# EP #21 The First Day You Stop Opening Apps

“EP #21 The First Day You Stop Opening Apps.” AI at Work by AscendAI, 2026-04-09. https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/

Credit the named speaker when quoting a passage, and link to the episode or the passage.

Transcript licensed under [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/).

### [Kevin Williams · Passage 1](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-1)

Hey, good morning, ALI.

### [Elijah Szasz · Passage 2](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-2)

Good morning, Kevin. little earlier than we usually record and on a different day as well. ⁓ yeah. ⁓ on a spring break here in St. George and there's like 25 people in this house so I'm hiding in an unoccupied room not speaking too loudly. ⁓

### [Kevin Williams · Passage 3](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-3)

Needs must, needs must. And I know that you had to sneak away from your family to do this podcast. So we all appreciate that. But what you need to tell your wife is what I tell mine is that this is one of my outlets. So I don't talk about AI to her for another hour. So it's going to spare her that blathering.

### [Elijah Szasz · Passage 4](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-4)

I did. Absolutely. I think that's a fair point. really do. by the way, happy ⁓ ⁓ of two weeks of peace. Because usually everywhere go, that's all I'm talking about is AI. And people are asking me about AI. But I got say, the few days leading up to this, ⁓ everyone was like, mm. talking more about ending civilizations and as they're going to be a nuclear war. That was kind of the topic everybody was bringing up. So now I'm looking forward to more AI news. Maybe that's going to start surfacing in the headlines again for the next two weeks before we go back into a regular news cycle. We'll get to hear more about what's going on with AI.

### [Kevin Williams · Passage 5](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-5)

One can hope.

### [Elijah Szasz · Passage 6](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-6)

I mean, you and I are the ones who are always running around yelling, why aren't more people talking about this? Oh my gosh, did you see this new model? This is self-recursive. And people look at us like we're crazy. I'm like, oh, OK, yeah, there's actually a bigger topic in the news right now. It absolutely involves AI. it really does. Speaking of, yeah, I've been trying to surround myself with more humans lately, this being one example.

### [Kevin Williams · Passage 7](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-7)

it probably involves AI too, so.

### [Elijah Szasz · Passage 8](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-8)

being in this house with 25 people, most of them kids before you think I'm on some fun spring break trip. It's mostly kids. Um, and I started doing some group bicycle rides again, and these are road rides. And this is a little different than when we would go on a group mountain bike ride, like you're used to, because those are usually going up the mountain where I try to say something to you and you yell back and say something like, are you honestly trying to have a conversation with me right now? All my heart rates at 185. And then there's the other side of the mountain where I'm just like hanging on for dear life, trying not to crash. Right. these, these group road rides, have these long stretches of flats where you can kind of ⁓ up next to somebody and have a long conversation. And had four different conversations about AI with four different types of professionals. And I to get your feedback on some of this. thought it was really interesting because three out of were really optimistic, like a really optimistic future. about AI. And these were people that I thought may or may not have their entire livelihoods disrupted by it, but they still felt super, super optimistic. this first one, strangely was the CEO of a big SaaS company. Not super, super big, but in the of millions of dollars of revenue, SaaS company. And ⁓ I ⁓ buddy, are ⁓ kind of worried the SaaS apocalypse or somebody coming in and vibe coding your whole product and just racing to the bottom for pricing. He shrugged. He's like, no, absolutely not. We've already had a couple of people take a run at knockoffs and then it had so many bugs that the few customers that left actually came back to us. And I don't know, Kevin, I was going to ask you, do you think that's just a snapshot of where we are in time? Like this is a very finite problem of somebody vibe coding a relatively complex SaaS solution. a few users because the pricing is so much less because they haven't had years and years and years of development, maybe tens or hundreds of thousands of dollars of development fees. Or is it just a now problem? Like ⁓ vibe is so early, right? That all of the shortcomings that people are finding with these products that are built really fast are going to be fixed, let's call it, know, ⁓ six months now. Like that'll... That'll be a thing of the past or maybe a year at the longest. Like, what do you, what do you think about that? I was, I was a little shocked at his feedback of being just completely unworried at all.

### [Kevin Williams · Passage 9](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-9)

I mean, a couple of different observations. One is that it's a point in time that the technology will continue to get better. And I can you that there is some general guidance that's floating around out there that if it's for personal or if it's for internal use, by all means, just like Vibecode it, build it, and it's fine. But if it's for commercial use, concurrent users, like network load, ⁓ are all kinds complexities that are involved in that. But as I can speak to from my own like vibe coding commercial adventure with my Leadworks lead magnet tool, now doing ⁓ today we're going to do testing with a thousand concurrent synthetic users. So I don't have to necessarily wait for to like deep into a beta phase for stuff to break. ⁓ I'm breaking things now. Now I'm also being a lot more diligent about the way that I'm building this than I think some people are, but can totally build something that's robust. takes longer and it does cost money. ⁓ yes, you do need to get dev involvement, ⁓ assuming all of, all of the things that are going to be produced from the, the, the vibe coding channels are going to be janky ⁓ is a mistake. And there's technology continues to improve. think I've given this example before, ⁓ but being able to ⁓ conceptualize something that you want to approach, be you knocking somebody else off or solving a problem, essentially the models continue to get better, you're probably going to get to an alpha from the time it takes you, it'll take you two hours to do the like ideation and back and forth, ⁓ and 10 minutes to do the first alpha, like long enough to go get a coffee and come back. And that rapid iteration will be a problem as those insurgents get better and better. The other bit that would concern me if I'm him is the nipping at the edges. That's okay, two different types of nipping at the edges. One is if companies are getting not necessarily leaner, but more focused in the way people are being applied, fundamentally they probably need fewer seats in whatever the SaaS is. And I don't know what this is, but But I'm not saying that like an accounting platform or whatever, you're going to fire all the accountants and you need fewer seats. But if they're doing their job internally correctly, they're probably deploying more agents to start accomplishing more tasks. So the five accountants that they had on staff that were five seats in whatever SAS might only need to be two seats, not because they fired the other three to be clear, but they've redeployed those people in a way that they're intercepting the data from this guy's SAS. because that's gonna be hard for him to deal with. And then there will be people who will decide, particularly in the SMB community, if the weight of the subscription is high enough, should they nip at it and be able to do whatever piece of it that they feel is most important for their business. And whatever that is, I think it's very easy ⁓ when building a platform to build it. build everything for everyone because you kind of have to because otherwise you just like end up with all these edge cases that drive people away. So you spend all of this time developing all those features. ⁓ if you look at the features that your core user actually uses, ⁓ might be surprisingly small. So the question becomes not can somebody replace my entire SaaS, but can somebody replace the five things that 30 % of my audience ⁓ on?

### [Elijah Szasz · Passage 10](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-10)

Right.

### [Kevin Williams · Passage 11](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-11)

And the answer might be yes to those five things and no to the entire thing.

### [Elijah Szasz · Passage 12](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-12)

Yeah, and in this case, you know, I think it's important to mention this is B2B very niche inside of manufacturing, right? And in reference to what you pointed out of what you have vibe coded, I think it's also important to point out that you're a team of one, right? And this company has 150 people in there. You're a team of one. And what you did is an extraordinarily complex, but it's pretty complex.

### [Kevin Williams · Passage 13](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-13)

Yeah.

### [Elijah Szasz · Passage 14](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-14)

You know, it wasn't very long ago in time that it would have taken a team of many Kevin's and by Kevin's, mean, actual developers, uh, to, to, put that, to put that thing together. And as a team of one, you were able to do that. And so the story that we've been hearing is that these teams of one will go out, find a piece of expensive SAS vibe, code it themselves, and then go after a company that does have 150 miles to feed and say,

### [Kevin Williams · Passage 15](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-15)

Hahaha

### [Elijah Szasz · Passage 16](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-16)

hey, users, we're going to give this to you at 50 % of the cost, which is what he was seeing, essentially. But when those users went over, they're like, ⁓ wait. ⁓ didn't really understand the nuances of this link in the supply chain or hooking up to this ERP. And ⁓ that token doesn't really work the way it was supposed to. And they ran into enough little edge cases that it made it too annoying. And they went back and paid the higher bill per month. And so I was just curious, is ⁓ you this a little window in time we're looking at right now and all of those nuances are going to start to evaporate ⁓ that team of one will start to be able to build what's closer to a perfect clone for a fraction of the price. And my feeling is yes, ⁓ think that will happen. You just like the other conversations we have about AI, when it's like, well, I can't really do this. And we're like, yet, you know, this changing every week. So that one a little, little surprising as far as like just the optimism in terms of what he has to offer the market. Second one, I wasn't too surprised. This was a creative in film, around the world of shorts for commercials. And seen lot of ⁓ AI marketing stuff up in the world of video. ⁓ People ⁓ coming with video models, building on top of old video models, ⁓ OpenAI killing off little form video model Sora. So a lot of churn in this corner of AI. And his position was people are always going to want organic people showing up with actors and cameras lights and makeup ⁓ getting permits. you know, in my mind, I'm just tallying up the bill for all this stuff. Right. And thinking, really, is everyone going to always want that? That sounds really, really expensive. And, know, the customers they go after have the money. to pay for those things, and they want organic, and they want it to be authentic. I was like, huh, OK. ⁓ super, super optimistic. I was like, I don't know. I don't know if you're really seeing end game of this or how fast it's going to change. I was bit undecided on that one. What do you feel like about as far as that niche of creative space where there's lots of video production involved?

### [Kevin Williams · Passage 17](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-17)

Hmm. I know I said this a few weeks ago, but we're in this place where everyone can be a storyteller and, you know, stories can be products, they can be businesses, they can be creative outputs like that. And there's a creative spark that's common with all of that. And then up until now, there have been all of these technicalities and technician type things, which the people who do them would probably call them art. Okay, fine. ⁓ a lot of money that's involved. So in product, it would be a whole stack of devs and such. creative, like you're describing, it's a whole set full of cameras and lights and people and grips and people who know how to use all of those things ⁓ such that story can be told. Well, the story is probably the most important part there ⁓ and the story gets told ⁓ is I think that you are going to see an appreciation for organic content certain veins. Like there's the same thing with art, that people who have been spinning up these AI-driven art companies are surprised because they don't do well. Well, who wants to spend like 300 bucks for a giant piece of AI art? Well, when you know there's a story behind an artist, and ⁓ you care about those things, ⁓ care that it was a human. there's artists as well.

### [Elijah Szasz · Passage 18](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-18)

It's what we decide to place value on, right? It's just a social ⁓ contract.

### [Kevin Williams · Passage 19](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-19)

Yeah, exactly. And might find that that organic storytelling of person to person, so storyteller to actor to film and everyone knowing in the audience that it's the actors, I buy that. I totally buy that there's a need for that. but I don't buy that you need the red camera and you have to rent it for $40,000 a weekend and you have to have all of these people to make this stuff happen. No, you don't. You can be using AI production technology to rapidly increase the speed of your filming and decrease the cost. That's great. But from a commercial side, okay. So again, like commercials, commercials, network and things like like that. it's ⁓ transient content. ⁓ I hate to tell you, but the data suggests that people like the AI content, whether that's good or bad from a humanity perspective. ⁓ commercial message is that highly tailored content ⁓ take take BTS, the Korean boy band. Like I read a piece that that their first off, I think everybody knows that it's not that it's synthetic.

### [Elijah Szasz · Passage 20](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-20)

transient content. Yeah.

### [Kevin Williams · Passage 21](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-21)

music, but it's incredibly algorithmically dictated music. And there are little tweaks between markets that they make such that it will if it's something that's destined for the Asian market, the beats might be a little bit different than music that's destined for the US market. It might even be the same track, but it sounds different to different audiences because they can optimize it that way. And the same is true with short form media.

### [Elijah Szasz · Passage 22](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-22)

Yeah.

### [Kevin Williams · Passage 23](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-23)

If you can optimize it to the nth degree based on data that you know, again, I don't think this is good for art. I'm not making that claim, but if that is the way that you can rapidly make money from art, I promise you that's the way that commerce is going to go. And perhaps it's going to hit a ceiling at some point where people reject it they want to go back to the artisanal style. You hear this term, this is a crazy term, artisanal code. So like,

### [Elijah Szasz · Passage 24](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-24)

Yeah. It's like kids buying Polaroid cameras in vinyl again. Yeah. ⁓

### [Kevin Williams · Passage 25](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-25)

Yes, and that could be a thing that could be a total trend or you know, am I really going to pay more for an app that that a guy dutifully coded in his basement? You know, like, I don't think so. But maybe

### [Elijah Szasz · Passage 26](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-26)

Yeah. Yeah. Yeah, I think you and I are netting out around the same on this. had the same opinion, especially when and synthetic is indistinguishable. That's ⁓ at that point, right. And it's again, you know, this kind of gets into that conversation of talking to an AI that you can't tell if it's human or not. Right. Like the Turing test, but actually ⁓ something a deeper where

### [Kevin Williams · Passage 27](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-27)

Yeah.

### [Elijah Szasz · Passage 28](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-28)

you really, really can't tell where there is no difference. And then you got to ask, well, is there a difference? So if you're looking at an end product of some creative pursuit in film, and you really can't even tell the difference. right now, again, maybe because this is just the trend, ads, short form ads on Metis platform are performing much, much better than the actual organic ads. That's just the way it's working right now. Who knows if there's going to be a correction? But when you can't tell the difference between the two, when nobody can, then I think it's just really going to be driven by economics 100%. Right? Well, what other incentive is there going to be? Like, I'm going to pay for the 35 people on this sheet of line items to get this little commercial put out, or I'm just going to have a couple prompt engineers sitting here working on this video until it gets right and then stitching it together for, you know, one one hundredth of the cost. So that's probably where it's going to head in my opinion.

### [Kevin Williams · Passage 29](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-29)

So, know, in the business news side, ⁓ we don't do a lot of this, but Ben Affleck, of all people created, created versus invested in versus like shepherded along an AI company. It's called Interpositive. He sold it to Netflix for $600 million, ⁓ but interesting to look at what it does ⁓ because not like purporting to do, you know, art in a can ⁓ type But what it does is what I was describing, that it can help fill in the lights, it can help do retakes, it can help with do these really expensive, fix really expensive mistakes that always happen in Hollywood. And Netflix, who is not like a pie in the sky type company, looked at it and was like, yep, shut up and take my money. And that's pretty telling. really not great for the artisans. ⁓ I called these people technicians earlier, but the artisans who are doing ⁓ design and things like that.

### [Elijah Szasz · Passage 30](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-30)

Yeah. Yep.

### [Kevin Williams · Passage 31](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-31)

It's not great for them, but you're also not making necessarily Oscar caliber films this way. you know, even commercial stuff, I can't even imagine or whoever having a choice between really good AI generated content that doesn't tick the uncanny valley ⁓ costs $50,000 or...

### [Elijah Szasz · Passage 32](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-32)

Right.

### [Kevin Williams · Passage 33](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-33)

a nine month process to make a toothpaste ad that is hiring at talent who you have to pay residuals to and all of this thing and all of these meetings and that costs $500,000. Like at some point, those decisions are gonna be pretty obvious as the quality continues to improve. Does the quality need to improve? Yes. Are we close? Yes. Will we get there? Yes.

### [Elijah Szasz · Passage 34](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-34)

Yeah. Yeah. It's, funny because I think just video as a genre of all the things that AI can do, feels really choppy where there's very little progress. And then all of a sudden you're like, Whoa, there's a bunch of progress. ⁓ then there's very little progress again. ⁓ then a company like, open AI says, yeah, we're just going to shut this all down and focus on more serious enterprise stuff for awhile. It's so again, ⁓ think, talk about a lot. Is there going to be improvements and rapid change? Yes. What does that rate of change look like in that specific part of the whole genre? Who ⁓ Who knows how fast it's going to move?

### [Kevin Williams · Passage 35](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-35)

So I want to go on to your next archetype, I think it is funny to note that, so my Leadworks thing is it's leadworksai.co. Like I'm saying that not to promote it, but because I'm using a Sora video on the homepage and it's a Lead Magnet Generator. And I really liked the video. I think it's really funny and I think it conveys the point. And it's a really good Sora output. And I done it around like,

### [Elijah Szasz · Passage 36](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-36)

Okay.

### [Kevin Williams · Passage 37](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-37)

boat brokers because my brother's a boat broker. And I was like, ⁓ man, I'm going to need to do this like more generally to actually sell this thing. And I went into Sora and I was so sad that it was dead and it was dead, dead, dead yesterday. And then I had to try and reinvent what I had just done with like VO3 and it was way harder. ⁓ I didn't really get there to be honest in the way that Sora was getting me there. So now I have a hole.

### [Elijah Szasz · Passage 38](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-38)

Dead dead, yeah.

### [Kevin Williams · Passage 39](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-39)

⁓ And there are other there are other approaches to this but you know the the rumor was that Sora was burning 15 million dollars a day for like cat videos and you know at some point even like, you know, Silicon Valley Bank is like hey guys Maybe you shouldn't be doing that

### [Elijah Szasz · Passage 40](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-40)

Yeah. So what is this product going to make money ever? What's the plan here? Yeah.

### [Kevin Williams · Passage 41](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-41)

Yeah, or just like drain the lakes of the Midwest, right? So, but I have, I now have a hole like, somebody will fill this hole and I will now have to pay for it. When I went to VO three, which is the Google video generator, lo and behold, free uses are gone now. So you have to hook it up to an API and you need to pay the Piper for your use, which okay. I'm fine with, I'm not going to, I'm not building a business where I need to create 400 videos.

### [Elijah Szasz · Passage 42](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-42)

Yeah. You can, you can, you can get pretty far for, you can get pretty far for 50 bucks. Yeah. Yeah.

### [Kevin Williams · Passage 43](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-43)

Yeah, yeah, exactly. And it was absolutely worth 50 bucks to me, but I was a little despondent because I went from having a good idea, working with a good prompt, because again, prompt engineering is still kind of important. And I think I did three, maybe I did three versions of it in Sora. then I hit one that I liked and I was like, cool, this is great. Move on with my day. ⁓ I'm back in this land of fiddliness, but I want to.

### [Elijah Szasz · Passage 44](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-44)

Yep. Especially with video. You can get really nuanced with those prompts.

### [Kevin Williams · Passage 45](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-45)

I want to hear about your next talk.

### [Elijah Szasz · Passage 46](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-46)

Yeah, yeah, I mean, just to put a bow on Sora, it really was the easy button of video generation. You know, it really was like you could have a very, very haphazard prompt and get something very close to what you're looking for. Where Vio is definitely more finicky. And I would say a lot more full of surprises, things that I didn't want where I'm having to rework it. like a lot of the other Gemini powered products, I find myself getting into these little ruts where now I'm kind of stuck in something that I don't want and then I have to start fresh again. So ⁓ very very easy. And this, this third guy actually was also very upset that Sora got shelved because he was using it heavily. ⁓ So another creative, but a different kind. So this is an entrepreneur with a very small team making products for himself, actual physical products. he's also a creative. And said that AI has completely changed how he feels about his work because all of the mundane stuff, all of the pushing around things and building reports or having to do presentations for different companies that he's trying to sell to. We're just done in a few clicks, like just totally, totally done. And he said these were things that would tank an entire day for him before. And he would absolutely dread. You know, we talked about people filling out their warranty cards for this product, right? ⁓ then you'd have to move all this data around and then put it inside of their CRM and then look at eligibility and track dates and all this stuff. And he said, yeah, that's just, that's just done. So for him, AI been this huge unlock of going from the work that he loves to do, the creative work and the creation of products, and then a bunch of administrative stuff on the back end of that to focusing almost entirely on the creative part, the part that he loves about his work, and then just offloading all the rest to AI. And he said, AI has just, it's improved the quality of my life in regards to loving what I'm doing more than anything else that's happened since I started this company. And he said, yeah, you know, I'm looking at this irrational lens and I do realize that some of the things that I I doing and that's now what I'm getting paid to do could also possibly get replaced by AI. Like maybe this is starting off all fun and but it could go in a dark place for me as far as me getting replaced. said, but you know, his his take was ⁓ I just think of all of that right now. and the tools that have been provided to me have really, really changed the quality of what I do. So I thought that was interesting.

### [Kevin Williams · Passage 47](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-47)

Okay, so this is, I made one note on this show, which was the end of fiddliness. And I just said that term a minute ago and like anybody who's bumped into me this week, I've been like saying fiddly all week and you and I are built differently. I know this, I am, I I'm probably very challenging to work for because I'm just busting with ideas and things and whatever. And then ⁓ I up in, in, in what's a personal hellscape. of trying to like track all of the things because I understand how important it is from a project and a organizational management perspective to make sure that everything's staying on track. But I hate, hate with capital letters being forced to go into like some platform that has all of these like fiddly bits to it. looking at you HubSpot or a click up or you know, ⁓ like

### [Elijah Szasz · Passage 48](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-48)

I was going to say a CRM. Yeah.

### [Kevin Williams · Passage 49](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-49)

You know, I get it. get how important it is. Like it is really, really important, but, the structure doesn't work with my brain. Like I, and communicating what I need in a delegatory way to staff, like it's really rough because I, I, I I speak really, really fast and there's like all of this information coming out of me. ⁓ then you have to catch it and run with it and get it into a system. And I an, did an experiment this week where I live my entire

### [Elijah Szasz · Passage 50](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-50)

The structure, yeah. Yeah.

### [Kevin Williams · Passage 51](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-51)

entire week or the last, I guess, six days only in Claude and using Claude connectors and apps such that, and there's, there is more here and we're not going to go into the details, but creating Claude cowork, like, like ⁓ that up certain things. They're running on a clock at five o'clock in the morning, like Slack accumulates all of this stuff from ClickUp and from HubSpot.

### [Elijah Szasz · Passage 52](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-52)

that are running on a clock. Yeah.

### [Kevin Williams · Passage 53](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-53)

And then I can move right over to Claude and I can manage my day from there. honest goodness, don't think that I ever have to go into ClickUp again because ⁓ I get the reporting I need as the head of the organization ⁓ from it from ClickUp. ⁓ I push things into it. ⁓ I a one-on-one with a staff member yesterday and I simply said, ⁓ happens to be another Elijah. ⁓ And just said, hey,

### [Elijah Szasz · Passage 54](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-54)

Yep.

### [Kevin Williams · Passage 55](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-55)

give me an update on all of Eliza's tasks and status and et cetera. And bam, it was all there. I didn't have to like jump over to click up to do the thing. And it's been incredibly liberating for me because ⁓ would dread that, that, that, that part of my day ⁓ the fiddliness. And now ⁓ am, I'm, I'm free of it. So that's for the, for, if you want to move really fast, and this also ties into the team of one, ⁓ this is maybe.

### [Elijah Szasz · Passage 56](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-56)

Yeah.

### [Kevin Williams · Passage 57](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-57)

you know, now that I'm so enabled, it takes me less time just to do the thing than it would to explain the thing to somebody else on my team and then have to wait for it to come back. Like I end up doing things that I probably shouldn't be doing, but they get done so fast such that I can move on to higher value things. And organizationally, I think that's something I'm going to have to deal with again. I'm not, I'm certainly not going to like do away with people, but people who were doing certain things, they will no longer do those things because I'm enabled just to make them happen either completely automatically or with a couple of clicks.

### [Elijah Szasz · Passage 58](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-58)

So, okay, I wanna talk more about what you're working on, because I did something similar this week. But I just wanna be clear that what you're doing is far more sophisticated than the upside the sky is seeing, right? You're building things that are running while sleeping, connecting dots for you, plugged into a bunch of tools directly that you don't wanna have to bother going into. ⁓ where this guy is probably dumping in five different spreadsheets into a custom GPT that he built that then takes all of the data he's looking for in those five spreadsheets and puts them into one report, right? Or loves Notebook LM, right? Because he can put everything about his product and then everything about the customer inside one of these notebooks and then say, ⁓ me a pitch deck, you know, and then it just pops it out. And this is something he would have spent an entire day doing. What you're talking about, I think is worth touching on a little bit because it's very new. Like this idea of, of, of Claude having these connectors for other tools is relatively new. And I want to bring up that there's two flavors of these connectors, right? There's the ones that Claude publishes where you just press one button that asks you for your log credits, and then you go right in there, right? Then there's all of these other MCP tools that you can hook up. to Cloud are a little more brittle. And by a little more brittle, in my experience so far, you have to reconnect them at a minimum once a day. ⁓ So will be working. You're getting your information. Clickups is actually really good. I've only had a break one time over the course of two weeks ⁓ for to go ahead and reestablish the connection. Other ones, not so much. There's a great to-do app. that has a free tier that gets me just about everything I want called Todoist. I really like it. I like the way it's organized. I like the mobile app, the desktop app, everything else. And had this connector to Claude, but it wasn't the one that Anthropic published, right? And it every single day, every single day. I'll ask my little project, hey, know. what's going on in my to-do list right now? Like what's coming up? What's due? You know, what should ⁓ change the date on? Can you add some notes over here? It'll do all of that stuff just through the conversation. And again, that's great. I don't have to go into that specific app and type a bunch of stuff. I can just have a conversation the way that you do with ClickUp. But day it says, ⁓ I don't have access to that tool. Like, no, you do because we did it. It's like that. I don't see it here. ⁓ And I have to go back into, you know, not not the Claude app, but I have to go back into the browser version, into my settings, remove the connection, reestablish the connection. So the workaround I found, I thought you'd appreciate this, Kevin, that in Cowork, I actually ⁓ give it ⁓ ⁓ I it go into the browser-based version of Claude and reestablish that every day at three in the morning. Like it just disconnects, reconnects every day at 3 in the morning so that I don't have to go back and have this thing broken. And that the broken connection about 50 % of the time. But other connector tools I have in that same project, such as I also have Google Calendar. I have my email. have some forwarding rules set up. One caveat of all of this, especially in the world of ⁓ these Google products, is only get to pick one, right? So if you have five different email accounts, well, Pick the one you like the most because it's only going to support one of those connectors right now. I'm sure in the future you'll get all of them. Same thing with your calendar. So I have some fancy forwarding rules set up. But I now have the singular chat that is the closest thing I've had to a personal assistant so far. Because like you, can say, hey, what's going on in ClickUp? Can you tell me how much content we have for the next two months? will. It'll automatically look for stuff and flag it and say, oh, you know, this is past due. It'll do the same thing in my to do list. can say, Hey, I just got off the phone with Kevin. We're going to have a meeting with these guys next week. Can you please put it on the calendar? And then instead of me having to go into my, Google calendar app and create an event, it just does that for me. It puts in all the event details. It will put together email drafts for me and say, Oh, I forgot I was going to email Kevin and tell them about this great mountain biking spot. Can you create a draft in my Gmail? It'll do that.

### [Kevin Williams · Passage 59](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-59)

Mm-hmm.

### [Elijah Szasz · Passage 60](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-60)

So like you said, all that, all that fiddly tapping through different apps and connecting all the dots and typing things with your big fat thumbs on your iPhone. I feel like those days are in their last. Like the stuff is getting simple enough to set up where you really can have this. Uh, you know, it, we've been talking about it for a really, really long time, right? Like, what is the next, what is the next smartphone? Oh, S going to look like? Is it just going to be a little prompt box?

### [Kevin Williams · Passage 61](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-61)

Yes.

### [Elijah Szasz · Passage 62](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-62)

you know, like a little text box where you put in a prompt and that's, that is the OS, right? I feel like we're starting to get there with these tools.

### [Kevin Williams · Passage 63](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-63)

a couple of different things. feel like I want to more than double click on it. But is you're also doing this verbally. I happen to know, you know, we're both big on to dictation. So being to just basically talk to it is ⁓ is can muse, can meander like it is like having a real deal executive assistant ⁓ who's standing in your office at attention, like waiting for you, waiting for your missive to go actually do stuff, ⁓ not move some paper around, but actually go do it. ⁓

### [Elijah Szasz · Passage 64](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-64)

And when you Yeah. And make suggestions. ⁓ that's what's so cool, right? Because as you meander and just kind of like ramble into this dictation, it might ask like, Do you want me to put that on your calendar? Or would you like an email? Yeah.

### [Kevin Williams · Passage 65](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-65)

So yeah, exactly. And do it in a smart way. ⁓ Like so from the calendar connection perspective, you know, it's a virtual world. Most of us have virtual meetings, but I have an in-person meeting after this. Part of why we're meeting so early, but I have to drive a half an hour to get there. It picked that up that I was going to have to get into a car and it suggested, hey, do you want to block time before and after? Because it looks like you're going to be driving. And that's exactly the sort of thing that would have, would have caught me previously because I maintain blocks and whatever. But I had to instruct it to do that. wasn't like, Hey, I'm going to, I look for that type of thing and suggest that it's out there. But you said something else that's interesting that I think that, that our listeners need to know, which is it's one chat. So you're I'm creating a single chat in Claude.

### [Elijah Szasz · Passage 66](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-66)

Yeah. I would manually put those in those little blocks on my calendar so I wouldn't get stuck in the car. It's so annoying. Yeah.

### [Kevin Williams · Passage 67](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-67)

I'm contextualizing that based on the information of my day that I'm getting from, there's some stuff going on behind the scenes, it's not like overly complicated, ⁓ it's basically summing my day. And still a little janky in that I actually cut and paste it from Slack to browser. And there's some reasons for that, ⁓ but the browser has all of my context about everything that's going on in the rest of my world, ⁓ it's more powerful than trying to do it in the desktop version. But then I stay in that single chat throughout all of my tasks. And unless I'm going to go vibe code something or unless I'm going to work on a very specific project, I've gone from having 19 clod tabs open to having like two open. And I just run my day because it knows what's going on. And then it does a midday check in. The other observation for your friend with the SaaS company is by doing that, and this is a little bit of a divergence. Basically, I'm using ClickUp and HubSpot as just glorified databases that are really good at structuring CRM or project data. And they are really good at structuring those things. But if I never actually have to go there because UX, my experience is in Cloud, that's going to be a problem for any SaaS that allows it. ⁓ any SaaS that doesn't allow it is going to run smack into ⁓ preference. of needing to allow it. you know, okay, he's doing like sort of manufacturing stuff, but what happens when his big client is like, look, dude, we were starting to adapt to this other stuff. We need to have API access.

### [Elijah Szasz · Passage 68](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-68)

Yeah. We just want to have a conversation about it, not click through 15 screens.

### [Kevin Williams · Passage 69](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-69)

Exactly. then when he says no, they trigger the computer use to go and do it anyway. like, know, he's going to block his customers IP addresses? No, he's not going to be able to do that. ⁓ is you look at the SaaSpocalypse. And I had a ⁓ client asked me yesterday about SaaS valuations. It's like, well, this is what's happening. You're losing seats. People are nibbling at the edges. The essential business model is being decilled to just being a database.

### [Elijah Szasz · Passage 70](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-70)

Yeah.

### [Kevin Williams · Passage 71](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-71)

and you're not gonna really be able to prevent any of it by doing like clever pricing or barriers because that will just accelerate the cycle. So yeah, 20X valuation on a SaaS company maybe doesn't make sense anymore.

### [Elijah Szasz · Passage 72](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-72)

No. Well, that's the other thing I was going to say when you were talking about HubSpot and how you're almost at the point where you don't need their UI and you don't like their UI, you don't like going through all those different screens. also becomes quite an expensive database. ⁓ you remove the UI from it, you're like, ⁓ is okay. It's a database, but it's a very expensive database. So the utility I'm now getting from it, I feel like ⁓ a disproportionate value proposition here that, you know,

### [Kevin Williams · Passage 73](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-73)

Yes! Yes!

### [Elijah Szasz · Passage 74](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-74)

might need to reconsider in the future or vibe code something that is just using the features that you really need for that database and then having the same entry point that you have now. It's hard not to imagine a future where that's going to start happening more and more because things cost a lot of money. HubSpot seats with all the features you want are not cheap.

### [Kevin Williams · Passage 75](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-75)

They're not cheap. it's, it's material on a P and L. So you start paying attention to it. And that's, you know, we, a while ago, I talked about like my own matrix of risk for SaaS companies and like the more critical you are to a business, actually the more at risk you are, because the more critical you are, the more the business is going to notice the expense. Whereas if you're some like little tool that they use occasionally, they're not going to try and like recreate you. It's just easier to buy the thing. So. If you're something that people pay attention to, it's going to be a problem. And I have to laugh a little bit about just the speed of all of this. So almost two months ago, two and a half months ago, so right at the beginning of the year, I went on this big kick of trying to build my own click up. Okay. Like, ⁓ yay, I'm a great vibe coder and we can like figure all this stuff out. and I. I, we selected something called Plane, P-L-A-N-E, which is like an open source. Yeah. It's kind of like an air table type thing, but it's structured around project management. And I thought, okay, cool. We could use this thing to build our own system and we're going to activate it all through Slack and whatever. And the short story is it didn't work because it was going to be too fiddly to manage the platform itself that we built because people were going to need to be in the platform. ironically, I ponied up and I bought.

### [Elijah Szasz · Passage 76](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-76)

⁓ another one,

### [Kevin Williams · Passage 77](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-77)

up licenses for the team and I was like, okay moving on we got to focus on things that are actually gonna make us money ⁓ now ⁓ went to my ⁓ team member who's administering the whole thing and I was like I actually think that we could do plane ⁓ if you don't care about the UX and you don't care about how that works plane would work great because it doesn't matter that it's kind of ugly in the back end and it's kind of hard to navigate ⁓ long as the data is available in a way that that cloud can use it

### [Elijah Szasz · Passage 78](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-78)

Yeah.

### [Kevin Williams · Passage 79](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-79)

Point for our users that they shouldn't miss is, you and I were obviously become like big fans of Claude. Claude does a lot of stuff. GPT has something called skills ⁓ they've launched in their browser that works remarkably similar to Claude cowork, ⁓ it does it in the browser. And that actually can be really, really powerful. ⁓ don't despair. Like, don't get me wrong. Like GPT is super, super powerful. There's a lot going on. Cloud got the edge as far as a lot of these automations, but it's not like isn't paying attention. So if you're on GPT, check skills and check out the connectors that are connected to it. And you actually find that that's gonna work great for you. ⁓ You can them, they can connect to a whole bunch of other things. I haven't gone super on it, but...

### [Elijah Szasz · Passage 80](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-80)

Yeah.

### [Kevin Williams · Passage 81](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-81)

don't necessarily need to bail on GPT and go to clot. They're all going in the right direction.

### [Elijah Szasz · Passage 82](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-82)

For sure. mean, everybody's marching into the same territory, especially when they see something that really sticks and has promise. And you're like, ⁓ whoa, maybe this could be the future right here. Every other lab is going to follow the same direction. You said something I was really curious about, that you're still copying and pasting a lot of stuff in that kind of personal assistant workflow, where it's coming out of Slack, and you've got to take this, and you've got to paste it into there. And I'm guessing that's because you built this in Cowork. And I'm curious. why you're not just using like the browser-based chat inside of cloud to do that word. Cause you can, you can have those connectors in your settings there as well. Not as many. And some of them aren't as robust, but you can, you can do it.

### [Kevin Williams · Passage 83](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-83)

I have to admit that part of it is just a hardware restriction. My poor Mac, I have a Mac M1. And when you're running as many Chrome tabs as I'm doing, I run into system instability issues. So ⁓ basically, I need to pony up for a better Mac. I'm at the tail end it. But I think a lot of people who have consumer grade like,

### [Elijah Szasz · Passage 84](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-84)

Okay.

### [Kevin Williams · Passage 85](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-85)

equipment. This is a good Mac, but it's five years old and like it's it's at the tail end of of its life cycle and people are not necessarily realizing how much of this stuff is taking local load on their machines and we're not a hardware show at all. But if you have just like an off the shelf like Mac error MacBook air that has eight gig of RAM. you're actually gonna run into some problems if you're doing like cloud code or if you're running a thousand different browsers and things like that. also just kind of find the workflow a little kludgy with ⁓ the And my like talking point this week has been like, I feel like we're between user experiences right now ⁓ because there are bunch of ways that you can do this ⁓ and they all a little bit different.

### [Elijah Szasz · Passage 86](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-86)

Yes.

### [Kevin Williams · Passage 87](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-87)

⁓ And it hasn't quite been unified yet. And I think that we might be heading for a different experience that's a little bit more proactive. And yeah, I mean, I could spend the weekend and I could go out and I could build myself a dashboard that allows me to give like a single source of truth for all of this stuff and then trigger the cloud conversations or the cloud chats off of that. But I think that might be heading our way and that the way that we can all picture an LLM chat might be. I don't know what this is gonna look like, but it might be about to change in a way that it's encompassing a lot more information as more of like a control center. And if you want it super simple, ⁓ if you ⁓ to be able to pull in these things such that it's not all of this duct tape pieces of information, there's going to have to be a UX change. ⁓ for me, from a behavior perspective, I did try and automate this stuff. I tried a bunch of things.

### [Elijah Szasz · Passage 88](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-88)

I think you're right. Yeah.

### [Kevin Williams · Passage 89](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-89)

And, you know, maybe I need some tutelage on some other ways I can do it. But at the end of the day, triggering my awareness in Slack such that, that I'm always in Slack. So it's an unread message. It's their first thing in the morning. I, I first tried to build it such that I would just do everything in Slack, but one that would end up on the API. So that could get really expensive. ⁓ two, because it was like local knowledge, it didn't have the basis that browser had. So I gave up and decided, you know what? I'm just gonna put a copy button in Slack and yeah, couldn't be bothered to actually highlight and copy. So that I could just, you know, it's one motion, grab it in Slack, dump it in the browser and move on with my day and not.

### [Elijah Szasz · Passage 90](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-90)

Is there a robust Slack connector for you for Claude? OK. OK. I'm off of Slack now. And if you're paying for ClickUp, by the way, you can have all those same chat features that you have in Slack inside of ClickUp for the same license fee.

### [Kevin Williams · Passage 91](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-91)

Yeah, it's great. It, it, yeah. Don't get me started on Slack. Slack might be the most valuable part of, I'm actually not a hater at all. Except for the noise. We're not, we are a workflow show, it could get really expensive, but it might be the most durable asset that Salesforce has because it's cross client. ⁓ bet you still have a Slack account and if you needed, you could spin it up.

### [Elijah Szasz · Passage 92](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-92)

I used it for over a decade, yeah. It's expensive though. ⁓ It's really expensive. Yeah.

### [Kevin Williams · Passage 93](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-93)

because I have whatever nine clients who need to use Slack for communication and I have to use it. Like even in the future world, idea of me connecting to their ClickUp communications tool, like there's just no way. ⁓ we email as ubiquitous communication, depending on who you work with, it might be WhatsApp ⁓ and is the other one that

### [Elijah Szasz · Passage 94](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-94)

Mm-hmm. I know. Yeah.

### [Kevin Williams · Passage 95](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-95)

those are the only ways, I mean, that you can communicate. So, Slack's in a good place, and if they jacked up my price, I would find clever ways of getting around it, I'm sure, which again is a problem for them, but it's in my column of very unlikely to unsubscribe.

### [Elijah Szasz · Passage 96](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-96)

Yeah, I mean. again, have the things you do internally and ⁓ the that you do externally, client facing things, right? ⁓ And internally getting off of Slack and moving to ClickUp where we are already paying for seats saved us about $10,000 a year doing that. But as far as getting people outside of our organization into ClickUp chat and off of Slack, one out of four was just like, hard no, no, we're not going to get into another real time chat. already in Slack. You guys got to talk to us in Slack, ⁓ right? still, I'll take the 10 grand savings and just deal with Slack when we have to. one of the things you talked about as far as being in this with these tools where it's a bit fragmented and it's hard to stitch everything together right now, but it's getting there. ⁓ I one of the things with Claude Cowork that you're pretty deep in right now is that that is not a mobile experience, which is why I tried to build everything in their browser-based project or a singular chat is because I want to spend more time out doing things and not be tethered to the machine that all of this magic is running on. They do have a tool called dispatch. So right now I have like a daily marketing ⁓ automation runs every day at six. And I've been away from my machine that this is running on. And ⁓ also no syncing, by the way. So if I've got my laptop here while I'm traveling, that also has the desktop app of Cloud installed on it. of those co-work tools that I've built, they're not on this machine. They're not syncing with the Anthropic Cloud. ⁓ all standalone, right? So I've got this Big Mac studio ⁓ in Salt Lake City, ⁓ ⁓ five hours south from there. ⁓ ⁓ I'm using a free product called TailScale. and then screen sharing on my Mac so I can actually look at my desktop and make sure these things are running correctly. because I don't want to get up at 6 a.m. when this automation starts running while I'm on vacation, ⁓ been grabbing my iPhone in the morning, opening up the cloud app and then hitting this feature called dispatch ⁓ I can actually talk to my computer and say, hey, how is that marketing automation doing today? And it'll say, good, on turn five hundred and seventeen. I saw got caught up on this and it'll give me a little report. of how that automation is doing. So you can see that it's getting there, that the pieces are all coming together, but it's all still a little bit of, like you said, copying and pasting and duct tape and everything else. But rationale for trying to do all of this inside of a cloud project or a cloud thread is because ⁓ want to have that continuity between devices. I want to walk away from my desktop and then check in at lunch. and maybe ask that same assistant who has that persistent memory of what we've talked about, what's coming up on my Google Calendar. I want it to all be tied together. And I'm getting closer. one that I came across this last week that I think is worth mentioning is that when you're ⁓ these different ⁓ or projects or anything else, you have a set of instructions. You think of it as a very evolved prompt. very long and it thinks of everything and it might have a whole series of steps for whatever you're doing, right? ⁓ you've got this set of instructions ⁓ ⁓ are different ways that Cloud can go reference those instructions. And in Co-Work, one of my absolute favorite things about this ⁓ makes it very different than having ⁓ the instructions for like, let's call it a custom GPT or a Gemini Gem or a Cloud project, any of these reusable things you're going to be doing ⁓ that Just like you don't want to go into your calendar to type in all the details of how long is my commute time and how long should I block and who do I need to invite, I just want to have a conversation about it. I have one of these tasks ⁓ I see a better way to do it based on an output, I also want to have a conversation about how to make that thing run better without having the conversation first, then looking at the instructions and either editing some lines in those instructions or pasting in a new file or uploading an attachment. And in Cloud Cowork, you can do that. So if you have these tasks that you've built, you can say, the way, when you're this bit of data from HubSpot, I only want these two fields. I don't want these four other fields you're giving me. ⁓ And say, great, would you like me to update your instruction file to reference that? And then you don't have to do any of fiddly stuff, right? So I've putting my instructions inside of a document in ClickUp and using that connector. Basically, before I start that conversation, the first thing it always does is it reads those instructions and it reads a second file in ClickUp, which is a memory file of everything we've been talking about. And that way, I don't get this huge token cost of having to go through an endless, really long thread to pack everything up. just writes it to a memory file. So I'm using a fraction of the tokens by being able to ⁓ start a new conversation in the same project in Cloud. ⁓ it just references a memory file it's been writing to and its instructions. And whenever I find something that I don't like the way this assistant is doing something or I want to add something to it, we just have a conversation and it writes it in those instructions for me. Huge, huge, on luck, huge on luck.

### [Kevin Williams · Passage 97](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-97)

Okay. I have to, I have to laugh because this was like last night. My Elijah and I were going back and forth about this, because, ⁓ we were trying to do something clever with ClickUp to create personas for the staff such that Claude would have, they'd know Elijah as a dev background ⁓ you know, he handles technical client engagement and whatever. And, is working on this.

### [Elijah Szasz · Passage 98](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-98)

For like task routing, like just so it would be smart about task routing. Okay.

### [Kevin Williams · Passage 99](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-99)

Yeah, exactly. Task-granting. But the challenge that we were having was where does this document live? And we landed on ClickUp for the same reason, because it's so accessible. I had actually just done the same thing that you're describing using ClickUp. ⁓ using ClickUp as the document repository because the data connection is so clean and stable ⁓ trying to mess with a document. ⁓

### [Elijah Szasz · Passage 100](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-100)

Yeah. Google Drive, for example, very finicky about how Claude gets their files. ⁓ no, it's not a wait. No, this is a .md format. I can't read it with this connector tool. great. You put it on a Google Doc, but now I can't write to it. There's all these hangups. But the ClickUp document, and the way, it doesn't have to be ClickUp. Obsidian is another one of these products that also has a huge free tier. ⁓ we'll the same thing. Motion, yeah.

### [Kevin Williams · Passage 101](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-101)

Very finicky. Yeah. So. Yeah! Sure, Notion, you know, yeah, a lot of people are doing stuff with Notion. Obsidian is kind of a scary road for mere mortals, like, ⁓ little bit more developing. ⁓ like, a lot of people are familiar with Notion, ⁓ just think about that. You have a document that's on Notion, ⁓ like, just you would with a task,

### [Elijah Szasz · Passage 102](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-102)

It's a little more developer-y.

### [Kevin Williams · Passage 103](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-103)

You're using Claude with some automations in it just to keep that document up to date. So it knows all the things that happened yesterday. This is a point in time though. Like obviously that's stupid that over time, like Claude itself, it's just all going to be in Claude.

### [Elijah Szasz · Passage 104](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-104)

It would just all, it would just all be in clock. Yeah. We wouldn't have to be paying playing like pass the baton between a bunch of different platforms and jumping through hoops. yeah.

### [Kevin Williams · Passage 105](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-105)

And that is, it is challenging. It's I made a note to myself and it's funny. I, I learned things from these conversations as well or things come to mind. So I have a, I have a daily task that, looks for really neat ways of using Claude go work to do other things. And I realized that I hadn't actually connected it to my daily planning. So I have to, I have my daily planning task, but off to the side, I have all of these neat uses of Claude code and cowork. So what I'll do today is I'll basically link the two such that whenever I'm doing something repetitive, Claude can recognize that I'm doing something repetitive. ⁓ can look at its own repository of clever things that other humans are doing with Claude ⁓ make a suggestion for an automation or a workflow that I might be able to just take off the shelf from GitHub or somewhere else and put into place.

### [Elijah Szasz · Passage 106](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-106)

Exactly. Yeah. And talk about this concept of portability a lot, right? ⁓ ⁓ another point is when you have ⁓ dynamically updated instructions sitting in a plain text file somewhere that your LLM is accessing, if OpenAI or Google wakes up and says,

### [Kevin Williams · Passage 107](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-107)

Yeah.

### [Elijah Szasz · Passage 108](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-108)

Wow, we've done all of this a lot better. It's like, guess what? You have your entire workflow there in one document. It's not like you have to bunch of tease a bunch of stuff out or like recreate this GPT as a Gemini gem with all of its new, it's you have the whole thing right there, right? Which I think is a super fascinating part about all of this.

### [Kevin Williams · Passage 109](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-109)

We haven't really seen how big SAS is going to react because they're definitely trying to figure it out. talked about that a minute ago, but there's going to be a bit of a fight or flight moment here. Maybe it's not tomorrow, but in the next couple of years, going to see pushback ⁓ it's probably going to be enterprise level stupidity as far as the way that they do it. And will either change the way that a lot of these different pieces work or it will accelerate. adoption of ways of doing things. And that was a little bit mealy mouth, like, you the corpus these big companies is not going to go down quietly. So ⁓ like, they to figure this out. Like they absolutely have to figure this out.

### [Elijah Szasz · Passage 110](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-110)

Hahaha No. And there's so many, there's so many different paths to how they could potentially figure that out. Right. Is it going to be, you know, ClickUp has been trying to figure this out. Like they, they get it right. understand what's happening. The writing's on the wall. And so you can choose a tier that gives you all of this AI functionality inside of ClickUp that I think is just calling like chat GPT for something. ⁓ and it's incredibly expensive. It's incredibly expensive. And you only get a fraction of the power that you'd get from using the models natively. Right. again, I'll get way more utility for way less money by just using the connector that Anthropic provides me to have my own tools talk to ClickUp than I would paying for that seat and ClickUp that gives me the AI tools. ⁓ it's not a singular path. Like there's many different ways they can approach this. That to me is probably not the one that's going to win. So anyhow.

### [Kevin Williams · Passage 111](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-111)

Yes.

### [Elijah Szasz · Passage 112](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-112)

Yeah, yeah, we'll see how that one shakes out. think we've covered a lot here, Kevin. There's a lot more, of course, that we want to cover.

### [Kevin Williams · Passage 113](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-113)

We have, I want to go build. really do. I wish I could, I've got a busy day. There'll be no visit building today. Well, thank you for joining us on your vacation.

### [Elijah Szasz · Passage 114](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-114)

Yeah, I know. know. Instead, you have to go to a meeting. Yeah. Absolutely. Yeah, I had to take a break from mountain biking at some point. there you go. I know. know. right, Kevin, next week.

### [Kevin Williams · Passage 115](https://ascendlabs.ai/podcast/ep-21-the-first-day-you-stop-opening-apps/#transcript-115)

⁓ that's rough. That's super rough.

