Pablo Solare
Sprightful
Pablo Soto Day, founder and CEO of Spryteful, a tech-enabled solar company, is rethinking how to hire in an AI-first world. He recently moved from Miami to Santa Cruz to be closer to the AI revolution reshaping how he builds teams, manages contractors, and thinks about the future of work. His insight: the best new hires might be people who’ve never used ChatGPT.
When Soto Day moved his family to California, he didn’t move to set up an office. He moved to be near the epicenter of AI development while keeping his team distributed across Portugal, Colombia, and the United States. The conventional thinking says you need to be in an office together to innovate. Soto Day sees it differently.
The value of remote work, he argues, is irreplaceable. You cannot bring the freedom and adventure of remote work into a physical office, but you can approximate the benefits of in-office collaboration through intentional practices online. Google Meet sessions where teams “vibe together” without a formal agenda. Walking meetings that spark ideas. Asynchronous updates that keep everyone informed without forcing synchronous presence.
The freedom and adventure you provide to individual members trump the benefits of in-office work. There’s nothing you can do to bring remote work’s value into an office.
Pablo Soto Day, Spryteful
The tradeoff Soto Day has implemented is three months per year of in-person time spread throughout the season, not consecutive. Three weeks here, a month there. This beats both the all-remote isolation and the all-office overhead. It forces alignment, kills doubt, and maintains the tight-knit feeling that prevents churn.
Installing solar panels requires boots on the ground. Spryteful has tried hiring its own crews and hiring separate labor companies. Now it has landed on a hybrid: four core team members and three contractor teams in the Miami area who handle installations. These teams are usually brothers who’ve been working together for years. Soto Day vets them, tries them, and sticks with the ones who deliver.
The real work isn’t managing calendars. It’s communicating clearly about what needs to happen, when, and then following up on quality. The contractor teams are often not great communicators, so Spryteful has to be the bridge. They learned this the hard way over ten years of iteration.
The tool that changed everything was Telegram. Voice notes, pictures, video updates—hundreds of images a day showing project progress. Asynchronous. No pressure to respond immediately. The teams get flexibility to schedule around graduation events and family needs. And they get to say yes or no to projects, which gives them skin in the game. If they reject one, Spryteful moves to the next team. No hard feelings. For people living paycheck to paycheck, this matters.
We have to be super on the ball with communication: what happens, when it happens, then follow up with quality control.
Pablo Soto Day, Spryteful
Soto Day has always hated CRMs. Now he can chat with his. He uses Monday for project management and CRM, then built an OpenAI agent on a Vercel server that watches the Monday database constantly. When a deadline is approaching and hasn’t been updated, the agent sends a Telegram message asking the team what’s happened. When they respond, the agent updates the CRM automatically.
This is the beginning of what Soto Day thinks will become the dominant pattern: one human managing thousands of agents. He predicts that within three to four years, the ratio will be one human for every 10,000 agents running inside a typical organization.
To move from talking about agents to building them, Soto Day hired his brother, a Georgia Tech computer engineer, two months ago. Now when Soto Day has an idea about agent-human workflows, his brother implements it in a few days. They test. Sometimes it works. Sometimes they throw it out. The iteration speed is what matters. Within the solar industry, Soto Day and his brother are building capabilities that 99.99% of solar companies don’t have.
The technology is improving so quickly that people get stuck with what LLMs could do a year ago. The difference is so dramatic.
Pablo Soto Day, Spryteful
Here’s where Soto Day breaks from conventional AI hiring wisdom. When bringing someone onto Spryteful today, after five months of building agents and AI-first processes, he doesn’t require AI fluency. He requires high agency: the self-motivating drive to see things through without getting stuck asking for help.
If someone is smart, motivated, and has never heard of ChatGPT, Spryteful can teach them in a month. They’ll see what the tools do inside the company. Their creativity juices will flow. And in some ways, this is better than hiring someone already locked into how LLMs worked last year.
Soto Day compares it to humans discovering fire. The wonder and possibility are still there. Someone who hasn’t been trapped in the limitations of yesterday’s models will imagine better uses for today’s tools.
Someone hasn’t ever heard of ChatGPT, within a month we could have them using a lot of these tools. If they see what we’re using them for, their creativity juices start flowing.
Pablo Soto Day, Spryteful
Soto Day doesn’t think humans disappear from the solar business. He’s planning to hire people to open the California market, to study how to enter new states, to eventually expand beyond the 99% of Spryteful’s work that now happens in Miami. These are roles where a human has to look at chaotic market conditions, understand sentiment, spot opportunity, and have the willpower to spend what’s needed to win. LLMs hallucinate too much for that kind of high-stakes decision-making.
Robots will install solar alongside humans within five to seven years. An autonomous van arrives with four or five robots. A human shows up, confirms everything is working, lets them operate. The human is there for liability, for judgment calls, for taking action when something goes wrong. You can’t yell at a robot and expect it to understand.
But Soto Day sees robots as empowering the contractor teams he works with now, not replacing them. Better equipment, less back-breaking work, more jobs per day, new skills learned. When he talks about the future, he talks about giving his installation crews Teslas with full self-driving and Starlink connections in the back so they can work from anywhere while robots do the heavy lifting.
I’m excited to empower these teams with robots because if we show them how powerful they are, it’s only going to improve their ability to execute.
Pablo Soto Day, Spryteful
For founders who don’t have a CTO, Soto Day’s advice is direct: learn Claude, learn the coding assistants, learn the AI tools of the moment. Don’t wait for permission. Don’t wait to understand everything. Pick one tool, watch a few YouTube videos, start integrating it into your workflow and file system. Give it access to your ideas and content. The privacy tradeoff is real, but the speed gain is real too.
The tools are improving every week. The only way to stay ahead is to be using them. Get the skeleton of what you’re building out fast using AI. You’ll do the final brushstrokes yourself, but at least you’re not staring at a blank page for months. And Soto Day’s team recently moved from ChatGPT to Claude and gave everyone Claude Pro subscriptions because he believes Claude has a stronger ethical foundation. He’s running AI days to make sure the whole company stays current.
The alternative to learning these tools is being left behind by founders who do. That’s the real risk. Not that AI is dangerous, but that you didn’t learn to use it.
I think that the ratio between humans and agents is only going to grow exponentially. It would surprise me if in three to four years the ratio isn't close to like one human for every 10,000 agents running within an organization.
It's a little harder to hire now because we're like five months into this project. If you're going to bring in somebody from Spryteful now and you've had two years of AI agents running of bought usage, are you just not going to bring anybody who's not AI fluent on the team or you're willing to train them? Where do you draw the line?
I think that they're like AI fluent, they have to be exceptional in some other way. I would be okay with them not having super strong AI knowledge, mainly because it's not that difficult to bring them up to speed. Because if they see what we're using the tools for within the company, the creativity juices will start flowing.
Yeah.
In many ways it might be better, mainly because the technology is improving so quickly. We see it time and time again that people kind of get stuck with like whatever the LLMs could do a year ago.
It's definitely rewired how I'm thinking about it. Like the way I'm thinking about it is like humans discovering fire. It's woah, like what is happening here? I can do so much with this. I can light things, I can cook. That element is now AI.
I'm Louis Cent and this is Who We Hire. Today's guest is Pablo Solar Day. Pablo, founder and CEO of Spryteful, a tech-enabled solar company with an ambition to enable the world to harness the power of the sun. Pablo, thanks for joining us on Who We Hire.
It's a pleasure. I love you, man. You know that. I I'm I'm always excited to talk to you and as I was just mentioning before the recording started, I think we should just record every conversation we have, you and me, cuz there's always some sort of golden nugget of information there just for us. But obviously it's great to make it public too and and see, you know, if there's any any anyone else who could possibly benefit from hearing our crazy ideas.
I love you, too. Thank you so much for being here. I I think we should have recorded our entire trip in in San Francisco cruising in in your Tesla along the streets there cuz that that would have definitely made a made for a good show. But I'm really excited to chat because I've been following your journey for a decade now and I mean you invented products in the solar market that didn't exist before. You launched them into the Florida market with great success. Now 10 years later you packed your Teslas and I think the one you're in right now and uh yep, you drove across the country with your family, your one-year-old friends. Now you're in Santa Cruz and like one of the big reasons was to be near San Francisco, near this AI tech and revolution. Can you tell me a little bit about that and how you know, how how that's inspiring how you're building teams and and how you're thinking about work today.
I think primarily the the excitement I I I don't think that our generation has really come close to technological revolution within anywhere near the scale of what this AI revolution is is is showcasing already with even it's not it's not even close to being done yet and it's already changing absolutely everything. And there aren't that many places that have the same density of people who are thinking about how to how this applies to not just work, but life in general moving forward because it is absolutely going to change the way we live. And it's nice to be in a place where, you know, you can strike up a conversation with the person who's waiting for caught like in the coffee line right next to you about what they think this is about and and to hear their thoughts. And that's sort of spontaneous serendipitous conversation is what really excited me to move to Santa Cruz and the bay, you know, it's technically adjacent to the Bay Area, but but it is pretty Bay Area-ish. You know, there's a lot of Bay Area culture here as well.
Well, yeah, when you can turn your your Tesla into a co-working space and cruise over to the Bay, I consider anything in a 2-hour radius now the the Bay Area. So, does it not conflict a little bit with you know, your your remote work when you're not installing on a house, you're working remote, you have a team, you know, that's in Portugal, Colombia, like all over the world. And like that's an I typically hear that as an argument for in-office work, right? So, you you need to go in office so you can have these four spontaneous moments. Does that not go against a little bit of what, you know, your policy is for working?
Yes, you could say that. I I also you know, there's no there's no like clear winner in the whole remote work versus office work debate. Like it's absolutely clear that there's benefit to being in the office with a lot of people who are working in the on the same problems, and there's there's so much benefit to like efficiency and and inspiration and alignment with that. However, my perspective continues to be that the freedom and the adventure that can be provided to the individual members of the organization are they trump the benefits that come from in-office work. I've seen that time and time again with friends, with employees, with myself. There's It's It's really hard. You You can You can mimic and you can approach the benefits of being in the office with different you know, you can you can you can you can do even even if you do like a Google Meet session where nobody's necessarily in a meeting, but you're just kind of vibing together, you're working together, and you can you can easily like shout out someone I'm "Hillary, have you thought about this? And you know, if you're around, you can answer." That That kind of mimicry can exist for remote work. You cannot There's nothing you can do to bring the value of remote work into the office. So, from that perspective, if I if I extrapolate that towards the future, it makes me more excited about remote work in general. Like, we'll always be able to improve remote work. Hard to improve in-office work.
I love that answer cuz I think I can ask that question a thousand times and I'll never hear adventure in it. And it you know, I think I think that's an important part of your ethos and how you build your companies is because it's, you know, your culture is a little bit of who you are and your adventure. So, you could have simply just said that and it would have it would have answered it.
And adventure is extremely important to I think everyone, right? That sense of uncertainty and freedom is is a remarkable essence of humanity. And that's one of the reasons why I all Another one of the reasons why I moved to California is because there's a lot more of that nat- nature that's accessible within like a 4-hour drive here than from Miami. Miami, love Miami, and I still consider Miami home. However, it's hard to go adventure in the Everglades, if you know what I mean. Yeah, you basically any any Miami adventure more than likely is going to take uh a boat ride, which is, you know, some people like that.
Um I'm in Puerto Rico right now, so the the weather is very similar to to Miami. If we're not in the water, you're you don't want to be outside essentially, where in California, you have the hikes and the waterfalls and there's so many weekend trips you can do through through that. So, that's that's great that you're taking advantage of that and enjoying it. So, on you know, I was in San Francisco a couple weeks ago, we got a chance to meet up. And you said something interesting I really I think resonated with some of my team way too much, where you said that yeah, I need one of our leaders. They they need to spend 3 months with me, you know, any point throughout the year." And that kind of forces that, you know, spontaneity, like forced, you know, interactions that you really, you know, double down on those 3 months on on work. How, you know, how did how did that come to fruition? If it was that, like, plan, like, "Hey, we need to do this every year and and you're coming." Or, you know, how how how would you encourage that for a founder looking to build a team that's fully remote, but still wants to get them together every year?
One of the things that I hope happens as remote work becomes more widespread is that that freedom leads to a desire to experience different geographical places, right? So, if you have a remote worker who's based out of Lisbon, like I do, I hope that he or she has the willpower and the and and desire to go spend some time out of the year somewhere else. So, whether that's, I don't know, Thailand or San Francisco or Miami, I think that that those things come hand in hand. The people There's There's There's the Venn diagram between There's a lot of overlap in the Venn diagram between remote workers and people who like to travel.
And And so, if you take that idea and and then directly apply it to how it can benefit work then coming together for 3 months out of the year appears to be, at least from my experiments the the ideal balance. And that doesn't mean it's 3 months straight, it means 3 months spread throughout the entire year. So, it could be 3 weeks here, a month here, 2 months, then a month later. Whatever you can make work, those 3 months, which is a quarter of the year, being together are essential for alignment because that's the only thing and and for really like the alignment is mainly there to limit churn, make sure that there's still a tight-knit feeling between the team. But also to align goals and express doubt if needed and and then extinguish that doubt. That's so so important. Yeah, I mean I I think 3 months is is is how I we've defined that's the sweet spot for us. I don't know we're a small team. So there's only four of us that makes it a lot easier. If there were more of us could get a little bit more difficult and that's where you starting getting into the whole company outing, you know, where you go spend a week or so, you know, doing something special together as like a group of 40 or 50 or whatever. I think in in even bigger companies it's possible. It's just that each individual company unit would probably have to do their own thing just to maintain that flexibility. Hard to get hundreds of people to align on like where to go for, you know, a specific amount of time. I mean destination weddings are the clear example of how how difficult that can be. You have to plan it years in advance for people to to be able to go to those. So yeah it's it it's it's interesting that you mentioned the whole 3-month thing because I don't really think about it much and it feels so natural to me. I love spending 3 months out of the year with my team. I love hosting my team at my house or nearby. It's it's special and I think in the future that's only going to become more ingrained into the Spryteful culture.
I love that. That's that does sound magical especially when you talk about hosting them, right? You're inviting them into your world. They see the human element beside it and you you you make a good point in terms of retention, right? It we we need to meet each other and communicate and and talk through things and it's you know it is hard to do that on video sometimes where if you're intentional about it and you're seeing them face-to-face and for me one of my favorite things is to do walking meetings, right? If I if I figure out good studio quality here, I'll be doing these podcasts walking. Where I speak the best, it's where I get the best ideas. So I think you know, if if and when we met our team in San Francisco, we were walking you know, we were doing meetings walking around the the city and for us that was no it was absolutely magical and you could definitely see you know, that that there's a because you're putting in the effort they're also going to go and put in more effort. So I you know, I'm going to borrow some of those ideas from you and implement them into our culture. You mentioned your your four people now you're definitely punching above your weight in in terms of like what the company looks like from the outside but you also have to hire a ton of contractors for different jobs. You have to organize not only you know, but most of the founders that we interviewed they have purely software company. They don't have to do anything physical. You do. How does that you know, how do you change when you're hiring somebody fully remote or somebody who needs to show up at a certain time and finish a job like you're you're smiling because
I'm smiling I'm smiling because this has been such a fun element to to figure out for for us cuz we've been around for 10 years. We've tried it all. We've tried hiring our own you know, physical labor crews. We've tried hiring separate companies that handle the labor portion and they hired and then they have their own teams and I think we've landed into a really really wonderful sweet sweet spot once again. The sweet finding those sweet spots is just so nice. We the way we see it now is we are four people and we have a lot of experience and know-how and what that leads us to be able to do is to hire teams for projects. So, we we currently hire in in the Miami area, which is our primary place where we install solar, we have three teams. And these three teams have been working with us for a couple of years now. Depending Now, we know depending on the kind of project, we will know which team to hire. We will know how to communicate with them because every team has a different communication structure and and and uh uh The I I would I would say I would say more than likely these people are not great communicators. The the the the teams of people who are doing labor are unlikely to have the best communication skills. So, that's where we have to come in and supplement. So, we have to be super on the ball when it comes to communicating about what's going to happen, when it's going to happen, and then follow up with the quality control. So, we That's the part that we've really figured out. We've got teams of Usually, it's brothers, surprisingly. Teams of brothers who uh work together installing solar systems in South Florida. We vet them. We try them out. And the ones that we really like, we stick with. And we just have to It's a It's work The work The real work that we have to do is like keep tabs on them and uh sometimes ask five times before we get an answer like, "Hey, how many strings did did this system uh end up getting installed? And like, what what are the terminations like?" And all these things. And like, the the We have to know what language they speak. Meet them where they are. Usually, that's the voice notes over Telegram uh or video. Oh, yeah, Telegram has become our favorite communication platform. Telegram is ul- ultimately, that that asynchronous communication with with the teams, where we can get hundreds of pictures in a day showing how the project is going so we can check the quality and ask questions if we need to. And at the same time, give the teams the freedom they need to schedule it how they prefer. You know, this this week is an example for example. This is the great example. We have two of our crews are running with half of their guys because there's so many graduation events happening right now. All their kids are going from like eighth grade to ninth grade or you know, graduating from high school. So, we have to be able to give them that flexibility because again, just as just as with the remote work conversation, my perspective, my intuition is that to to win long-term, we need to give the individuals the freedom that they need to live their lives how they want to live it. And and that's another thing is that these teams obviously when we send them a project, they get to say yes, you know, they give us a thumbs up. We want to do this. And if they say no, then we just we choose one of the other teams. But in general, that gives them a little bit of skin in the game. Like they have they have said yes to this project. They didn't just get assigned this project. It's not like a like a like a team of workers who has to do what they what they're sent to do. They they have a a little bit of they have a little bit of a free will element which I mean, isn't that what we're all after? A little bit of free way.
Sure. For sure. I mean, and you've it sounds like you've gamified it in a way where like you have first option here, you know, within 24 hours let us know. If not, no worries. You know, we'll we'll give you the next one.
Yep. Exactly. No no hard feelings either. It's like you I we understand you don't have you don't have to take all the projects we send you. Depends on how hungry you are right now. A lot of these guys live paycheck to paycheck. Like they they I I have to admit like it's very rare that they reject reject a project because they they want more. They want more. And and so that's that's how I see it is me and my team, our goal is to feed them as many projects as we can, make sure that the customer stay happy understanding how everything is going throughout. That that's primary the the the pri- our primary goal is to make sure that things happen when we say they're going to happen and that we communicate when things go a little bit off from the plan.
And you're achieving this. If we go to your reviews online, Pablo Andrea is there so many five-star remarks in the communication and the warranty and the service. And it's amazing to me that you're able to, you know, go and do all this remotely. You really have to be intentional with the communication and over communicate.
Has has AI help you know, create processes for this or automate any of this?
A lot. Yeah, we we are we are tapping into AI as much as we can. It seems like every day we come up with a new use case for how we can take advantage of this new technology internally, externally. In general, the like for example, we I've always hated CRMs. I've never been able to like keep tabs with like all the stuff. I'm really excited to say that now I can officially chat with my CRM. And and it'll try and it'll respond back with like
What's your CRM?
insightful. Well, right now we're using Monday, a platform that I think is mainly for project management, but we've it has a CRM component.
Wait a minute. You chat with Monday now?
Well, yeah, I mean it the the I think the tech stack that we're currently using is Monday is where the information lives and we can go edit it edit it manually if we want to. Then we have an open cloud living on a Vercel server, I believe, that is always looking at the Monday database and asking us questions. So for example, if it knows that there's a deadline coming up for a specific component of a project, but we haven't updated it yet, then it'll ask it'll send a a text message to the whole group on Telegram. And and it'll ask like, "Hey, has this happened yet?" And then when we respond, it's able to go back to that same CRM and update the information. So, it's nice to have It's nice to have a Yeah, I mean, I would I would argue that we have a hundreds. You know, it's it's that that's ultimately what I feel like is happening is that we are gaining so much in in just eight with the agents that are coming that my favorite thing right now is that I was able to hire my brother who is helping us. He's He's a computer engineer from Georgia Tech. Extreme brilliant guy. Very motivated to just do good work. And it's really exciting to see. He's He's We We We brought him on on the onto the team 2 months ago, which was a miracle. It was It almost seemed like the stars aligned and we were able to bring him to the team. But bringing him on was amazing because it a lot of the ideas that I have regarding this agent agent human relationship, I don't have the technical capabilities to bring them online. But he does. So, I can throw an idea his way and and and he can implement it in a couple of days and then we test it out and we're like, "Yeah, this works." A lot of the times we're like, "Yeah, no, it didn't quite work out the way we expected it." You know, throw it out. That's fine. You know, we just we keep going. It's It's It's nice to be able to iterate quickly on with these with these agents.
Yeah, I mean, having an open claw and a CTO, I think you're you know, ahead of 99.99% of all all solar companies in the in the world. I mean, that's one of our goals this year, right? Is to have agents running whether it's on open claw or, you know, a more economical token robot. Like, what are some of the thing Like, one I do think you're technical enough to go and do this. It's just going to take you longer than, you know, he's going to be able to do it in a few minutes versus you're going to have to spend all day on YouTube figuring figuring this out. There's definitely some learning curves to it, but what you know, in addition, like what are some agents you're seeing in the future, short-term, long-term that could be employees of your company? I mean, we we were talking, you know, 10 years less than 10 years ahead where well, I guess we were we were debating how long it would be until robots are installing solar. So, I want to hear like what does that the next 5 years look like? Are you hiring any more humans? Where do humans, you know, have a place in your company versus is it always AI first and if AI can't do this, then I'm going human? What How do you think about that?
Well, I think the human is always going to be a part of of the building value experience at the at that intersection between other humans, right? So, like every company is operating in some sort of space where it is bumping shoulders and it is interacting with other humans, whether it's the providers for the manufacturers that we work with or the distributors that we buy stuff from or obviously our our clients who are real human beings and they don't want to necessarily buy buy things directly from a robot, even though some might, but we're not we're not quite there yet. I would say human human hiring exists for those for those components of the company. So, I'm excited, for example, to bring on board people who can help me open up the market in California or other states and and and help me just spend time looking at how we want to enter these markets as Miami Miami is 99% of what we do, Miami and the and the surrounding areas. But obviously we have plans to grow and and eventually provide our expertise nationwide. So, because I'm in California and because the California market excites me a lot, it's extremely The weather is really it lends itself really well for solar here and the electricity rates are insane. I would say five times higher than that. Nobody else smiles at
Well, I got it because think about it. We got a lot of sun and crazy electricity rates. This is amazing. The utility companies are a little bit they've they've closed the doors on a lot of the things that made solar super attractive back in the day, but it's still very attractive. It's just that the the game has changed. Long story short is I am really excited about how like where the solar world is right now because it is so chaotic. And and so humans are really that that that element that brings that chaos into view. I think that the agents, the LLMs, are great tools that we can use right now, but there's I don't see a way for for a an LLM currently, and I say currently because in 6 months it might be different, for an LLM to be able to look at all the data that they're receiving, all the context of the industry, the the general sentiment, and understand how to go to market somewhere. You know, like how do how do we tackle how do we best tackle this market? And then and then have the willpower to say let's do it and spend, you know, whatever it needs to be spent in order to to tackle that market. Yeah. So, it's
You don't think you can do that right now with like three open claw agents?
it's great I think it's great at refining refining plans, but there's just a little bit too much hallucination going on. And even one tiny detail that it gets wrong could go very wrong. One tiny detail. So, so we still need the humans. I mean, I guess I I'll say I'll say it this way. I think that the ratio between humans and agents is only going to grow exponentially. It would surprise me if in three to three to five three to four years the ratio isn't close to like one human for every 10,000 agents running within an organization. You know, cuz an agent could be doing very small details or could be doing very big details, but it's it's it's easy to see how it's only going to be empowering to the humans to just be able to issue out agents for whatever it is that they want. Could be could be as simple as like, "Hey, you know, you're in charge of making sure that my house never runs out of milk." Like that that sort of that that could be an agent right there. And then let that agent's job is just to make sure that the house always has milk. I don't know how it's going to do it, but they can do it. Same thing within an within a within a company. It's going to be a lot easier to to run. And I know there's a lot of talk on Twitter about when the net when the first one-person billion-dollar company is going to happen. I have to agree it's going to happen. You know, what it was going to be
I I heard something about this these people selling something two two brothers selling something on on Yeah. Something about like nutrition or Yeah.
But I I don't know. I I looked a little bit I looked I looked into it a little bit and I I'm not convinced that that I mean it is
a health startup.
Yeah, you're not convinced.
Yeah, I'm not I'm not convinced that it's real. I I want to see the Stripe screenshots, you know, I I think uh I think we're still at the level where it's easy to to do to say these kinds of things for pure clout or marketing, but very soon I'm not surprised I wouldn't be surprised if sometime later this year, 2026, we will see a single individual find some sort of niche and it just kind of that explodes into a billion-dollar market sheet market cap quickly. Uh so, it's it's it's going to be interesting. I think I think I'm excited to to see it happen. The opportunities are endless with with all these agents. And that's honestly like one of the reasons why I love remote work because agents also help the remote work element to keep everyone connected, make sure that information doesn't get lost, and and ultimately to to uh perform high value work without the chain cuz really it is a chain of being tied up to either a single geographical location inside a physical space, an office. So, yeah, I I want more of my people to I will say this, I want them all to have full self-driving Teslas. Like if it could be I would much rather have employees that are all have full self-driving Teslas because those are soon going to be unsupervised self-driving and they can be on the road just like I am sometimes working. Yeah, I'm excited about that. I'd rather I'd rather spend CapEx on that versus maintaining an office space, you know what I mean?
I would I would I would Let's say your Tesla is coming. It's uh
Yeah, I mean look, I will do my best to get every every Spriteful employee Tesla with full self-driving and a Starlink in the back like I do right here, you know?
Is that a Starlink mini? Is that what I'm seeing?
Yeah, a little Starlink mini right in the back. It's important to you know, like sometimes cell cell service is pretty janky but you you know, if you're in a car more than likely you have your view of the sky. So,
have a backup with the with the podcast here if we had any issues. This is amazing.
Yeah, exactly. Yeah, I don't I don't see I have I have strong cell signal at the a strong strong Starlink signal at the moment, I'll say that.
So, for somebody who doesn't have a CTO and wants to start looking at agents and thinking about how to level up their company, what would you recommend to them?
I mean, the I think for anybody who wants to do anything, the first the first most important thing is learn how to harness Claude codes of the world. So, you pick your favorite, you know, you can you can start with Claude code work. See how See watch a couple YouTube videos, see how people are using that to integrate it into their file system and giving it access to you know, the your your ideas, your content. Giving it Ideally, you know, there there are some privacy concerns, but everyone's limit is different. Just get used to these tools because they're not going anywhere. They're only going to get better and better and better. So, Claude co-work, Claude code, OpenAI's Codex, really powerful. They're probably going to level up quite significantly in the next coming months as well because these these tools are are getting enhanced dramatically every week. So, there's nothing that you could do that could improve your likelihood of success at an individual let's call it center of ideas with a plan or motivation, passion, whatever it is, than empowering yourself with these tools. Nothing. And um I I I think there's some people who are still caught up with the idea of, you know, I want to do this myself. I I I want to I want to see how you know, I want to be the one doing the final brush strokes on this, whatever it is. Sure. You're going to do the final brush strokes, but I mean, at least for ideation, just to like get get yourself the the let's say the the skeleton of what it whatever it is you're trying to build, there's no quicker way to get that skeleton than using these tools. So, learn how to use the tools, stay up-to-date with uh what's going on in that world.
That's what we do. And
Do it. Just start doing it. Just uh get pick pick one, get in it. I mean, we have
to trade our privacy for convenience. I think that's where uh Claude is a it From what I read, it's some more ethics involved in in the clauses, though. We, you know, we recently moved the team over from ChatGPT over to Claude, and you know, everybody gets a Claude Pro subscription, which they're excited about when they join. And you know, we're doing AI days, where we're really pushing for it, but it's uh it it's interesting to see. And like when when I, you know, talk to our customers and our partners and even ourselves, it's so much it's a little harder to hire now because we're like 5 months into this project, right? And then we have to bring in somebody from the from the very beginning. What are some ideas on that, right? So, if you're going to bring in somebody from Spryful now and, you know, you've had 2 years of of AI agents running, of Claude usage, are you just not going to bring anybody who's not AI fluent on the team or you're willing to train them? Like, you know, where do you draw the line?
I think that the AI if they're not AI fluent, they have to be exceptional in some other way. I mean, obviously like if if it's someone extremely charismatic, trustworthy, high agency, I think would be my my critical element for anybody that I bring onto the team, which basically means that they have a self-motivating desire or need, drive to just see things through and not get stuck and have to, you know, ask for help too quickly. Just someone who is driven, I would I would be okay with them not having super strong AI knowledge, mainly because it's not that difficult to bring them up to speed. So, I would imagine somebody who hasn't ever heard of ChatGPT, within a month we could have them using a lot of these tools, mainly because if they see what we're using the tools for in the in in in within within the company, the creativity juices will start flowing and and they'll and and in many ways it might be better. Yeah, it might be better to hire someone who doesn't have that much AI fluency, mainly because the technology is improving so quickly that we see it we see it time and time again that people kind of get stuck with like whatever the LLMs could do a year ago. And the difference is so dramatic between what they could do a year ago versus now that I'm kind of curious. I would I would love to be able to meet someone who's super high agency, intelligent capable and doesn't know anything about ChatGPT or any of the LLMs, and watch the watch their eyes grow big as we explain how these tools work, and and and ask them to to you know, use them. I'm sure that they would come with like they would they would come up with like really interesting ideas about how to how to use them.
I love that attitude and perspective towards it. I it definitely rewired how I'm thinking about it. Like the way I'm thinking about it is like humans discovering fire. Right? It's woah, like what is happening here? I can do so much with this. I can light things, I can cook, I can hunt. It's a magical element, and that element is now AI.
Yeah. Yeah, I mean, fire 1.0 was like, you know, you have to kind of wait for a lightning strike to to hit, and then you had you had to carry that that fire from one place to another, protect it because it would go out, and then you wouldn't know how to make it again. Fire 2.0 is when they learned how to make it from scratch, you know, rubbing the two sticks together, and then and then they they started making fire by themselves, you know, and you take that same technology all the way through, and then you get combustion engines and nuclear bombs. And you know, I think that AI is likely at the fire 2.0 or 3.0 level. We have no idea what the nuclear bomb version of this looks like. I'm hoping it's not as uh world-ending, catastr- catastrophic, but we're heading there, you know, like this is It is It is like I I agree with you. Like we are experiencing a new technology that is as much of a game-changer as fire was for humanity.
I I love it. We're going to that that may be a good good visual for for the podcast here. Pablo working with fire sun and fire.
Yeah, I'm excited to I mean I I feel extremely empowered by by this new technology cuz like as a as an entrepreneur the ideas are endless and it's nice to feel like my ability to execute is improving as it only improves and and it's nice to to see how the people around me are getting empowered all the way from my my mom, my obviously within my company all the people in my company, my friends, my wife and obviously my my daughter Amari is 18 months old. She's going to grow up in a world where this has existed from the beginning. She's younger than ChatGPT. What does that look like for a human being to to to never have experienced existence without these magical rocks that help us think? I'm excited. Like I'm I'm optimistic in general. Not a doomer at all. I I have to admit that it's it's it's fun to think about how things can go wrong too and I'm glad that there are a lot of people spending time thinking about how this can go wrong. But in general, extremely excited about hiring more agents. Extremely excited about hiring more humans that can have thousands of agents and benefit from from the increased ability to get things done.
You you and me both and I I've always seen you very passionate about what you do full of ideas but like what what you said there you know, you're able to execute a lot faster which I can see like your passion is reinvigorated. It's maybe Pablo like 7.0 at this point. You can run through a brick wall with with these tools which is really exciting to you know, see when I when I talk to you when I when I see what what your ambitions are. You know, before we wrap I do want to talk robots. You're in you know, what one of the you're in the Bay Area, you know, we're seeing robots everywhere. We have robo taxis, we have Waymos, robots, you know, delivering food in the street. We have light, you know, bedside robots folding laundry. You were you know, talking about robots installing solar and you know, where where do you see robots in your world? Would you hire one to be in your house? Would you you know, hire one for your company? Like how how are you thinking about that in the next few years?
I I love that that that we're about to experience this robot revolution. I think robots are going to be extremely handy. I see them as as like a set of helping hands. And so I imagine you know, 5, 6, 7 years from now when when these become mass manufactured and and easy to access and they've got great cameras and a great, you know, little robot brain so they can they can maneuver the world around them. I think it's going to be extremely useful to have them install solar alongside humans. So I I would imagine for at least the first 10 years of this technology, a human kind of shepherding a flock of robots is is very likely to be somewhere where somewhere where we're heading. Where there's a you know, like let's say that a an autonomous van arrives with you know, a cup five four four five different robots. They start working. A human comes along, confirms everything's working well for the solar installation and and lets them be. Primarily like you know, if if if this technology allows us as a company to install solar better, to make sure that quality control is adhered to better, then it's something we're going to we're going to look forward to. It's It's you know, going back to that the groups of the groups of of teams that we currently hire, I'm excited to empower these teams with robots because more than likely they won't take up this technology by themselves, but if we if we show them just like we show people right now how to work with agents, if we show the teams that we currently work with how powerful these robots are, it's only going to improve their ability to execute.
Yeah, and they do more jobs for the day. They're not breaking their back. They're They're learning something new and and I love the way you think about that. It's empowering humans with AI with robots. It's not about replacing them. It's empowering.
If I can if I can have my current guys who are working for us, you know, installing solar right now up on the roof in a self-driving car in computer like right outside watching the robots do the work, making sure everything's going well, I'll I'll take that. I'll take that as as a as a future as an ideal future. I think the the humans in the loop are going to be amazing for obviously making sure that there's an a responsible person who is nearby to take action if things go awry, if there's a robot insurrection. Just kidding. But if but also just to make sure that there's someone who can be considered responsible. Like if something if something does go wrong and cuz you can't just point at robot, you know, XZ1B and be like, "What did you do, robot XZ1B? You were supposed to, you know, use this other solar panel, not this one." What's the robot going to say? "Sorry, sir." Like, no. We need a human being who is making sure that the robots are doing the appropriate work. And and then they can respond quickly if something needs to needs needs action.
You don't want to humanize your robots or agents? Are they Are they really like ABC or are you giving them names and putting them, you know, giving them pictures and
Well, it's enticing. It's enticing to to give them names. Um We're I don't think we're there just yet. You know, like these It's It's still important It's still important to recognize that these are not conscious beings and and they have to be I mean, at least from my perspective, we have to recognize that they can be killed and reborn at any time and they don't suffer and we should not suffer for their deaths and rebirths. They are These are These are instances of neural networks that exist. Love them. Humanizing them, giving them names, sure. Getting too attached to to, let's say, I don't know. Like, for example, I think it's really interesting what's happened with GPT-40. I don't know if you've seen this. There's a massive group of people who had created these personas and emotional relationships with GPT-40 from ChatGPT, from OpenAI. And once they killed that model, there's been like revolt. Like, everyone was asking Chad OpenAI to bring back 40, but obviously it's not in OpenAI's best interest because they're using that compute for their more efficient newer models. Long story short is there's a lot of There's a lot of nuance in in getting emotionally attached to these to these creations. I I would say I would say fight fight the urge to become super emotionally connected to them.
A very fair warning. We'll We'll wrap up on that high note. Pablos, thank you so much for being on. Where can people find you? And your work?
Well, I I think I think the the best place to find me is brightful.com. Lately, I haven't really really been using I haven't been using social media that much, but I am pab.sun on Instagram. I I do check my messages there as well. I'm on Twitter, pab.sun. I don't use Twitter that much, but but but if you want to reach out to me there, I'm there as well. And uh yeah, I would love to hear any thoughts, any ideas that that anybody has about any of these topics. So, I'm open.
Bob Sun, thanks for being on Who We Hire.
Yeah, thank you for having me.