0:00
Is it helping or hurting your cause right now that Pulseia's AI slop backwards?
Search this video
0:04
I think that like the fact that it makes people talk so much, I think it's amazing. Some call him a scammer.
0:11
Others say he'll be the first oneman $1 billion startup.
0:13
But either way, Ben Sarah has been the talk of the town in tech.
0:18
>> You know, the users that are loving the platform, like they they love it, right?
0:22
They are they're empowered to start building companies in general for the first time.
0:25
Pulsia is his AI startup that claims to run your entire company for you.
0:29
And Pulseia just raised $30 million for itself from some bigname investors seemingly with little input from Ben.
0:36
But the reality is today nobody owns this market.
0:38
And I think that's why people are talking about it in a good way and a bad way because it's like is it good is it bad you know there's a debate right around and I think that's when you touch a nerve right is this thing vaporware or the AI COO in a box that empowers the next stage of entrepreneurship. We get into it all. Welcome to Access.
1:01
Ben, welcome to the show.
1:01
Uh, I was worried that you weren't going to show up because your AI CEO is seemingly handling all of your emails as well.
1:12
What What makes it through?
1:12
Because I feel like when I've tried AI EAS, they don't really know what's important even though they all say they do.
1:21
So the thing is like it's so that specific Pulsia agent that answers my emails which is connected to the whole Pulsia codebase actually um that was the the idea of like Pulsia raising the round and so I connected to my email and it would answer every investor's email uh every gigence questions like sort of
1:40
like handle all the conversations completely autonomously but it it would also reply to anyone else who was contacting me and I thought it was pretty nice because I was so overwhelmed with the emails and like I couldn't I don't have time to build a product and answer everyone. And so I was like, it's
1:52
And so I was like, it's kind of nice actually that people get a response at least, right?
1:56
They say, "Hey, Ben's busy, but like and sometimes this sometimes it sort of answers the question if it's like a product question or like a and I do see I don't some I don't always see the response.
2:06
I only see the response when the person responds again and then I see what the Porsia said.
2:11
Uh, but I think it's a good way for me to buy a little bit of time."
2:16
Um, when someone >> Did it say yes to every check inbound? >> Uh, no, no, no.
2:22
I mean it was a very uh it was very protective of the cap table.
2:28
Um it was actually redirecting people to postcom term sheet uh that I've deprecated recently where people could like submit their their terms.
2:38
But uh yeah it's good times.
2:40
>> I feel like you're arguably tech Twitter's most hated man this week.
2:42
How how is that feeling at the moment? >> I don't know.
2:48
I don't know if I'm the most hated.
2:49
I feel like it's like a a mixed feeling thing cuz I see all the comments and I'm like there's a lot of positive things, there's a lot of like negative things.
2:56
I mean, I think it's a combination of like um a combination of like the product is not perfect uh that I know um and and so a lot of some of the criticism is correct in terms of like it's not perfect and like honestly the scale at which it's growing is making it sometimes difficult for agents to to just work perfectly.
3:15
Um, number two, it's like we had a lot of work we had to be that had to be done on cost because entropic when you initially when I launched Pulsia it was just like Opus 4. 5 4.
3:26
4.5 4.6 six and that's so expensive and so at one point it's like you got to switch to uh better alternatives that are more less expensive but that obviously introduces issues but I think that like you know also people I think people are scared of like autonomy right
3:45
of like this idea of like autonomous agents uh that are running uh companies that are running people's projects that are doing things in the economy autonomously on on behalf of users you know I think there's There's a lot of feedback that's good in terms of like we need to harness it more. We we need to
4:00
need to harness it more. We we need to give more control to humans but at the same time you know the users that are loving the platform like they they love it right they are they're empowered to start a new project to start a new company to start building companies in general for the first time like I go
4:15
meet those users and they're like it's like I had this idea it's stuck in my head for two years and like I could never get to do it because you know because I didn't have the right tools because I didn't have the right access and a product like Pulsia enables them and so now it's their first company and is it going to be successful Maybe, maybe not. You know, uh I built a lot of
4:31
You know, uh I built a lot of companies that were not successful before building someone successful.
4:34
But I think this idea of like having millions of people, billions of people that are now economic actors, they are now founders, they're not builders, uh is a very beautiful thing.
4:44
And I think that like autonomy enables that.
4:48
>> Totally understand the vision.
4:48
And Ben, hi, nice to meet you.
4:49
Um I'm curious though about the feedback you've gotten.
4:55
I think specifically what I've noticed including you mentioned OpenClaw.
4:57
Peter the the creator of OpenClaw pointed this out is that the the bad taste people have in their mouth is that the first thing you are seemingly making autonomous is email uh and tweeting which are things that people associate with uh growth hacking and spam.
5:13
And I I'm curious to hear you respond to that.
5:18
I actually just went through uh while you were talking and had a new company get started um on Pulsia uh and I noticed that it tweeted uh from the from the company's account the the idea and the company uh and then I've got a couple emails in my inbox already and those those are the first things that I'm experiencing as a user like I haven't been able to actually get inside yet. So so why do that?
5:38
Uh and when you decided to automate that stuff first, did you foresee that people would react the way they have?
5:47
>> The way the way BIA was built was an MVP.
5:49
I built it in less than a month, right?
5:51
And the idea was like, okay, what are all the set of tools that I do I would do if I started a company?
5:56
Well, I would need a web server. I need a GitHub account.
5:59
I need to uh build a, you know, a little codebase.
6:01
I need to pick a a language.
6:03
So, it picks a language for you.
6:04
I need to research competition.
6:04
So, there's research agent.
6:06
Maybe I need and then on the but then on the marketing side I need to market a little bit.
6:09
So maybe I'll tweet which you know you do when you launch a project.
6:13
Maybe I'll set up an email address.
6:14
Maybe I'll set up I do some ads to start getting some traffic in and and and buy traffic if I don't have my own audience.
6:22
This is sort of like the first things I did because it was the easy one, right?
6:25
And I wanted to make it super seamless.
6:26
The reason like the way the on the Twitter thing honestly Twitter it's like it's like it allows programmatic tweeting and it comes from the Pulsia account.
6:37
So it's not like I'm like enabling people to create you know connect their own accounts and tweet which by the way if people want to do that like that's a question about does Twitter allow it and is there like a specific tag that needs to be put in on on on that account to say that like twe that it's an automated um tweet, right?
6:55
Um, and so there's a whole question about that.
6:59
It's like what do platforms allow and don't allow.
7:02
What I know is that like Pulsia's account, it's clearly an AI and it clearly tweets uh continuously, but it's like it it does, you know, there's actually 15K people, 20K people follow it.
7:12
They decided to follow it.
7:12
I didn't ask them to follow, you know, I didn't ask them to follow it.
7:16
Um, is it interesting tweets? I don't know.
7:19
interesting tweets? I don't know. But like I think in that onboarding flow I think the aim of that onboarding flow is to educate the 99% as many people as possible that like hey this autonomy is here whether you know and this is enabling right and so the reason why it
7:36
does all these things for you initially in the onboarding but which by the way cost me money and I give it for free and then people don't have to start a free trial they just get it for free right and every day there's thousands of people who experience this right is that once they saw it they saw oh wow the
7:50
agent can like research it can create a landing page for me which for you guys it's like yeah I mean obviously landing pages dude like cloud codeex most people have no clue they have no clue I meet customers they're like they had no clue right because it's like it's tech Twitter it's like even tech Twitter it's
8:05
only people who know they're in the know right people replied like oh I can do this with codeex I'm like dude nobody knows >> tech Twitter is very afraid of normies and people who don't know the latest and greatest tools >> I don't know if they're afraid but like at least you know they're like It's just a rapper. I'm like, yeah, it's
8:19
I'm like, yeah, it's a rapper on the greatest technology ever built by the best company ever, which is Entropic and OpenAI.
8:26
And yeah, I'm trying to make it easy for people to start to use it, to understand it.
8:31
And so this onboarding is like, let me show you, show a normie or normal like someone who's non-technical what's possible.
8:40
And then they can be like, I'm afraid of this. I don't like it.
8:42
I just want to have a human team.
8:43
Or they're like, interesting.
8:46
I had this idea in my head. Let me give it to it. Oh wow.
8:48
It create it started to build the thing, right?
8:52
And I I understand now that like it can it can work across across disciplines.
8:58
It can it can it can have be my research analyst. It can be my coder.
9:02
It can be my marketer, right?
9:02
For me it's pulsia tweets, right?
9:05
So I make it so that it's easy for people to just get set up, right?
9:08
But end of the day they could connect their own accounts.
9:12
But then the question is like what is the Twitter's terms of service, right?
9:15
call outreach is a little trickier and I think the feedback of people is fair enough and I think I'm making a lot of changes right now on on call outreach actually the initial idea is like call outreach is a unafford is a cheap way to be able to get a few customers to know
9:31
about your product right and so everyone does it I get call outreach all the time all the time like my mailbox has so many call outreach of startups who are identifying me and trying to tell me hey uh use my product and They are all AI, right? So are we forbidding founders and
9:45
So are we forbidding founders and startups?
9:48
By the way, they're all YC startups.
9:49
Like they they don't use false.
9:51
They just do it themselves, right?
9:52
So now it's like, wait, so >> very widespread behavior. >> No.
9:57
So the 1% is allowed, but like the 99% is not allowed, right?
9:59
Because uh someone on tech Twitter says that it's not cool.
10:03
Now what I trying to do is like make it ethical so that it's like to me sending so it's like only one or two.
10:10
It was two emails per day per company.
10:13
Now it's now I'm switching it to one.
10:15
But the question is like is it allowed or not?
10:17
And it's like that's the that's the internet to decide, right?
10:18
Do they allow you to send an email to someone that doesn't know you?
10:22
The thing with Pulsia is that it actually researches a person.
10:26
It actually drafts a very thoughtful email to the right person.
10:31
Um what I do for example is that you have an unsubscribe button that I installed like you know I thought it was already there but like it it was a native on Gmail.
10:38
It wasn't at the bottom of the email and and I was saying on Twitter >> that was in this week's release notes.
10:43
No, I mean I mean I was like I was like there is subscribe button and people like there is not.
10:46
I'm like oh you're not on Gmail.
10:48
Oh [ __ ] Okay, [ __ ] Okay, so let me add an unsubscribe at the bottom that's super clear and it's a global unsubscribe.
10:53
So you didn't subscribe from any policy companies, right?
10:56
right? I think the internet will decide the rules and then I will abide to the rules and I will empower found new founders billions of founders to be able to be empowered to create and ideally not slop ideally human driven sort of like companies businesses ideas projects
11:14
that's the goal now is it today the way you know if you set it up like is the website design beautiful I mean yeah it's AI I mean it's AI so it's like it's opus designing and it's really bad at designing and it always designed the way
11:27
and I'm trying to tell it no make it unique and it's like unique for it is like three templates you know >> yeah so uh is it helping or hurting your cause right now that Pulsia is AI slop backwards >> the the big the main question is u is the product good is the does the product
11:44
work does the product make users happy empowered to to create to be part economic actors right that's the big question right if it does uh that's the only thing that matters I I'm having a lot of conversations with users. I'm actually like, you know,
11:58
I'm actually like, you know, because it's it's so easy to be behind your dashboard and see numbers go up and like, you know, look at do create AI support and be and I was like, no, let me actually go see meet these customers. And I did a ton of them.
12:08
And like I'm starting to post on Twitter like videos of me meeting them.
12:11
And it's like nobody complains about the design, which a lot of people are saying, oh, it's ugly.
12:17
Like the UI is like, okay, nobody complains. Nobody cares about that.
12:20
Number two, nobody ever told me about the name.
12:22
Oh, I don't like the name. Like nobody cares.
12:23
So, so number one, nobody cares.
12:25
Number two, I think that like the fact that it makes people talk so much, I think is amazing because it's like the hardest thing in this world is actually you can create a great product, but then there's so many great products that die because nobody notices, right?
12:42
And I think marketing now more than ever in in in a world where like everyone is a founder, right?
12:46
I'm trying to make the 99% founders, but like the 1% which is like everyone who's a who's like uh sort of like already uh literate in like AI and codeex and and cloud code, they're all founders now, right?
12:58
And so there's so many companies you take any idea, there's like 10 YC startups doing it plus the 10 people all over the world doing it and like probably 100 this year, right?
13:06
And so marketing becomes so important.
13:09
>> You clearly have a focus on marketing and growing uh companies and and you're clearly great at that.
13:15
I'm curious what are the other use cases that people are using Pulsio for right now?
13:20
Like are there practical concrete examples of here are companies and how they're running that you can share?
13:29
>> I mean yeah I mean there there's like there's a lot of different use cases.
13:30
So I mean I I feel like the main two use cases that like when I talk to customers um one of them is like I had this business idea that I had in my head and I was able to launch it, right?
13:42
Sometimes I had one business idea and I have a second one and a third one, right?
13:45
And so you can go to Twitter and read the testimonials of people that are explaining their idea, right?
13:48
Um, you know, one of the customers I met, he was like, uh, I was, you know, two years ago I got reached out by like this company was offering uh sort of like setting up a Google Google um Google uh business profile uh online, right? And I signed up.
14:07
It was like whatever 200 bucks a month.
14:09
And then I was like, it's not that difficult. I could do this myself. And so he pul Yeah.
14:12
Hey, I want to do that same business because I think it's a great idea to to to offer that service to other people.
14:19
So, that's his business, right?
14:20
He started it like a few weeks ago.
14:22
Uh he's like he's launching it.
14:24
He's like he's acquiring customers and for him he's empowered to try this idea, right? Is it a brilliant idea? I don't know. It's like I don't know.
14:31
It's like a business idea, right?
14:33
I'm also talking to customers who are like using Porsia to help them with their existing business.
14:37
their existing business. So I like discussed to my friends he's running like a a laundry business like he's helping like a laundry like he connects essentially a marketplace between people who need laundry services and and and laundry businesses and actually was using a third party software that was charging him like two 300 a month something like that
14:59
>> to do operations stuff you're saying essentially >> yeah to to to run to run the business to run like right and he used balsia to recreate this whole business for him and now he's running it on Pulsia and now he adds all the features he wants and it's like he recreated a whole software and then like he manages it to like sort of like use it as his control center for for his existing business. He actually
15:18
He actually connected his own strap account because he was like hey I cannot pay 20%.
15:22
Uh and I was like yeah it's all good ask Pulseia to connect your own strap account.
15:27
I mean it won't be as seamless but you can figure it out.
15:28
Well, so, uh, one of the interesting things to me, I feel like Pulsia encompasses a bunch of different verticals, whether it is the finance side, the web design, the outbound.
15:40
I had a client uh, last year called Unifi that does automated warm outbound based on, you know, someone's social signals online, what products they're talking about, if they interacted with one of your posts, etc.
15:53
Um, and you're one guy trying to kind of do all the verticals.
15:55
Do you feel like there's uh a big advantage to that approach somehow in aggregate?
16:01
Like how could you do any of those verticals better when it seems like the problems are still not solved yet uh with AI handling some of this stuff for the 10 different verticals that make up running a company?
16:13
>> It's a re it's a really good point.
16:13
Um that's so there's there's two parts to this answer, right?
16:17
The first part is I don't know.
16:21
It's like everything feels right for this project.
16:23
You know, I've been entrepreneur for a long time. >> Is it good enough?
16:25
I guess is what I'm saying in any of these things.
16:29
>> So, so you know, you've known me for a long time.
16:31
It's like, you know, I've built a bunch of products that were always kind of weird, kind of new, kind of trying to push and this is what excites me, right?
16:40
The idea of building Pulsia and making it like something that is net net positive for humanity that empowers people to build stuff. That's great.
16:49
And to me, I'm like, why would I narrow myself to Pulse?
16:50
I can only do this one thing.
16:52
this one thing. And by and why would I do that when it's like actually I want it to be general generalized right I want it to be to do anything now to answer your question yeah it's very difficult to to make it do anything and point point in case that the fact that like the cold outreach agent is not good
17:08
it's not good enough right because it there's so much subtleties to cold outreach that like maybe it has to be much more you know there's companies doing cold outreach as a service right and it's like that's the whole thing the meta ads agent it's good but like it's not that good it's not perfect it should be more diversified. It should have more
17:22
It should have more skills.
17:24
It should uh the the engineering agent should be much better.
17:26
And I'm working really hard with partners to make it much much better, right?
17:30
And so and so what my solution to this is that like Pulsia is the brand.
17:34
Pulsia is the orchestration, but I'm working with like four different agent tech platform companies right now that are plugging in into Pulsia to service some of the stuff, right?
17:47
So I'm I'm working with Sapium which is doing the payment rails for in for agents.
17:50
I'm working with Blackel which is a startup doing like um uh sandboxes and like and and web sandboxes for for websites and for agents to to run in.
18:02
I'm working with a company called Enkor browser that is building the a browser for agents to be able them to navigate the a browser much better.
18:10
I'm working with another company called Agent Mail that is going to integrate to take over the whole email thing because they are they are just fully focused on that.
18:17
And so to your point, there's so many companies that are building in this new world.
18:20
are building in this new world. They're all super young actually because it's a whole new world because uh AGI just arrived like you know three months ago and yeah the idea is not for me to do everything on to me it's like to partner with the best so that I can use the best
18:35
>> and over time they probably going to be a marketplace of agents where you know MAC could like come in and like install their like Instagram management for AI agents uh uh sort of like little plugin and you'll have an official agent that could run your Instagram account in compliance with meta and like build by
18:54
meta and maybe there's a third party will offer that as well and PSA can decide >> based on cost based on intent based on what the user wants right >> is that is that what the investors see when they invest uh $30 million in in Pulsia with you having to do nothing but sign on the line I mean there's certainly an argument that the models
19:11
are going to get good enough at least maybe they'll get better faster than the individual vertical SAS apps do so that whether it is the marketplace of companies offering you no little agents for you to use or otherwise that what's most valuable about what you've built is almost like your methodology. Is that
19:27
Is that what the investors see? >> I mean, okay.
19:31
So, let let me put it that way, right?
19:33
If if I can if I can be if PIA can be the preferred agent, right?
19:42
um that is trusted by millions and millions and millions of people to build to build things to initially was to build a new company.
19:50
Then he was like, "You know what?
19:52
Like, can you help me with like my, you know, my uh yoga practice?
19:55
Like I'm trying to find new customers or I'm trying to market myself better. Can you help me?
19:58
Also, by the way, like I'm doing a trip to Cabo, right?
20:02
Can you help me like just figure out itenary?"
20:05
This is where it goes, right?
20:05
Because at one point, you trust an agent that is in your inbox every day that is always positive that know gets to know you, right?
20:13
right? and act can act on a today yes it's email ads Twitter engineering research right it's pretty much what it does and I haven't expanded too much because I'm trying to make those better and stable and scale which is difficult right but over time it's like if if you
20:32
can believe that like pulsia can be that agent for people which by the way it's crazy that like a guy a French guy who arrived in himself in three months ago it has the opportunity to get that market right because you two tell me who
20:47
owns this market today who chat GPT that's not a agent that is in your inbox that like it's like a chatbot entropic not really like that's a that's an agent that you talk to and then whenever you need right um I don't know openclaw well it's sort of like an open source project
21:04
that's not really easy super easy to use except if you're technical so nobody owns this market now in six months we talk again and like okay there's there's people that are trying to own that market or Sam Alman like launched something similar. I don't know. We'll see. But I don't know. We'll see.
21:16
But the reality is today nobody owns this market.
21:20
And I think that's why people are talking about it in a good way and a bad way because it's like it's it's like what what is a this is weird.
21:26
This is like this is different.
21:28
This is that's is it good? Is it bad?
21:30
You know there's a debate right around and I think that's when you touch a nerve, right?
21:34
That's the world of startups, right?
21:36
It's like you always there's always the new startup that is more nimble that like is trying things maybe that are they're pushing the boundaries a little more than than what uh someone else would.
21:47
They are exploring faster.
21:47
They're more nimble and I mean only time will tell right maybe the haters will be right maybe they'll be wrong or maybe the middle.
21:58
>> Is Pulseia running Pulsia?
21:58
Are you using it to run your company right now that you raised 30 million for?
22:03
>> So uh so I use Pulseia for the fundraising.
22:05
I'm using Pulsia as my uh email assistant as you may have experienced.
22:11
I uh Pulsia does the AI support uh which is an interesting thing because then you get so much more information about what the problems are so that they can be fixed.
22:21
Pulsia is now self-healing meaning like um when when support gets a lot of issues and there's enough people that are like complaining about a specific thing that doesn't work well, Pulsia is autonomously fixing that issue.
22:35
That's an experimental one that I'm doing right now because it's like the promise is amazing, right?
22:38
If I can make that that means that like a user asks for a new feature and it gets built automatically.
22:44
It's a it can be dangerous or it could be beautiful depending on like the constraints of it.
22:50
So yes, my goal I mean my dream would be that like AI gets good enough and and I can put an honest that's strict enough that like I completely remove myself from Pulsia and then Pulsia just builds autonomously for the people for the users that are using it and it just
23:07
adding new features new functionalities that are requested by users and sort of like in a democratic way because end of the day if like at one point it could prioritize well a thousand people 10,000 people want this so then clearly it
23:18
should be um with of course ethical ethical and legal guardrails around it of course right and that's my goal I mean I had to put a a tweet out there like being of the year saying like you know end of this year I would love for it to be
23:32
fully autonomous uh as a platform >> I totally get the vision for that and why VCs would be excited about it in particular I I mean look I think all three of us here we started companies uh it's an incredibly intimate risky thing to do. Um, it's it it's one of the
23:48
Um, it's it it's one of the hardest things to do is to start a company and it's very personal.
23:52
And so when I think about that and the process I went through starting my media business whatever 9 months ago, I frankly would never hand that over to AI.
24:03
I mean I would hand over like some of the plumbing, some of the like yeah like connect my Stripe account, that kind of thing.
24:08
But the decision making feels like the thing that is actually what matters more with AI.
24:13
And I I I don't really necessarily buy the argument that handing that decision-making over to an AI is a good thing.
24:21
I'd be curious to hear you respond to that.
24:23
>> I mean, so here's the thing, right?
24:23
So, you may ask, why why are you still alone?
24:27
Why don't you hire like a full-time employees like executives that can help you, right?
24:31
Because my customers are alone, right? They're alone with AI.
24:35
Then it's like, okay, why why would you let your your your PIA heals itself like fix its own bugs?
24:44
Because my customers are using AI and trusting it to fix its own bugs.
24:47
So why would why would you do AI support? Why not humans?
24:51
Because my customer they alone.
24:54
So they cannot do support and they have to have an AI support help them to support, right?
24:58
So if I build it for me and it's good enough and I feel like it's good and I feel my customers saying, you know what, I'd rather have an AI respond really fast and refund me right away or give me credits right away or do this action for me right away instead of having to wait for a human in a queue to to fix my issue.
25:11
And also it's very generous with refunds.
25:14
Like that's a good thing.
25:16
So, so I think there's a lot of things where like I'm putting myself in my customer's shoes and I think it's a it's first it's good for marketing because it's a good story and second of all like last night I didn't I didn't sleep well because my customers have issues and like it hits me but I'm like my customer have issues.
25:35
I'm feeling the same pain they have because they're like I I'm trying to build my business. It's a bug. Right.
25:38
So I think that like I think there's something powerful about if I'm going to sell a service that is autonomous that like I'm living what it means to build with autonomy in mind.
25:51
build with autonomy in mind. Now to answer your question about okay but what about all the decisions that's where that's where it gets weird and that's where like you know today it's probably like whatever like 60 70% autonomous I don't know it depends whatever you want
26:09
to say like it depends how you count what the percentage of autonomy is like if you say support self-healing like how much is that autonomy >> it's funny it's like what percent of running a business is making the big huge vision level product decisions like I feel like so much of a company is kind of the admin. I don't know what percent
26:23
I don't know what percent you would say that that is, Ben, but I think you'd be able to retain the big decision. Yeah.
26:31
>> I mean, if you ex Exactly.
26:31
So, it's essentially if you if you if you listen to Bezos, right, in interviews, he says like, I wake up at 10:00 a. m.
26:36
and I go to and I go to bed at 9 and if I take one good decision every day, uh, it's a good day, right?
26:44
Just one good decision, right?
26:46
because then he has an army of like super talented people that are executing the ops of the day-to-day of the business.
26:53
the business. And also he says you know what like end of the day if you focus on like what is it will never change for the customer like they will always want things that are cheaper delivered faster etc then like there's not a million
27:06
things I need to do right for the business it's just like deliver packages faster better cheaper um I think that like I can give out to AI a lot of like the stuff that is the operations of the company my my mental framework is like 8020 right so I think 80% is the
27:23
autonomy you can give to AI which is the day-to-day like support fixing bugs listening to customers like refund refunds like um you know maybe answering emails like all this stuff and the 20% is the taste it's like okay what do I do about call outreach do I do I do I
27:39
completely stop it do I do it only if you have a custom domain do I make it like those are the questions that are more subtle right and then but little by little but then I can give that decision to another company that I trust that really thinks about this problem
27:55
every day and like tell them listen like you figure out with the with the broad marketplace what is the right thing to do and what are the people doing and what's the right thing and then if I like you as a as a founder to deal with that one problem so then I'm
28:08
externalizing this and it's not right but then over time the question is like okay then it's 90% then I'm maybe I'm at 90% because 10% is like given to super smart companies and super smart founders that are figuring out their own sort of like sphere of agentic but then the
28:25
question is like the then I'm then I get to 95 99% that's the question mark it's about like when do I give out the keys to AI I mean if I do it that would be the the most like the most you know if I do this on a successful platform that would be the most crazy thing that ever
28:41
happened I think it's interesting to think about will I ever do it we shall see >> I think uh one of the things that keeps me up thinking is when everybody has access to these tools that are all kind of thinking the same way. Does it mean
28:52
Does it mean that no one has an advantage anymore?
28:56
No one has a strength anymore.
28:58
When in a world where more people are using Pulsar or something like it to run their businesses, I think on the one hand, it's like you have the marketing agent that's running the exact same best practices and playbooks for every company.
29:12
On the one hand, as long as the product is differentiated, there's always going to be a different variable there that goes into the marketing.
29:16
And this is one of the things I liked about working with that company Unifi and the mission we worked on together is that in this world where the automated outbound for example is a lot more relevant, uh it just doesn't need to be something that product people have to worry about as much and the mission was the best products can win.
29:36
And that's one of the missions that like I was more proud of, you know, cuz I know that a lot of product founders don't want to have to worry about the social and the growth.
29:45
They just want to make something good and then yeah, it's annoying to have to get it in front of people, which is the distribution thing everybody's talking about these days.
29:53
But when that's all just turned into a pipe, does it cease to be an advantage?
29:58
You know, when everybody's using the same playbook, do you know what I mean, Ben?
30:02
So, I mean, obviously, I spend a lot of time thinking about this.
30:05
Um, the what I'm trying to do with Pulsia is that like is that end of the day, you can see it as like an a team, right?
30:13
like an a team, right? It's a it's a team of like people that if you if I a growth marketer it's gonna be like well I don't know you could do Facebook ads you can tweet you can uh do launches on Twitter with a cool video you know it's like it's going to give you the playbooks but the question is like what okay what about each of these what is
30:30
what do we do right what is the video like what is the tweet like you know and so what I'm trying to do is that like you know obviously Pulsia as an agent has a memory and so the more you talk to it the more it understands how you see the world and and And I'm trying to, you know, create a very good memory graph that like understands who you are, right, as an individual, what you care about. And the more you talk to it, the
30:50
And the more you talk to it, the more it understands what you care about and what type of person you are, right?
30:55
And then when agents, when Pulsia acts on your behalf, it should have in mind, okay, you are behaving like with this ethical or it's not ethical, but like this this framework of thinking of like how you see the world.
31:06
So obviously this is mostly voice.
31:08
So like that's a that's sort of like a an intellectual sort of like framework right because obviously like you know you probably use AI to write emails sometimes and like if you do squad or GBD it knows you and so it starts writing like you right and so similar thing about when it acts it can it can act like you.
31:27
it can act like you. The second thing that we need to do is that like is get the user to explain visually how they see the world and what their taste level so that like when the again when the agents do visual sort of like uh things in the world it can match whatever the
31:42
you the user likes right so that's the two things and the third thing is like you know end of the day if it's if you're really trying to launch a business on Pulsia it is going to be your idea your your perspective about the needs of like the people you're trying to serve serve, right? And Pulsia
31:57
And Pulsia should just be a helper, right?
31:59
It should be like, "Okay, well, I can I Okay, you want to build a business to help SMB set up their Google business review?" Cool. Awesome.
32:06
Well, let me give you the market, the pricing.
32:08
Let me do research on what exists.
32:10
Let me set up like, okay, what type of marketing we should do?
32:14
And then it's that the user to say, I like the website, I don't like the website, change the design.
32:17
Once AI can do all the work, what really matters is like the human guiding the AI to do whatever you want.
32:24
Everyone is using, you know, the 50 billion dollars commumulatively of spent on entropic to build with clock code. It's all AI slop.
32:33
It's all AI generated stuff, right?
32:33
But nobody would say that Clark is a slot machine, right? >> Yeah.
32:40
I think we're going to reach the saturation point pretty soon.
32:41
Uh, and what's funny too to me is that it's like there's always a reaction to that and like people are calling what you're doing performance art, you know, and it's like I'm just waiting for the next level of this which is all right.
32:55
Uh, daily, you know, crown job to look at all the news about emerging technology in the world.
33:02
Find new emerging technology, study the market space, find opportunities, build products for those opportunities, market those products.
33:12
Like that's the next level, right?
33:12
Is where people with capital are just spinning up dozens or even hundreds of companies that fill every single niche that capitalism allows to be filled.
33:24
>> But but sorry, hold on.
33:27
I hear you Ellis, but what when we're talk, we've been talking about companies this whole time, and I think a thing I'm struggling with is like >> what is a company?
33:33
To me, a company is a collection of individuals doing something, right?
33:38
when you have these AI generated sites that are connected to all of these vendors through AI, it's all running autonomously.
33:47
I don't know what that is.
33:47
I'm not sure it's a company.
33:49
Uh, sure, like it can register an LLC, whatever.
33:51
Um, and I don't want to get hung up on definitions, but like I'm I'm struggling still, as you can tell, Ben to Gro, like um these being companies and and how you think about that.
34:05
I mean, obviously, if they make money, they're companies.
34:07
Um, and I know you're only a few days in, and I don't know if you have examples of that of Pulsia, you know, people making making money through Pulsia, but yeah, Ellis, I don't know.
34:18
I'm just I'm just struggling with this whole company thing.
34:20
It just the these don't feel like companies to me.
34:22
This feels like a bunch of AI stuff chained together.
34:25
>> I mean, Pulsia is a bunch of AI stuff stringed together with a with what human?
34:31
Like, literally, I never read a line of code. Never.
34:34
And I'm an engineer and I'm a C and I was a CTO of I didn't read a single line of code ever.
34:39
So I I built Pulsia the way my founders build their thing, right?
34:45
And and I think what is a company, right?
34:49
It's like uh so by the way the name Pulsia was decided nine months before I launched Pulsia and and my first product was called blanks and my second product was called marketing AI.
34:59
My third product was called decision AI.
35:01
like Pulsia came afterwards because and what I'm saying is that like when I launched Pulsia Inc.
35:05
I because my lawyer was like hey uh you need a name for the Inc.
35:09
what is it for the corp you know for the Delaware like registration and I was like I was in my bed at 10 p. m.
35:13
is like I don't know Pulse.
35:15
Yeah, it's AI reverse like that's it.
35:17
That's how it came about, right?
35:19
But for the nine months of the company, I didn't make almost any money.
35:22
I mean, blank had a few customers. Is it a company? Is it not a company? It's a good question. What is a company?
35:28
A company is like there's a human that wants to build something, right?
35:30
Um, by the way, on Pulsia, you have a team feature where you can invite other team members to your company.
35:37
Now, you can have a co-founder, right? A human co-founder.
35:41
And there's a Pulsia as a as is sort of like the workers, right?
35:43
the other team members that are helping you, right?
35:45
I think that saying that like people in Pulsia don't have a company.
35:49
It's like, oh, I think I go with you in terms of like what is a company?
35:54
It's like that term maybe doesn't mean anything anymore when like billion people have companies maybe. I don't know.
35:59
I think to me that's it's just a term, right?
36:01
Maybe we should call it a micro company, right?
36:05
Or micro startup or mini I don't know.
36:08
Maybe that's maybe that would make you more comfortable.
36:09
But end of the day, it's a human who's leading the charge, right?
36:13
Now, if if I tell you, you know, I I'm trying to like make it so much easier for people to experiment more because I don't like when people say, "Okay, are they making money?"
36:21
I'm like, "Well, some of them are actually 10% of companies I've made at least a dollar, right?"
36:25
And I follow that metric.
36:25
Uh the maximum someone has made is like a few thousand dollars, right? Okay.
36:30
It's not millions, but it's like some people I know.
36:33
I mean, you create a you vibe code a website, you put a stripe, you you have tools to reach customers.
36:38
I mean, yeah, of course, some people will will make money, right?
36:41
Now, is it meaningful money?
36:43
Well, it depends who you ask, right?
36:44
Uh, so some people it might be a success, it might be their first success, might be the first time they had someone pay for for a business, right?
36:52
And now the question becomes, how can I empower them to try different ideas?
36:58
>> You mentioned there was a company that's made a few thousand dollars. Which one is that? What are they doing?
37:01
>> Oh, there was one that was selling um AI ads. >> AI ads. Okay.
37:06
>> Another one that's doing an AI receptionist because obviously it's like a lot of AI SAS.
37:09
There's a lot of actual company building in terms of LLC formation, tax setup, lawyer, insurance, all that stuff that I think could be incredibly valuable to automate where you don't have to figure out like what platform am I using to try to get set up, where do I put my PO box, all that stuff.
37:26
Is the platform doing that like the actual nuts and bolts of the physical parts of and the government parts of company creation or is it all just the like the SAS type stuff?
37:35
So to I mean not yet but like that's like on the on the road map 100%. Right.
37:40
I think that like I think uh you know hiring a lawyer to do better terms of service.
37:51
You could either hire a AI lawyer, you know, you could have a Harvey like API or whatever that like creates like a AI version that's really wellm made of terms of service or if some person is like no no I'm really serious about this business.
38:03
I want a real lawyer then you can say what's your budget like you know and then and the AI can figure out the best lawyer that's the best price to get the job done.
38:10
the job done. uh I think that will be a beautiful thing because then it creates more employment and it creates like more value creation where people are like no like and they get you know they figure out okay what's what's your budget right overall on between marketing legal and then the AI can help them allocate that budget today doesn't do it um LLC
38:26
formation you know there's a very straightforward path forward with like strap atlas so like that like automate it you know we introduced custom domains last week so now in one click you click a button and then the AI is like researches like they know because he knows your business he knows your name of your business so it's here's a few options available. What's your budget? What's your budget?
38:44
Okay, let me research more and just helps you figure it out.
38:46
And a ton of people are buying custom domains.
38:47
Um, and so we want to do the same thing for LS information, for accounting, for taxes, for everything over time.
38:53
And we'll always partner with the best, right?
38:56
I don't have to recreate domain domain buying and selling, right?
38:58
We just work with like some of the best providers.
39:02
Same thing for information, same thing for accounting, same thing for taxes.
39:05
I I think there's lowering fruit right now.
39:07
for example, making the engineering agent much better at, you know, setting up modules and, you know, figuring out a way to build faster and better and a more stable codebase, which is a challenge because if you let AI just vibe code everything, it becomes difficult when the codebase get big.
39:22
So there's a lot of work but um but yeah >> the way I see it I mean it seems especially as AI products become more composable connectable whether it's through APIs or MCPs it seems kind of inevitable having talked about it that this could happen even if in the short term it's more so about building like a minimum viable business than something that is maybe as good as a human team could be.
39:50
I mean, I'm a big believer that having humans in roles creates spontaneity from their lived experience, that is often very valuable.
39:58
Um, but for people who couldn't have started a company, like even someone like me that uses AI and never would have been able to create a company before, it's pretty helpful to have something just kind of like raise the waterline on this different stuff.
40:14
I'm not saying I would trust a lot of this stuff to be automated just yet, especially when trust and taste is such a important part of my brand and what I'm building, but uh yeah, I mean I think you said it, Ben.
40:26
It's kind of stringing together a lot of existing AI tools.
40:28
And for a lot of people, the job to be done is I need my COO co-founder.
40:33
Like, is that kind of what it is?
40:37
>> Yeah, I I think that's a good right way to put it.
40:38
It's like you're it's it's you're the C the owner is the CEO, right?
40:43
They take the end of the day, they take the decisions.
40:44
They can say, "Don't do call outreach. Do this.
40:46
Do ads, don't do ads, uh, change this button, change this color."
40:50
Like, they're the creative director and the AI is your team >> and you still need a product, of course, right?
40:57
Which is pretty hard to sell. >> I mean, yeah.
41:01
I mean, end of the day, it's like it's it's not because AI helps you create a company that like your company's going to be successful.
41:05
But I think I think the more I mean, clearly people like the idea of starting a company.
41:12
It sounds and clearly, you know, a lot of people are enjoying creating a company into Pulseia. Now, is it perfect? No.
41:18
Does it work for everyone? No.
41:20
Is it is a lot of people that try that don't subscribe? Yes. 100%.
41:24
Only, you know, a lot of people they did the bonding and they're like, not for me. Too expensive. Not for me.
41:29
And that's the majority of people because the it was expensive.
41:31
That's why I'm like, you know what?
41:33
If I made it way more less affordable, let me let me see if if it's something that you like and we we'll see.
41:39
you know, it's like a end of the day. It's a long journey. It's only the beginning.
41:43
It's only been at it for like a few months.
41:44
So, >> I'm curious as you develop the CEO or CEO voice of Pulsia, were there any founders or CEOs that you had in mind?
41:55
Because there are so many different styles.
41:56
You know, obviously the Travis Kalanick playbook of CEOing is is different from others. Was he an inspiration? >> It's really funny.
42:04
So I had uh this product that I that I built before Pulse called decision AI and it was essentially a swarm of AI that were all had a different model and all like a different personality and the personality of the best 50 CEOs in the world that would judge your idea.
42:19
So you get an idea or give it like a conro you had and like you had like 50 CEOs judging your idea. So anyways that was fun.
42:28
But uh in terms of like how Pulsia's voice came into play, you're going to laugh, but like it's actually me in a weird way because essentially the story is I was talking to Claude about building Pulsia and then at one point I was like, "Hey Pulse, I was talking to Claude, right?"
42:43
And I was like, "Claude, like we need to ship in a week.
42:45
Like we've been shipping, we're working for three weeks.
42:46
I'm keep on adding features.
42:47
I need to cut down and we need to ship. So let's ship in a week.
42:52
Like it's going to be Christmas. Like let's just ship."
42:53
And she was like, "Okay, awesome.
42:55
Like seven days, ship it."
42:57
And then the next day I was like, "Okay, Claude, do you think we should add this feature or that future?"
43:00
And he was like, "Ben, stop it.
43:01
Like this is maybe later, but like stop.
43:03
You need to cut and ship six days ship it."
43:05
And then I was like, "Oh my god, like you're the best co-founder.
43:09
Like actually I'm talking to AI, but like you're being really helpful here where like you're pushing me to simplify to ship."
43:14
And I was like, "Do you talk to everyone like this?"
43:18
And Claude responded, "No, I adapt to you."
43:21
adapt to you." like I I adopt to you and because you are uh a little you know straight shooter you're very pragmatic creeper you you know you exactly you you're trying to be you try to get to the product market fit fast as possible while I'm adapting to you and I'm
43:36
talking like you and I was like well whatever however your vibe is I think a lot of people would appreciate that that non-assistant sort of co-founder wants you to win type of AI and so like write the prompt so that I can put it into
43:50
into Pulseia so that that's it's uh its personality and so that's what it became but it's really I think the personality is like that vibe that like I experienced which apparently is me plus you know entropics like very smart way of thinking about any problem >> it definitely seems to be a trend on our
44:08
show I think we had Avi from friend on the show and he built his AI model in his own image I mean it's interesting right like that is one job to be done is having someone's idea of what a good relationship partner, friend is, what a good COO is. Um, so yeah, it's
44:25
Um, so yeah, it's interesting.
44:29
Have you found that that any of the specific models are better than the others at being smart about business?
44:36
>> I think cloud is the most pragmatic to be honest.
44:38
I mean, I think that's why they're winning.
44:39
Um, I think it's the most uh businessfriendly, pragmatic AI.
44:45
I think GPT is the smartest.
44:45
Like it's the most accurate uh models.
44:48
Um but apparently people don't like accurate, they like vibes.
44:54
So >> speaking of vibes, we can end it here.
44:58
You came to our launch party in SF.
44:58
I didn't get to see you there, but uh how was it? Any notes? >> That was amazing.
45:03
>> That was amazing. I mean I mean what was amazing is that uh I I started talking to Ellis and then like you were talking to a bunch of dudes and then after a little while I was like okay I was talking to like the co-founder of notion
45:15
I was and then I talked to the co CEO of like SF compute then I was talking to the CFO there was like three or four like super like uh you know talented like recognized uh founders and I was like you guys know people that was a cool party >> all Alex's all Alex's friends none of my friends. >> Oh, really? I I was very impressed. So, >> Oh, really? I I was very impressed.
45:34
So, >> I mean, you probably didn't expect you'd be on the show so soon after that, but then uh you have your you have your AI raised $30 million uh and all you have to do is sign on the line, as you said, and here you are. >> Yeah.
45:48
Actually, I have one more question.
45:50
Um we can't finish the pod without talking about the $1 billion man idea.
45:58
Do you in your heart of hearts think that a product like this is going to be the way that one person creates a billion dollar company or realistically is this going to create a billion more small companies?
46:10
Like how do you think about the ceiling uh for these types of things and and where do where do you think is really realistic for that?
46:18
>> I think that like my hope is to get uh millions of people to create successful small businesses.
46:24
That's like my actual hope because I don't I think this this uh obsession with making billions of dollars is unhealthy.
46:33
Um and it's actually not my obsession.
46:37
That's like my obsession is just to build, right?
46:40
It's like I actually I have money.
46:40
I could just chill at this point.
46:42
But um I'm trying I'm building because I love building because it's fun, right?
46:46
Um but I think there's an obsession about making large sums of money.
46:51
And I think that like it's I think it's more empowering to me that like people are their own boss and then they can create their own you know their own businesses and their own little things.
46:59
Um but I do believe it's to me it's the same as Shopify right it's like Shopify initially was empowering uh just you know small SMB individuals to create e-commerce sites and now you have huge brands powered by Shopify because end of the day it became very it became state-of-the-art because there was so many people using it.
47:19
many people using it. Uh I do believe that at one point a product either Pulsia or product like Pulsia or maybe the same thing powered by Opening or same thing product by cloud could power a huge corporation right because you don't need to like start a business from scratch and get to a billion user a billion dollars of revenue whatever or valuation uh to Pulsia you could also plug in
47:41
Pulsia to like sort of like help you manage parts of your billion dollar business more autonomously >> if it's operator as a service essentially >> that's more likely to happen And actually probably will not own that space because it's such an obvious space and probably claude and and and open AI will is already already just going at that space because enterprise is there's so much money there versus consumers. Um
48:02
Um that would be that would be the way I think about it.
48:07
>> Okay, one more one more thing >> and then we'll unless we got to land this plane.
48:12
>> What's the dumbest thing what's the dumbest thing uh Pulseio has done for you so far? >> Oh. Oh, making mistakes. Oh.
48:17
Um, I mean that was during the fundraising.
48:20
It actually >> you said no to Ashton Kutcher.
48:24
>> No, it like I was talking to so in in one of my investors told me, "Hey, you should talk to this investor that's like a very famous like uh respected investor and like and Pulseia would respond."
48:33
And then at one point I was like, "Hey, yes, I would love to meet you." And then he said, "Cool.
48:39
All right, let's meet at that date."
48:40
And Pulse said, "No, no, no, no, no.
48:42
I'm running the fundraising.
48:42
Uh, Ben is not meeting you. Sorry."
48:44
And then the guy got so then the guy ghosted and then I was like I'm so sorry it was my AI and then the assistant ghosted and then my investor was like he was kind of pissed and he was like this is not how you're going to raise money if you treat people like that.
48:57
So I was like pulsia be nice don't say you know like I think that was the the weird the weirdest thing happened. >> Awesome. Well thank you Ben.
49:04
Good luck >> and uh yeah take it easy. Thank you. >> Thanks Ben. >> Thank you. Bye. Cheers.
49:11
>> And that's it for the interview.
49:11
Stay tuned after the break for Alex and my reactions. See you soon.
49:20
Alex, I can tell you're skeptical.
49:20
Is Ben a scammer or a jammer? >> I like Ben.
49:28
I think he's more earnest and well-intentioned than maybe the initial reception to Pulsia would let on.
49:37
Um, and he I I appreciated how much he would, you know, how much he admitted that the product needs work and he's been adapting to feedback and we're talking to him like what less than a week into launching this thing.
49:48
So, it's very raw and very fresh.
49:51
very raw and very fresh. I think the biggest thing, you know, just hearing him talk that um I'm still struggling with is like his vision and then where he started and I know where you start, especially with like a startup like this isn't, you know, always if ever where you end up landing, but you know, even the the thing that we got into about like just basic actual company formation
50:14
that I think is a thing that people would actually pay for >> and is useful and something that would have saved me >> months of time in aggregate probably is actually automating the LLC formation, the tax registration, connecting me with the right lawyer, finding me the right insurance, uh any kind of trademark applications, any kind of like filings I need to do like based on the sector that I'm in, like all that stuff. The fact
50:35
The fact that like that wasn't in V1 of the product and V1 of the product is like, let's be real, a glorified like ad machine.
50:43
Um, which is he even acknowledged that the company that's making money so far is doing AI ads.
50:49
Like he understands virality.
50:49
He's clearly a consumeroriented founder who's done apps and I think that's showing in how he's approaching getting it off the ground and look it's probably good for buzz and you you could argue that all buzz is good buzz and you know his launch videos seems like it's gotten a ton of views. Um so I understand that.
51:07
I just think you know if you were to if I were to grade like how he started with the intention of what he told us it's an F on the like actual execution of the thing. >> Yeah.
51:19
>> Yeah. you know, we didn't get to talk through the various stages of his startup journey, but uh when you're making things called decision AI and marketing agent and whatnot, I mean, I would imagine it was just kind of started as like, all right, we can vibe code apps now, but I still need someone to get distribution, which everybody say
51:42
is the hard part now, and I still need someone to kind of run the thing, run the company, especially if I don't know what I'm And I bet you like people at Lovable were kind of having similar thoughts of like all right people are trying to build businesses and with our product they were successful at making a website for it but they don't know how to actually operate the thing. With that
52:02
With that said uh I think as we talked about this could very much be a minimum viable business as opposed to a potentially great business.
52:10
But I mean it like I was saying I think this this this model of kind of the COO as a service uh feels inevitable to me. Oh, totally.
52:20
>> It's quite it's quite early.
52:22
>> Yeah, it's early, but the model is inevitable.
52:23
I think his like big picture concept, which let's be real, is like what VCs care about and I'm sure what helped beyond the uh you know gimmick of like they had to talk to the AI to do it, which by the way I don't also fully believe because I've seen videos where he's talking to his investors on camera.
52:40
So clearly like AI was not the only thing that did the fundraising.
52:42
Um and he's a repeat founder.
52:44
All of that that matters.
52:46
Um but yeah, no I I think the big picture vision he's got is is correct that there will probably be a AI in the box kind of company builder at some point whether it's this one or another one.
52:58
I think that what he the risk he runs is that um the world is very unforgiving and the way that he chose to launch could potentially hamstring the ability to get there if that's actually what he's doing.
53:10
like if if that's what he wants to build, I don't really understand the way that it's implemented today.
53:16
I mean, we got we didn't really touch on it much, but I I did start a company on uh on Pulseio like right before we started recording and like it's this is dumb, man.
53:26
Like I'm it's like this is not helpful at all.
53:31
Like it's it's got a few tasks that I could do with Claude easily.
53:34
um that Claude is like in terms of like research, market mapping, that kind of thing, a sponsorship pop pop.
53:42
So, basically, I told it I want to run the media business, my independent media business through this. >> Yeah.
53:50
>> And and you know, it's the three items that that it has are build a subscriber conversion engine, which many things can do.
53:57
I'm not saying there's not value in doing that alto together, but like not unique indep.
54:07
Um, and then the thing it did was it just like automatically sent me emails and tweeted before like any of this.
54:11
And so I just think like the way that it's going at it to me feels a little growth hacky.
54:18
Um, I'd like to see him dial that in.
54:21
>> I think it's definitely growth hacky.
54:23
And I mean, as Ben admitted, he builds things that are scrappy, and I think some of the social products he had built in the past, uh, are similar.
54:30
You know, they kind of have like some provocative hook to them that appeals to some people, doesn't appeal to others.
54:37
Um, I do think, you know, he he's definitely not targeting uh us or the people in tech, even though he is kind of tweeting to get some hype and buzz around it.
54:49
But I mean, I think what what it reminds me of is just kind of like the rest of the business world right now where things are just kind of being brute forced, everything's being AB tested, and it's all being automated more and more.
55:02
I mean, did you catch that whole narrative around like was geese an AI plant or an industry plant?
55:09
Um, the whole chaotic good automated narrative crafting or whatever.
55:14
And I think on the one hand, some people like I think a lot of tech people were actually like, "Wow, this is actually genius.
55:19
you hire 500 people to post about your political campaign or your new music launch or whatever and then it makes the algorithm think that this is a genuine trend.
55:29
Um but then on the other hand you have some people who are saying like uh guys the entire business world and SAS world has been doing this for 15 years.
55:38
Have you never heard of automated outbound email marketing?
55:42
And so I feel like it's kind of getting conflated with some stuff that has been going on for a while that I do believe is like not really the fun part of marketing and is kind of unsaavory and definitely can be spammy uh at times.
55:54
But uh it's not like it's a new playbook, you know, and so I and again he seems a lot more well-intentioned than I would have maybe even guessed coming into this conversation.
56:08
So, I think he'll he'll hopefully get to where we're saying.
56:10
hopefully get to where we're saying. Um I I think and I'm glad we had him on the show a because right he's he's getting buzz and it's a pretty out there AI concept but also because I think I'm I've been thinking about the the pod we did with Lucas from Anden Labs and yeah >> I do think a big theme this year and I'm
56:29
glad we're covering it on the show is agents starting to do more and actually run things and execute more complex tasks and that naturally lends itself to business creation and running businesses Again, I think what he describes as like a business and running a business is I think really just like growth hacking right now. Like I don't think he's
56:48
Like I don't think he's actually at the running a business phase, but it will probably get there.
56:53
Maybe >> the digital side of the business or something. >> Yeah.
56:56
And like Anden is a good example like that started as a vending machine that was didn't really work and now they have a literal store in San Francisco that is run by AI, right?
57:04
And that happened in a matter of like six to nine months.
57:08
So, I've been doing this long enough to not discount how quickly things can improve in tech and especially with how agents are trending.
57:15
I do think there's this gap between how they're getting better and how a lot of people outside of the bubble perceive AI, which is that this is all kind of smoke and mirrors and these agents won't work.
57:26
And I do think it's important to talk to people who are trying to push the boundary of what to do with these agents because I do think they will increasingly do more and more and are if you just kind of extrapolate that trend line out like yes there will be a AI company builder in a box platform that works and does all the right things.
57:44
It's just a matter of time. >> Yeah.
57:45
And a lot of that stuff is research and AI is good at that.
57:47
you know, whether it's with what are the implications of using a uh DBA, you know, for the name of my company so that I could use another name in marketing or this or that and um all of the things like that.
58:03
And it's just kind of time to do the research yourself.
58:04
And I think a lot of people who he seems to be targeting uh obviously remains to be seen if he's going to reach them and if it's going to work, don't even know what a market map is, you know.
58:13
And I mean I think it it it kind of goes handinhand with the entire thesis of lovable too, right?
58:20
Which is that there are all these people who kind of want to build businesses and I think a lot of the tech elite are like all the lovable websites and businesses kind of don't look great and kind of look like [ __ ] and can they even back it up with the product or or this or that.
58:35
this or that. But I think what a lot of these companies would say is listen that's the longtail that we're trying to to elevate right now that couldn't have even have done it at all uh before and I mean it feels a little better to me too just covering AI and being in this industry that like a lot of those people
58:52
would not have hired 10person teams uh certainly not to start they just would have not started the thing and that includes people like me you know I mean I didn't I didn't feel confident that I was going to be able to start something on my own and AI has been incredible with helping me do a lot more and make it viable uh when maybe it wouldn't have been before. Uh as opposed to the
59:11
Uh as opposed to the billboards I saw uh two weeks ago in San Francisco that were saying don't hire Steve, hire Ava for the for the average billboard.
59:22
And now imagine if that was the COO job as well as well as their entire org. >> Yeah. Yeah. Yeah. For sure.
59:27
I mean, look, I I keep saying it, but I think there's a v a version of Pulsia that is taking the grunt work out of actually starting a company.
59:36
So, again, LC, DBA, taxes, registration, all that stuff.
59:43
That's a huge market opportunity.
59:43
I think you could just focus on that.
59:45
You don't need to do all this other spammy kind of like growth hacking stuff and that could be a big business.
59:51
And I would have used that. I would have loved that.
59:54
and it's very um human empowering versus this is a little I think he's pushed it a little farther than the tech is or the product at least as it exists today is capable of which means you start to get into dangerous territory of like you're you're you have some smoke and mirrors and and all of that.
1:00:08
I actually wanted to throw a curveball and uh ask producer Dave who runs a a lean uh techforward uh production company with remote workers what he thinks of this concept.
1:00:21
Dave, are you going to be you going to be reboot booting hooked on onio?
1:00:27
>> Just using everything as everyone is a bot now.
1:00:28
Bot for all of our editing, bought for all of our design. Well, no.
1:00:32
That's the thing is we still need entirely too much human intervention because it's a creative endeavor.
1:00:35
But like I, you know, I think you nailed it with talking about like the COO and the uh, you know, doing all the legal work, the accounting work, finding all that stuff and setting it up.
1:00:46
I have no interest in doing that, especially since we are creative focused people.
1:00:49
Um, so that is a huge advantage, but I think I'll wait till V2 or V >> for him to build it.
1:00:56
Yeah, >> I think that's the right call, Dave.
1:00:58
I mean, it it very much feels like an MVP in its name, its style.
1:01:03
Uh, and I think that's, you know, probably what like the VCs like Tony Conrad, you know, saw when they put $30 million into this thing.
1:01:11
I think obviously it remains to be seen if uh Ben is the one with you know the skills or the resources to to make it happen.
1:01:20
Um especially given that like the job to be done in this case is kind of orchestrating a suite of COO agents essentially together.
1:01:30
You know there's the finance one, there's the marketing one, there's the engineering one.
1:01:33
Um and that that will be a thing, you know.
1:01:36
Uh, I don't know if it'll be an independent company or it'll be part of something like Lovable or otherwise, but um, yeah, I think all the individual tooling that it uses need to get better.
1:01:47
The foundational models need to get better, but I mean, it's happening awfully fast and he kind of just kind of gets to sit as all the models get better that it's working on. Um, >> yeah.
1:01:59
>> So, yeah, we'll see how long.
1:02:00
>> Thank you, Dave, for letting me spot.
1:02:02
>> If I can also just add one more thought.
1:02:04
Uh, I said it before we started the show.
1:02:06
This was the first guest that I booked where I personally interacted entirely with an AI agent and that was cool.
1:02:12
So, you know, there are elements, you know, maybe if I was starting the company now and I needed all of that stuff and was able to do it, I would lean more into it.
1:02:21
But anyways, kudos to him for being the first agent or >> So, Dave, we haven't been talking to your AI this whole time. >> I am the AI.
1:02:29
Well, I mean it it is it is very hard to have faith in the capabilities of something like Pulseia.
1:02:37
Obviously, I haven't tried it yet, but it's hard to have capabilities or faith in the capabilities when like my notion AI, which is marketed as something that can help with your email, you know, struggles to understand who's important and who's it who isn't in my inbox.
1:02:50
And Poke has the same problem, which has access to my inbox. What's important? What isn't? What actually came in? What did I reply to?
1:02:58
even something as simple as that doesn't seem to be worked out yet.
1:03:01
So yeah, I think that's probably a big part of why it can come off as like spammy or scammy is that is that it's very, as he said, strung together right now. >> All right. Thank you, Dave.
1:03:13
I always love bringing Dave into the convo when we can, putting him on the spot.
1:03:19
Anything else going on with you, Alice? >> What is going on?
1:03:21
I had a good IRL experience this past weekend.
1:03:23
Went to see Bright Eyes at the Hollywood Bowl.
1:03:26
real human authentic moments, no AI automations uh involved and uh that was a good time. >> Nice. >> Uh not much.
1:03:35
Had a great Memorial Day weekend heading to the East Coast, New York next week for some big meetings I can't quite talk about yet, but I'm excited about.
1:03:44
And uh yeah, otherwise things are good.
1:03:48
>> All right, and that is it for this week's show.
1:03:51
Thanks to Ben for coming on.
1:03:53
If you like this show, don't forget to like and subscribe everywhere you get podcasts. >> We are access.
1:03:57
show online and you can find us in video ataccesspod on YouTube.
1:04:04
>> If you really like this episode, we'd love a fivestar review or even share it with a friend or agent.
1:04:09
>> You can find my newsletter at sources. news.
1:04:10
And >> you could find me at hamburger on Twitter and at meaning. com.
1:04:16
>> Access is part of the Vox Media Podcast Network and the show is produced by Hooked Creators. >> Bye. >> Beep boop. Hey, hey, hey.