think at the outset, every single person who has been in Google Meet maybe a little bit less on Microsoft Teams, um, and as well on Zoom. They have this as well. A lot of people experience, the thing pops in, it has the name. Tell us the experience of what it's like to have Fellow AI inside of the call. From the, uh, people on the call perspective that aren't using it, and then from the perspective that they are using the product. Like, there's a couple of those types of tools. So just want to lay that out for the answer.

Speaker A: Yeah, definitely. Um, yeah, we can definitely talk about that. And you know, we can talk a lot about, like, the philosophical, uh, things and how this transforms companies, but also some very futuristic things of what the world's going to look like, uh, with AI being more and more infused in conversations. But yeah, so Fellow is like our philosophy behind the product is we want it to be effortless. And so there are, um, depending on what your use case is, Fellow will just behave the way it should. So there's two modalities. One modality is you have a meeting and Fellow shows up as an attendee. Like, it's another, um, tile on your screen. Like, you and I have a tile during this podcast. And so Fellow would be a third one in this case. Um, and that is useful for a bunch of reasons, but we also have a modality which is what, uh, we call a bot. Bot free modality. So in other words, there is no tile and Fellow is there in the background. It is taking notes and doing all the magic that it does. But there is also no bot on the screen. Um, so each one of these, there's reasons you would want each one. So for example, for the, ah, when would you want Fellow to join as a participant? Well, Demetri, if you have meetings happening at your company and you're not going to go to them, but you want your bot to attend on your behalf, your agent to attend on your behalf, that's, that's a reason why you would want the agent to show up. But if you're going to be there and um, you know, your computer is there and accessible, then of course you don't need to have fellowship as an attendee. It can be like a background process, uh, that runs and captures the notes and everything else. Um, but our philosophy has been different. So the nice thing about Fellow is it has all the modalities It'll capture, um, you know, both the ones we talked about. It'll do in person meetings, it'll do slack huddles, phone calls, it'll do all the modalities. Uh, but our approach is different. So we don't believe in this idea that someone should be recording you without you knowing.

Speaker B: Okay.

Speaker A: To me that is spyware, that surveillance. Like I was in my company and like somebody was just recording, they had

Speaker B: their voice memo out.

Speaker A: Yeah, that's just like, to me that doesn't build any, any trust. So even in the, um, the version of Fellow that is bought free and you don't see a tile on the screen, the way it works inside of a company is both parties can still see that the call is being recorded. It's just, it doesn't have the annoying interface of, you know, an extra participant in the room.

Speaker B: Yeah. It didn't say like Fellow AI and have your logo.

Speaker A: Yeah. So it's basically so we, you know, so that's the way that we approach it. It's the, we think that any tool that doesn't do it, and we're the only ones that if you are doing like a bot free version of this, um, yeah. And like both parties are aware, uh, we're the only ones that do that. And to me anything other than that is surveillance. But the other thing that we do which is, which is very valuable is we're also a collaborative platform from the get go. Because to me the hardest thing, you know, a lot of people will say, and I agree, which if you walk into a meeting, the most important thing that you want out of the meeting is the action items that come from it. Right. And even humans, you throw five humans in a room and you tell them and then ask them afterwards what were the action items. They might think different things. So, uh, the other philosophy that we have is that fellow, uh, creates this collaborative environment so everybody's on the same page with what the action items are. So it's not like you have your note taker and I have my note taker. There is one fellow for the meeting, for the company and it will tell you, all of you together, what the action items, um, are. So everybody's working off the same list. So these are some of the philosophies just on a user level. And we can talk about the other thing, which is the most important thing for us, which is privacy and security, um, which is kind of like a core philosophy and why organizations that care about this stuff choose Fellow. Uh, but from a user perspective, that's some of the differentiating stuff.

Speaker B: What do you, um, think is what people most find good about the. There's two aspects to this during the call and then there's post call. I know you said the action items and stuff like that. Is there stuff during the call that you're getting real time feedback of or, um, um, is there like a transcript going? Because there is a tool I use, won't name competitors on the call. It's one of the ones where you can. It kind of just basically takes the transcript from captions, uh. Right. Um, in Google. So it's a nice chrome extension. But I really like how it has the ability to show the running thing of what everyone's been saying. And is there anything like that in there? Or how does the in call and post call kind of look?

Speaker A: Yeah, yeah, great question. So, uh, it does a lot of really cool things. So let's start from before the call. Um, as you know, like, in the most ideal world, what everybody would do is they would look at their calendar and they would think, this is what's going to happen today. Let me prep so that in this time that is very expensive time because it's synchronous time. Synchronous time is the most expensive time. Right. That we're both prime time, like live attention that there's so many things we could discuss that of all the things that we could discuss that we are talking about the most important things and also in the right order. Like, ideally I would spend, uh, a bunch of time planning and then I would take that plan and then we, we would do the meeting. Right. But the problem is that nobody actually does this or doesn't do it consistently. And you know, this is a thing that AI can really help with. So one of the things that fellow starts with is the surface area that says here's what, like, here's what you should probably discuss during this meeting based on stuff that you've discussed in the past with that person, but also stuff that has occurred with your other meetings. Right. Um, you know, you, your VP of marketing. Like, this is what you guys talked about in your discussion. And it turns out that that could be related to your VP of sales that you're about to talk to. Um, now it'll tell me I make the decision. It's not going to, you know, show information, uh, to people who shouldn't see it, but, but this is the kind of preparation that fellow will help you do. And then you actually have like, the agenda. And so you can follow the agenda during the Meeting it's a collaborative surface. Um, a lot of people, you know pre fellow would use maybe like a Google Doc as a collaborative agenda. Uh, so fellow has that baked into. Except it's just powered by AI and will prep you. So now again your time is very valuable. So now you know that you had a PhD in every topic. Um, be able to help you figure out the most important things that you should discuss in the right order. So now you have that. So now and now you're just walking into the meeting. Now during the meeting there's a bunch of real time stuff so you have the fellow agent present you, you've got like a live chat, just like a chatgpt or Claude Nice. And then you can, you know, it'll give you suggestions, it'll say hey, like it knows that we're, you know, you're making a point and then there's all these shortcuts. So one of them for example is Debate Assistant. So I click on Debate Assistant and I have like the most powerful model that exists that's going to tell me all the counterpoints like the attack vectors in that's ridiculous that we're having or like what are the smartest follow up questions that I could possibly ask. Um, right. So it just, it's going to help you be on top of your game in a way that only the world's, I don't know most powerful people that have chiefs of staff, uh, would be able to do. Um, and now, now we're talking about say after the meeting, right? So from after the meeting there's all these actions that you might want to do. You might want to do a follow up email, you might want to update your CRM, you might want to um, create a knowledge base article from this discussion so that other people could benefit from it. You might want to create a clip of the conversation. Um, and then send that in Slack. So all of these actions are now also available in this like beautiful agentic format where you just say do this, do that, click a button and then all those things start to happen. Um, it really starts to become, I mean the vision for fellow over the long term is to be this AI Chief of Staff. Um, but just an always on chief of staff that's going to help you be on top of your game. And yeah, so that, that's kind of the what, what we've put together. So it, it's really like once you learn it it is more powerful than anything uh, else I uh, think in this category.

Speaker B: Well tell me a little bit more about, uh, how you even got from point? I don't want to say zero, but how did you. And how do you continue to figure out, hey, this person that's our ideal customer, would like that feature, would like this feature. And I have to just say I think that the debate example you just gave, really good. There's not a lot of them out there. I know Zoom might do a decent job of you have a chat with it as it's going, but I don't think a lot of the tools in your space are actively doing that. I think they're mainly post hoc or post, uh, call. I think you can ask questions of it afterwards. But the in real time thing is actually very good. So really cool, um, thing there. Uh, how did you, though, figure out this is what people would find to be awesome versus, you know, maybe there's some features you did, you could have gone after, but you didn't.

Speaker A: Yeah, you know, um, there's a bunch of thoughts I have, um, with this, and we're in a very special time. As, you know, you're talking to a lot of, you know, people building products, like, very innovative stuff. And so we really are in this moment where there's only so much that you can get from, you know, your customers telling you, or your users telling you what you should build. You should always talk to your customers. This is like one of the fundamental principles of building any company. But a lot of times you have to understand that the realm of things that your customers might ask are going to be grounded by the things that they know are possible. Right. And so the challenge is now we have this thing called AI that's advancing very rapidly. Um, and so the number of things that are possible is just increasing like crazy. Uh, so the set of possibility is so much larger. So the first thing that I would, um, would say is, like, you have to be careful almost not to only do the things that, and the features that come from your customers. Like, I think a lot of us were in this space where it was like, vote on features, vote on this or that, and then just like, do the top things.

Speaker B: The change log was made, the roadmap was, you know, pushed by the voting, uh, board. Yeah.

Speaker A: And, and like, you definitely have to listen. But, uh, but now we're kind of in this stage where, like, it's like the rule book has changed. So it really requires, you know, all of us to take a step back and imagine, Imagine, um, what the future could look like, the compelling future, and then think about the things that you would need to, to take that there. So, one, I, you know, I, I learned this and, you know, inspired by Brian Chesky, uh, from Airbnb. I don't know if you, you heard the story where he talks about the 10 star, the 10 star, um, experience, but we all know hotels are rated by like one star through five stars, right? Five star being a great hotel. And I think some, some places are six stars, maybe a few of them in the world. And so he basically came in and said, like, what would be like a four star experience, uh, in an Airbnb booking? And he kind of imagined that. He's like, well, you kind of go, the experience was really good, but it was kind of hard to check in. Maybe the owner wasn't quite there to give you the keys or somehow the code didn't work, but you got in and you had a good experience. And then he goes about the five star experience and then he does like the six star experience where like, you go in and someone already made you food and they all, they planned your, um, entire trip and they gave you these really good experiences. And he kind of goes there all the way up until like a ten star experience. And the ten star experience is crazy. Like, I don't know, a celebrity comes and picks you up in a limo. And it's not necessarily about building those experiences, but when you go through that kind of a creative process, all of a sudden you kind of see what the most amazing thing might look like. And then you kind of build back to, how can I take us there? So for us, the way we thought about it is we went through and we said, like, what does a five star meeting look like? What does a six star meeting look like? And we started putting all those things together and we realized to get to say, a 10 star meeting, I mean, that is really, really hard to do. Uh, because nobody has enough time to do all the prep. Like, you'd have to prep for an entire year to get to that 10 star experience. And so we thought with AI, can we make it so that everybody can have a 10 star experience all the time? And so a lot of, a lot of it comes from there. But the other part of it, and the second part, which is a little bit more on the visionary side, is you start to think about, like, what are meetings to begin with? They're these constructs that help companies operate, right? They're a tool at the end of the day. But when you think about companies in a world of AI that's rapidly getting better, you start to even question you Know what are meetings for? What does it mean if a meeting had multiple agents in it? Is there this concept of agents meeting with each other? What does that look like? And so part of it is also imagining what future companies actually look like. Um, and then start starting to bring elements of that back to everyday reality.

Speaker B: Yeah, I think it's a good approach. And you kept saying six star for a second. I was like, are you talking about six out of five? Um, yes, six out of five, yeah. Ah, it was the first time I'd ever heard that and I was like, you know what? True. There are things that somehow quite do go above and beyond, uh, I guess the five star limit in such a way. Everyone's had that experience. Like I don't even, like I can, I can say this is better than perfect in my mind.

Speaker A: Yeah, I actually, you know, interestingly enough I actually think that there are now again, I don't know exactly how the real, I mean this is a, this is a framework more for first creative exercise.

Speaker B: Ah.

Speaker A: But just, just for fun, there are actual six star or seven star hotels in the world. Um, I think the, there's one in Dubai for example. Uh, so I don't know if that's an official rating or they just call themselves six star experiences but apparently I

Speaker B: feel like anything in Dubai they're on their own set of rules to be fair. Um, so fair enough. Okay, well no, no, I think that's a really good uh, mental framework to have because I mean ultimately it is about a lot of things with software and AI at this point it is about the experience, um, more than anything. And I think that's something that's definitely been um, by some tools that have adopted it. You know the world of Claude code was very good for technical people like me for a while, um, but then they introduced Claude cowork and the average person on my team is able to take my level of output that I could do and now do it in the front end in a way that's a little bit more accessible. Um, and you continue to see that trend, uh, of we just need to make it the best experience possible and people will come. Uh, what about on the integration side? Um, want to ask a little bit about that? Obviously you have your own ecosystem. You said you want to make it a orchestrator for the entire team. I think that's actually a really good idea. I think especially if your tool has great integrations with some of the communication platforms that exist. It could in theory with like a slack, uh, bot config or Whatever. Um, if you have your meetings, you have your Gmail and you have your Slack, for most people, uh, there's no other communication so you could orchestrate anything. So curious to hear about the integration side, what that looks like. And um, I, uh, uh, what are some of the main types of companies that you're finding using some of these features and uh, becoming to your product in general?

Speaker A: Yeah, um, um, yeah. So, so on, on the integration side, we, our philosophy is that the, the first Surface, there's a few, few different Surface areas. But you know, surface area number one is the action items that come from meetings. So everybody in a company has a different way that they want to handle theirs, right? Some people will do them inside a fellow, some people will use Asana or Monday or Google Tasks or Microsoft Tasks or. There's so many different task management systems and it's really, really hard to standardize on this because everybody's going to, on an individual basis, really want to system, right? And so you basically have to integrate with them all. Uh, and so what we try to do is make sure that we can push these action items from fellow into all these systems, uh, regardless of what it is. And by the way, this is why getting action items right is so, so important because I, uh, I'll tell you exactly what will happen. The easy one is if you ever miss an action item so fellow didn't capture an action item, it would be over. Well if, if that happened just one time, you would lose all trust right? Now if we did the safe thing and we just said, oh yeah, anything that could be an action, and we're going to track it as an action item, right? M. Many, uh, like if you use some, some products in this category, they'll just like produce a bunch of action items and then what will happen is, and especially if these things get pushed to your personal system now, you're going to be flooded with all these action items. And, and we all know when there's too many things, nobody's going to do any of them. And so, so it's really, really hard. So if someone in the meeting says something like, yeah, we should like, like we should do that. Yeah, I think I'll, I'll do that. Like, is that an action item? It, it depends. It depends on many, many things. And so even again, humans in a room will sometimes debate whether something was an action item or not. And so it's actually really, really hard to do this. And we have this um, you know, know, very interesting approach where we get like a bunch of Agents to argue with each other, to kind of figure out like, what the actual action items are. But it's just like this sort of craft that really makes, um, all the difference in, in being able to do this well. Ah, so, yeah, a lot of action items are like, uh, a lot of integrations are like that. Um, some of them are updating CRM fields and, you know, being very smart about that. So, for example, everybody knows updating a CRM is the most annoying thing ever.

Speaker B: It is the worst.

Speaker A: No one likes it and nobody does it well. And so, but sometimes you have these strange fields in the CRM. So maybe you have a field in the CRM which is something like, what is the potential of this customer over the next two years? That's a hard question to answer. Like, you have to sit back, think about it. But it turns out that again, the way our integration works is you type your prompt and you map that to a field and you, you know, and then we're going to run that prompt on the, on the discussion, and then we're going to put that field, you know, the answer into the CRM. So we give you, like, that level of flexibility. So you can be very prescriptive about what specifically goes into the different CRM fields. Um, and then, you know, you can translate the same thing to wikis or, uh, whatever other system you use. So that's kind of been our approach, uh, overall.

Speaker B: Yeah, I think, you know, I'm trying to parse through my thoughts towards the end there. I think there is a couple of different, uh, aspects of, uh, what you guys are doing that it's kind of hard to pit down. Like you said, like, I'm a notion guy. Um, I've been recently disillusioned with its bulkiness for my taste. But I'm a notion guy. Love using it for tasks. I've reviewed literally every task platform 20 times over on my YouTube channel. So I know exactly how everyone has their own thing. I even use Microsoft tasks for like six months when I started, like, in the productivity space, which makes me cringe in retrospect. But people work, uh, at companies that have Microsoft, so it's all good. It's really interesting though, you bring up the CRM management. Cause I find that that makes a lot of sense for a meeting person. Um, you have all the context if you are in the meeting. Right? You do. It's like, is this person gonna, is this person gonna move to the next stage? I feel like I made my own, uh, little agents with other tools to, based on the context that I give about my business and how conversations can go, it'll move it from one stage to the other in my notion CRM. Um, but um, I do think having that naturally implemented is a really important thing because like we were saying earlier, meetings, slack, email, it's kind of the only components of communication that you have with people. And I would say meetings have the best context for sure.

Speaker A: Yeah, yeah, yeah, totally. And again, so we integrate with notion. Um, we do, yeah. So whatever the system is. Like, our point is that like, we just want to be the agent that can communicate with all the right, uh, tools and make sure that we populate them and, and people can decide. So the other thing that's happening now, of course is the world of um, mcps and people are going to have their, their central or their primary AI that they're going to use. And everybody wants to be the primary AI. But you know, for, for example, for our developers, a lot of them use cursor. Uh, and so what they do with cursor is they have the fellow mcp. And so what they'll do is they'll have.

Speaker B: Oh, you guys have an mcp, right?

Speaker A: Yeah, so, so cursor will read the meeting, we'll go get the meeting, and then from the meeting it will effectively produce code. Ah, and that's a super powerful thing, right? It just kind of short circuits. Like imagine you all are discussing how you would solve a problem in a meeting and then you get the cursor to go consume the meeting and then create the solution based on the discussion that was had. Like, this is super powerful short circuiting of many things. And eventually we'll get to a point where, you know, we'll just create that action inside of fellow. So it'll just know and say like, you know, create, um, you know, basically launch the cursor agent. Uh, so, so it's not there yet, but, but you can do it the other way around. Or we have a native cloud connector. If you use Claude, for example, uh, you, you can do the same thing. And so no matter what main AI you are using, you know, we do this with fellow, but you should do this with other tools as well. The idea is you really want to make sure that these tools are connected so that you can consume the knowledge in whatever way that you want.

Speaker B: Yeah, no, I think that's a very good point. And honestly it's, it's really great that you do have that mcp and I think that's a really good example. So you're saying basically you finish a call internally and the people who use cursor that are developers, it will automatically go from meeting complete to working on the code based on the action items that you determined in the call. Is that correct?

Speaker A: Yeah. Or based on the discussion, like the entire discussion in the call or.

Speaker B: That's incredible. I mean you could do that for anything, right? Like if you have anything. Yeah, not just coding like with this, with the ability now to like I said with Claude, coworker, whatever it is, I guess Claude code for people who are more high tech at the moment. Like you can basically have action items be acted on. I've been thinking about that a lot recently. So with your mcp you're saying that's a real thing you can do at the moment.

Speaker A: But even more than that, I'll give you a uh, kind of a non technical example. Now imagine that you're doing a customer interview, right? Okay. And. Or it's a customer feedback session. Say that it's not even meant to like for a ah, pm, but say that you're a customer success person, you're having a call with a customer, you're onboarding them, you're doing whatever. And in the uh, in the call they bring up like one feedback item and also one bug. Okay. So now what you can do is you can ask fellow, you can say, hey, can you create linear tickets or Jira tickets based on the feature request and the bug and it will do that. And then you click post to post to linear or post to Jira and it'll just do that. So again, all this stuff that used to be annoying, you can just automate it away. And when you make something so easy, what it will end up doing is it will make every employee act like your best employee because the best employee will always do these things consistently. But the problem is that life happens, things happen. Right. Uh, and again every single person can be like the most perfect person using AI.

Speaker B: Yeah. And the proactiveness I think is a big thing too. Right? Like someone will be like, oh, I'll get around to it. And then it just. They won't. But it could be caused by the friction. Not necessarily anything nefarious. They're just like, they have friction in the sense that when they're getting started, maybe there's some sort of blocker mentally, um, or otherwise they don't want to get started on it. And then now we have this thing that kind of, I think prevents uh, friction from happening or at least bursts you through that friction too.

Speaker A: Yeah, yeah, exactly. So a lot of it is, you're right, it's a proactivity and I think this is the place that we're getting a lot of the way that we interact with AI today still is on an ask basis. Right. You ask things and things will happen and we are getting to a place. I think one example of this was ChatGPT's, uh, point pulse, um, which was effectively, you know, use ChatGPT, use it the way you do and then overnight it will say, oh, Aiden was asking a lot about these sorts of things and let me go out and do some research and then in the morning just give him some things that are relevant based on the things that he's interested in. Right. So, so there, there is this like, I think this is the place that we're going to go where AI will become proactive. And today most of AI is, you know, driven by us or signals that uh, that, that we give it and so that, that is going to be the next stage.

Speaker B: Yeah, no, it's, it's, it is crazy. I just uh, man, I'd like the, the whole world has kind of opened up to me. It feels like more and more as the years gone on. I mean when MCPS kind of became mainstream and now I messed around with Claude code pretty early on and was just mind boggled by like sub. When like sub agents came out for Claude code. I feel like that was a turning point for people who were very, uh, tech savvy, if not they didn't even have to be developers. I'm not a developer, um, but I've been making workflows for a long time. So I was just able to figure it out. Wasn't that big of a deal. And then there are. Now it's actually in the forefront for people. Right. Because all you have to do is open up CLAUDE or whatever it is and be like, how do I configure this MCP server in my Claude? How do I. It'll tell you. I was having this conversation with a buddy of mine who's a really smart guy. He's like a client success, success person. And we were setting up Claude desktop for him and then he was kind of struggling because there was an issue with installing the MCP servers. It's because when you close it on Windows, for some reason it doesn't force quit everything in the background. So we're like, why isn't saving or whatever. And um, I was like, I don't want to be this on the nose, but ask the magic robot. And he was like, oh, duh People are still not even at that level yet, you know?

Speaker A: Yeah, I think it is really hard, Demetri, because we all have. It's kind of, uh, you know, the way that I would describe it. So today when I. We're wired to rely on ourselves first, but now we're starting to get to the point where that is changing and our first instinct needs to be to go to AI And I know that sounds scary a little bit, and I'm sure we could debate that concept, but the way that I describe it is almost like the way that I drive today. So the way I drive today is I enter wherever I'm going into the GPS by default. Like, and obviously I know where I'm going. I could go without the gps, but I do it anyway because what the GPS knows is it knows my. The traffic and it knows the conditions of the road and. And things like this, and it will tell me the best possible way to get somewhere. And I might not know the best possible way to get somewhere. And so now it's gotten to the point that I don't even trust my own instincts when it comes to this to driving. I'm just like, I completely outsource that, you know, to. To the car. And of course, like, you know, the next step is self driving. And, uh, you know, I drive a Tesla and it does drive me most of the time. But. But the point is, like, this is going to happen for a lot of problems, right? Like, yes, you could go solve everything yourself. Um, but like, the first source of action should be like, have you also asked AI right, to get that second opinion? Because what AI will be able to do over the course of time is it will be able to have more context than you will be able to fit. And so that's the thing. And, you know, maybe we're not quite there for everything, but I think we're going to get to a place where, yes, you could solve the problem. But again, just going back to the driving analogy, you won't know every single best way to get somewhere and which one of those best ways is more likely to get you there the fastest. Um, you might know a way, but if you want the most efficient and best way, you will then start to rely more on AI. Now, we're not fully there yet, but I think that's the place that we're going to get to, uh, for a lot of problem solving. And then the trick becomes, do you know what questions to ask? So it doesn't all of a sudden become the. So it starts to Become like, where should I go? And that's the thing. You decide, but you don't decide how to go there. And so then the hard thing and the. The thing that we have to focus on is we have to decide where do we want to go.

Speaker B: Yeah, I think that's a fair point. And, you know, a lot of people are probably going to be skeptical of, um. How do I say this? Uh, I think a lot of people are concerned with what you just said, probably because it's like, oh, you're just completely offloading everything in your brain to something else. I am actually the same way about driving. I think it's pretty stupid. People don't put it in every time, even if they know where they're going. Because what I found out is that there are things my parents would drive a certain way to a location because they know that's the way to get there. And it was four minutes slower all the time. Like, I found this out, uh, the second I started being the one who just had airplay or. Sorry. Yeah. Airplay. Is that what it is?

Speaker A: Yeah, Carpet.

Speaker B: Whatever it's called. Yeah. Carplay. Carplay. Yeah. And I was just like, my parents have been, like, taking an extra five minutes to get here, like always. Ah. I'm like, this is psychotic. Um, because it's just like, taking main roads or whatever. And I found out, like, getting to my high school, I kid you not, was five minutes quicker through a route of, like, going through some of the suburb, and it was just through the, uh, subdivisions. And I was just baffled. So I think this is something where it's great to acknowledge that, you know, people should. And I even think it's a context thing. Like, I'm not offloading my ability to think this through. I'm offloading the initial jump start. And if it looks dumb, I still can analyze it. Right. Or if, like, say, the GPS tells me go a certain way and I know for a fact there's going to be a traffic spike because it's right before the Bears game. Of course I'm not going to go there. But practically speaking, at this moment, it's correct. Or almost correct.

Speaker A: Yeah. I mean, I think this can be taken to many places. I mean, to each their own. Right? So I, you know, for anyone watching and I'm wearing a black T shirt, looks like you are, too. I don't know if that's.

Speaker B: I'm the minimalism guy.

Speaker A: So are you, like, every single day black.

Speaker B: I only wear the black T shirt. Yeah. I'm a minimalism person.

Speaker A: Yeah. Same Here. So the reason that we do it, I assume, is for the same reason, which is why, like, why the hell would I bother about it, right? Like, we could, you know, if I'm going to go do some special event or something, like maybe I'm going to do the brain processing, but I much rather my, my thinking time and processing time be used for other things. So I could think about the best way to get somewhere or I could spend the time musing about something else. That's like much more valuable to me. Now. It depends on what you want to do. The option is there, you could always do it. But I think we're just getting to a place where it just gives us even more optionality. There was a time where you really had to look at the map and focus on that, but now you have the luxury of being able to decide where do you want to spend your brain power.

Speaker B: Yeah, I totally agree with you, man. And I think, I think we're kindred spirits in that sense. I mean, uh, the people for years who, uh. When I was getting into the minimalism shtick when I was in college, and people would make jokes about it because I expressed it verbally. And then as I had gotten more and more out of the environment where people knew that that was a thing. Nobody has acknowledged once that I only wear a black shirt. And there is a reason because nobody cares. And the only thing I care about is this guy being offloaded. So I do think, um, we're both on the same page there, but we are kind of reaching the end of the episode. So one last question that I'd like to ask you is just to kind of close things out, if you had to give one piece of advice for everybody that's trying to implement FELLOW into their company, uh, give your final thoughts on how Fellow, uh, is going to improve even more this year so that they can save time and then tell them where to go to find you.

Speaker A: Yeah, I, I would say, like, I'm gonna get a little bit philosophical on this one, which is, it used to be when you and I would have a meeting, like, primarily the beneficiaries would be the both of us, right? We're. We're in the meeting and there's the dialogue and, and it benefits us. But we're now getting to a place where when we have a meeting, what we're actually doing is we're producing context. Right? And context is the most important thing in the world of AI. And to not capture the context is. I think, uh, it's. It's a big crime. Um, and the benefit of having the context in this meeting captured is think about all the ways that we can remix this content. Um, we can say that we have a problem that we've been debating and we just can't figure it out. We could take the contacts and then say, um, you know, send it to an AI to be able to help us solve it. We might take parts of it, turn it into a memo, and send it to the rest of the team. Uh, we may take portions of it and turn it into a wiki article. We might. There's so many things that we could do with it. And all of a sudden, every time we have a meeting, it's not just for the benefit of us, but it's actually for the benefit of the. The organization. Every time we talk, it's like the entire organization gets smarter. And so this is the power of AI in meetings. Um, it starts from what fellow is able to do. But this is kind of the direction that we're going. We're going into this modern organization, this futuristic organization where what a meeting means even changes. So, yeah, and if people want to find it, uh, it's easy to get to fellow AI and. Yeah. Check it out and let me know what you all think.

Speaker B: Perfect. Well, thank you everyone for listening to this episode. If you liked it, make sure to hit that like button. Subscribe on YouTube and also make sure to hit us up on Apple podcasts. Leave a review and let us know what you all think. Thank you so much for watching and we'll see you in the next one. Bye, Sam.