Lead With AI
15
min read

Building AI-First Companies Without a Tech Background

Peek CEO Sherry Jiang shares how non-technical leaders can leverage AI to build AI-first companies, rethink SaaS, and drive team productivity.
Published:
September 10, 2024
Last updated:
September 14, 2024 23:03
building-ai-first-companies-without-a-tech-background

🎧 Listen Now:

In today’s episode, we speak to Sherry Jiang, CEO of Peek, about how non-technical leaders can successfully build AI-first companies.

Sherry Jiang is an entrepreneur and tech innovator. After honing her skills as a Product Marketing Manager for Google Pay, Sherry set her sights on revolutionizing personal financial management. 

Now, as the Co-Founder and CEO of Peek, she's leveraging cutting-edge AI technology to help individuals track, manage, and grow their net worth with unprecedented ease and insight.

In this episode, you'll learn how Sherry leverages AI to bridge the gap between technical and non-technical roles, rethinks traditional SaaS models, and tackles the challenges of AI adoption in larger organizations. We’ll also dive into how AI is helping her optimize team productivity and streamline innovation.

Key Insights from Sherry Jiang

Here are the actionable key takeaways from the conversation:

1. Non-technical Founders Are Supercharged with AI

Sherry, a non-technical founder, used AI tools to bridge the gap between her non-technical background and the demands of developing a technical product.
She discussed how AI tools like Cursor and Streamlit enabled her to build functional prototypes quickly and collaborate more effectively with technical teams while also giving them the space to work on the core roadmap.
AI can level the playing field for non-technical founders, enabling them to participate actively in the technical aspects of product development and more effectively bring their visions to life.
And I think this applies more broadly – all of us now have the tools and abilities to develop ideas and maybe even apps.

2. The Future of SaaS May Shift from Centralized Platforms to DIY Solutions

This ability to quickly spin up apps also makes Sherry believe that demand for traditional SaaS products, particularly those with low barriers to entry, could decline.
She suggested that, thanks to AI's capabilities, founders and small teams might increasingly build their own tools rather than subscribing to a multitude of SaaS products.
Businesses offering SaaS products should be aware of this potential shift and consider how they can add value beyond what can be easily replicated with AI, perhaps by focusing on specialized services, customer support, or integrations.

For everyone else, look for opportunities to cut your software bills! 

3. Resistance to AI Adoption in Large Organizations Stems from Incentive Misalignment

Sherry noted the resistance to AI adoption in larger organizations, even in tech companies, and even among younger employees who should theoretically be more open to it.
She discussed how employees might fear that automating their tasks could lead to job loss, and how the lack of immediate, personal incentives diminishes the appeal of AI tools like ChatGPT.
Contrast this with Sherry, who’s highly motivated and can inspire the team to feel the same way.
For AI to be embraced in larger organizations, leaders must align incentives with AI adoption, ensuring employees see clear personal and professional benefits from increasing their productivity with AI.
Leaders should also follow Sherry’s leading by example, and asking employees to submit work to AI before they send it to their manager.

4. AI's Current Limitations: Confidence vs. Competence

Sherry pointed out the limitations of AI, especially in tasks requiring deep analysis and nuanced judgment.
She compared AI’s outputs to those of a confident yet inexperienced junior employee, noting that AI often lacks the depth of understanding needed for complex decision-making.
While AI is powerful for automating routine tasks, leaders should exercise caution when relying on it for high-stakes decisions requiring expert judgment.
Finally, Sherry also underscored what Matt Kropp from BCG X said: businesses should focus on using AI to remove toil, and increase joy at work.

Learn More and Lead with AI

Subscribe now to Lead with AI and be the first to hear my conversation with another AI leader who's transforming their organization.

Want to level up your AI skills today?

Want to get your whole team involved? Check out our offerings for customized AI courses for leadership teams.

If you have any other questions or feedback, or would like to be considered for the podcast, just send me an email: daan@flexos.work

🔔 Available on:

Transcript:

Daan van Rossum: This is super exciting because you're in the Lead with AI community, and you're probably one of our best examples of someone who's really using AI in the Workplace in every facet of at least their working life, as far as I've seen.

So, I'm really curious, as people just heard from your bio that you're a non-technical founder, or maybe a semi-technical founder. How did you approach getting into building a technical product? And I know that AI played some role in that. 

Sherry Jiang Peek: I have to say it started out with motivation. Because anytime you want to learn something new, it takes a lot to get past the learning curve because it is completely different. I'm a gross and consumer marketing person with a background. The motivation for me was really, I want to help, speed up product development, and build the best possible product and service for our users.

And how can I do it in a way that doesn't require necessarily shifting around the core roadmap that my engineering team is working on? It really started out as, Can I create these like mini apps, spin them up in a couple hours in a day, get some validation, and see how that can maybe be embedded into the core product later on?

So that started out as the motivation for getting more technical. Now, how did I do it? I spent some time actually talking to people that I thought were good; they're technical, but they're a good blend of technical people that have been teaching other non-technical folks how to get started. So somebody adversely shared with me how to build products, actually using a software called Cursor, which recently I think has taken Twitter by a storm.

I actually want to say I'm an early user of Cursor actually. I used it before some of those really loud developers on Twitter. But I started using it back in the May/June period to start building my own app. So I was using Cursor on Python, which actually is the most logical language for me because I've done SQL before. I've done Excel.

So, again, it's very readable. It's much more understandable for me to debug and figure out what's going on. (See also our guide to Preventing AI Hallucinations.) So, I was using Cursor in Python, and then I was using Streamlit, which is essentially like a library where you can basically deploy front-end apps using Python. So it doesn't require you to dig into things like JavaScript or JavaScript-based languages that are maybe a little bit harder for somebody from zero to be able to learn. (See here for the best generative AI courses.)

So, then secondly, I found the easiest way to get started, which is let's start with a language that I think I can be familiar with and with they'll, a way to deploy it that's much simpler than building like a full-like front-end app.

Daan van Rossum: That makes a lot of sense. And I love that you started with the motivation because you somehow felt that it would help you in your role in this case as a founder to have some of these technical capabilities.

How did the technical people in your company respond to it? Did they welcome you with open arms? Oh, that's great. We have another technical person or semi-technical person, or was there also some hesitancy of, okay, now the founder, who usually is just focused on the big picture, is now going to worry about the product. Like, how was it received? 

Sherry Jiang Peek: I think generally it's been well-received because now we can basically run in parallel. There's something what I call an 80:20. Because 80% of what our team should be doing is executing on the core stuff related to the product.

So with Peek, we need to make sure that core tracking works. We track PNL, we make sure the API integrations work well, that I will probably not touch because I think far beyond my capabilities right now as an engineer, but there's a 20% that helps us test out the ideas, maybe two, three steps. (See all AI Statistics.)

That's beyond that 80% core. And that's actually been really nice for me to liberate that from the team so that I can get started with testing and validating some stuff. So maybe I'll give you an example. We hear a lot from our beta testers that we want to be able to project our net worth or see what our fire number is for those who are not familiar.

Fires is known for and called Financial Independence Retire Early, which a lot of people on Reddit like because it allows you to envision a different life for yourself and work 9 to 5 and retire at 65. So, people are really into this idea of projecting out your net worth.

So a lot of times when people ask for these things, I'm always like, do I really want to spend one or two months of effort building out a really nice feature to see if they really want it? Probably not. So, what's a cheap way to test it?

I actually built an app in Streamlit that uses static data. There's no connectivity with Peek, which is your real-time portfolio data. Static data, but it projects out with some simulations of Monte Carlo as well, what your range of outcomes can be for your portfolio that you manually enter. And tested two things to see if this is something we actually want to build into the product.

Number one, I waited to see how many people complained that they couldn't integrate their portfolio into this tool, and I got a few; they were like, I really liked this, but it doesn't integrate. It's not real time. I'm like, I know.

And then I had other people essentially vote on it in this like community chat that I also run with our beta testers. And I was like, 1 to 5, like, how important is this to actually integrate with your data? And a lot of people voted 4 or 5, which to me is again, not super scientific.

Again, you always want to look at product analytics, but at least an indication that, okay, this might actually be a feature that we want to have some level of persistence and build into the core app versus just a standalone app.

I guess in general, it's been a welcome thing for my team, but I'm also very mindful to touch the things that wouldn't necessarily break things at the foundational level, because I am aware that my skillset is still much less than probably a senior engineer.

I'm still learning everything, but I want to make sure that I'm not the reason why Peek goes down in a given day's deployment.

Daan van Rossum: Maybe an idea that you have. Maybe a product feature or new app that you are thinking of launching, you can now, instead of just telling people about it, say, What do you think?

And it's like it's very intangible. People don't know what they're or what you're talking about. Now you can create a demo version of it by yourself and actually show it to people, and people can interact with it. 

So it's a way advanced version of a mockup, basically for people to actually feel it and then say, Oh, I want this. I love these ideas: if you wait for people to complain that it's not pulling in the real-time data, then we've got to actually develop it for real. 

And at the same time, I could actually see this as a benefit for the technical team, because now they're not getting bothered by all these little ideas that the founder has of why don't we do this? Why don't we do that? Because you're doing that on your own time. And they get to focus on building out the bigger roadmap and making sure the core of the platform works well. I think that's a fascinating combination.

I know you mentioned somewhere that, maybe now with all these tools, you can almost be your own CTO. Do you think that eventually there will just be one central person in a startup doing both the product market side and the technical side, or will we always need that truly technical person?

Sherry Jiang Peek: I think that's going to depend on the skill level of the founder or the single founder and also the product area as well. So I would say for companies that are deeply technical, where your edge is something very technical.

Let's say you're actually working on a few layers above the foundational AI models. If those types of co-founders have been like PhDs for many years and use their academic research to form the basis of that product, then I think it's a lot harder for a non-technical founder to take it upon themselves to do that by themselves.

I'm not saying it's impossible, but I do think that there are products where the technical aspect is the edge. There's a product edge there, but I do feel like for a lot of products where that might not be the case, like I don't necessarily think you need to actually start out having a CTO.

And I think there's a lot of examples actually of Indie Hackers or Solopreneurs on Twitter that build really good SaaS products for a very niche audience. And they're not necessarily technically challenging, right? So they could be like new services for email automation or small standalone products. I don't think they require you to necessarily have a full CTO co-founder, right?

So again, it also depends on the person. I like doing this stuff. I like rolling up my sleeves and figuring out how to do this. Other people might not like it, and they'll always want to have a CTO. And in that case, that's up to personal preference.

Daan van Rossum: Yeah, absolutely. But what you're saying about these Indie Hackers that you're seeing on Twitter, for example, is that it does now enable way more people to “build products.”

I also think that a lot of the time we're overstating the importance of software or technology. And we've heard some voices around maybe the end of enterprise software because everyone can just build apps on the fly. 

Sometimes you just need something pretty small and simple to solve an actual problem rather than building a whole platform. So it sounds like definitely there's going to be changes in how the product is going to get developed, how it's going to get built, and how it's going to get marketed. 

Sherry Jiang Peek: Things that I think might happen in the future are fewer kinds of pure B2B SaaS products that are being sold to startups. And I say that because right now the tech stack for the startup is just, there's so many things that we subscribe to.

I subscribe to a to-do list. I subscribe to Calendly. I subscribe to so many things, and I didn't know, deep down, can I find some way to probably rebuild this? I think if I had the time, I could.

But I think in the future, this tech of building apps is going to get even better, where people are going to realize why am I paying a couple hundred dollars or maybe a thousand dollars a month for these like clonable products because now the barrier to AI adoption might be a lot lower and you can self-service to some extent.

So there are some apps that should be a little bit concerned, I would say. They feel like there's something that somebody can clone in a few hours; with AI, there's no moat anymore. Like so many people can either create ones for their own community that are very specialized, or they might just create it for themselves.

Daan van Rossum: It seemed like such a different world, the technical and product world. And now anyone can spin up a mini-app using a cloud artifact. I think when people discover that, when I think I saw an example of a tip calculator, you're in a restaurant, it's like, Oh, I need a tip calculator. Oh, I could find some apps on the app store. 

I can just literally prompt Claude (see also our prompt generator), and it will just make it for me. And I have it right there in front of me. So that is going to shift how people think about it. 

And I know you had a career as well in big tech before you started your own company. How about when we shift the frame from being a startup founder, where obviously you can do these things? You can build a product feature, throw it in a WhatsApp group, and let people vote on it. Even put it into the main app. This is obviously very different for people who work in larger companies and maybe even like non-product people. 

Do you think there's going to be changes on that side as well in the velocity of the work or how we work when we think about it outside of the startup world?

Sherry Jiang Peek: I think it's going to have to, but there is just so much resistance, Daan. I feel like sometimes I live in this startup founder bubble where I assume everybody is obsessed with how they can make their lives 10x more productive using AI, but I'm hit with reality when I actually talk to people who are still at big tech companies. And I just realized how big of a gulf there is.

So, I'll give you an example. My first team that I joined at Google was on the strategy and operations team that actually did the annual budgeting and planning for the global marketing organization. So what I worked on actually went up to the CMO of Google. Now, you could imagine that this team is like a bunch of ex-consultants for McKinsey, BCG, whatever.

A lot of the stuff, theoretically, can be automated with AI, like chart creation, dashboarding, analytics, and all of that. And so speaking to my former manager of that team, we were just catching up in Singapore. He happened to be here for a business trip, and he was telling me that there is literally an effort he is leading.

For his team and the rest of the organization, he called it migrating to AI Websites. They're trying to educate people about why it's important. And even the people who are like fresh out of school, who I thought would be much more motivated to learn this stuff. They're like young and hungry would be like, yeah, like I would love to save time and use Gemini or whatever to improve my workflows, but he's there's so much resistance.

There's so much resistance. And I was like,

Daan van Rossum: Where does it stem from?

Sherry Jiang Peek: I'm not in big tech anymore, but I do feel like there is some concern that if someone automates their job too easily, they automate themselves out of a job. I can understand why that fear would be there. So then somebody is Oh, here's a couple buttons you're pressing, and you have all these workflows done.

Like I'll just. Take it up and not pay for this headcount. So I think sometimes the incentives are not necessarily there for everybody. I mean, that's like, maybe on the more negative side. But I think on the other side of it is just that if you're in an organization that's quite large, like Google, how 10x your productivity might not necessarily impact how well your performance review is.

It's still very much tied to perception of the work that you do, so if there is not a change in the perception of the work that you do from utilizing these tools and just being so much faster at doing shit, you don't necessarily have the incentive.

I feel like it's an incentive thing, on both extreme ends. It could also be like an exposure thing as well because, again, I feel like we're surrounded by people that feel, wow, I just extended my runway 30% by doing this. It's so tangible.

That's so real for us, but for other people, there's a lot less of that kind of fervor, that kind of conversation, day to day. So it's not really as top of mind.

Daan van Rossum: No, I think you're totally right. I think that the idea of the incentive is that if no one around you is adopting it and you feel the pressure to also take it on, you're not going to be the first one in your team to suddenly work a lot faster.

You're just not going to do it. And, definitely not; if you don't see the vision, I think that some of the AI Change Management principles we've heard about in the season so far. It's you need to show people that here's where you're going to get going with AI. So if you use all these AI tools and principles, then this is what your role could look like.

This is what your career could look like. Show them the roadmap. Show them the picture of where they're heading. If they don't have that, of course, there would be no incentive for them to start using AI suddenly, at least not publicly, because you also see a lot of people are using it secretly.

Sherry Jiang Peek: I think the main motivation I can see people doing this secretly for is they don't have to work as much and they don't have to tell their manager. They just clock off at work earlier. That would probably be my guess of why some people would do that. It's not so much like I want to do more.

It's Hey, now I get to have more time with my hobbies. That would motivate; I can see that being a motivator.

Daan van Rossum: I know that you also spoke about it for your team. You're very purposefully trying to build an AI-first company and an AI-first team. So what does that look like? Are people as motivated as you in the company to use AI? Are they using it very publicly? Or what are some things that you're doing if it's necessary to get them to use the AI tools?

Sherry Jiang Peek: I think it definitely starts at the top, right? And so both my co-founder and I are like obsessed with it. Actually, she even before me, right? I think he's somebody who's just obsessed with productivity, even before AI became like a buzzword.

So for us, I guess there's a bit of that lead by example. I will just share sometimes with my team, like, Hey guys, I did this as a fun hackathon project and I built it in AI. So, it's okay, like she's doing it. She has time to learn this stuff. Maybe I can do it too. So there's that lead by example component.

But I also make sure I do relevant sharing because I think if you're just going to your team and be like, you must figure out how to use AI to do better. It just feels like too generic and not very actionable. But I'll point it out in specific spots. So I'll give...

Daan van Rossum: Like using the internet. This is okay,

Sherry Jiang Peek: Exactly, just Google it. I just use the internet. Let's make it practical. So I know I have one person on my team, and she'll basically ask me, Oh, like, how do I do this one thing in notion? And I'll be like, use GPT first. If you don't know the answer and it's not telling you what you need, then come talk to me. So I've gotten some team members into the habit of at least using AI to self-service, some of these questions. And a lot of times they just use it. I would just be like, just GPT it.

And it's very like in the situation. It's not like I'm saying it generically when they come to me, then I get them to think, Okay, I can ask AI first. And then another example is that just recently I've been really obsessed with a V0 development by Vercel. I think it's a Claude artifact on crack, actually.

Artifacts is great, by the way, but I think there's some of the ways that they like that Claude actually produces apps. It doesn't look like production-ready, right? It feels very mocky, which is fine, but when you use for sell, it actually gives you what can actually be quite close to production-ready websites that I've actually seen across the internet.

So, today I suggested to my designer; actually, I was like, Hey, I think that there's some ways that we can actually improve this particular page, but I want you to use the tool and see what it tells you versus me being like, Hey, this is what I think, right? So I sent her the tool, and then you know how to look at it as well.

I try to find like little ways to show in situations how it works, and then I also want to lead by example. It's not just hey guys like You all figure out how to use AI, but I'm just going to sit on my old habits and not do anything at all. That doesn't really work. So I think you have to have an actual tactical playbook as well as like your own example to get people to realize the full potential it has.

Daan van Rossum: No, I love that the people have to see that it's going to benefit them personally. Otherwise, again, there's no point going back to your example of when you're in a big company when you're in big tech and you don't see the immediate incentive for you to use it. It's obviously not going to work, and you showing that it works for you and then getting your team inspired to do it as well. Going back to motivation. That's very motivating.

At the same time, you're obviously seeing the benefit because your team will work faster. You will get more out of the time that they spend at work. So it works for everyone. 

Sherry Jiang Peek: Yeah, absolutely. And like you get to automate the stuff that's not as enjoyable sometimes. I think a lot of people enjoy, let's say, thinking of how to create a really comprehensive yet clear and friendly. Personal finance experience. But if there's ways to automate, like the pixel pushing of it or how to round out some of the edges, that may get you just 90% to 95%. If there's a way to get there without using all of your cognitive energy, I think people would like that. So then you reserve your energies for the creative stuff, the human stuff. That is enjoyable.

Daan van Rossum: That's what we heard from Matt Kropp from the BCG X first episode of the season. AI Implementation is really about looking at people's current jobs and just looking at how we can remove the toil. How can we increase the joy? Debbie Lovich at BCG talks about this a lot. This should be the opportunity of, like you said, if someone is just pushing a couple of buttons, then why are they doing that?

It really is not necessary anymore with AI because all that stuff where it's just these button pushing, we can remove it already. So there's that opportunity.

I also got to ask, Are there any examples of where you were trying to use AI either in the way that you're working to be more productive or actually in building the product where it didn't work out so well, where it wasn't as good? 

Sherry Jiang Peek: Oh man. I definitely have a lot of examples. I'll give one example on the writing side.

Daan van Rossum: I'm glad. I think otherwise. Everybody will feel like an imposter here. Good. 

Sherry Jiang Peek: I'll tell you where I've tried to use AI for writing. It's sometimes when you fill out those, like accelerator applications, or I just don't like filling out applications because it feels really repetitive.

It's very different than the content I put out on LinkedIn, which again, like you want heart and soul in it. And it needs to come from you. There's so much joy in writing that I don't feel the joy of filling out applications sometimes. I don't think anyone likes it.

So I've tried to see if AI can help with that. And I think it helps maybe in structuring thoughts, but still, I just don't think it's great at writing. And if there's anyone who's found a way to do it, like automating some of these, like writing tests without making it sound like AI, hit me up because I would be happy to be able to streamline that part of my workflow.

So that's an area where I think it's still quite difficult for AI to not sound like AI, which makes a lot of sense. It's like Sherry. Don't sound like Shirley when you're writing. I'd be like, Oh, I'll really try, but it's hard because I am sharing.

And then, by the way, this is not me saying it's not going to be in the future, but I think there's just limitations. I think that doing analysis is still a little bit difficult for AI. Anything that's factual, right? What has happened in the past? In our case, like what transaction someone did in the month of May, when it comes to finances, it's easy for AI to play a querying function, so you save yourself a couple steps of navigating something. That's super useful.

I find that using AI for help center support is actually really nice. You don't give it a lot of interpretive freedom. You tell it like this is just how this API integration works.

Daan van Rossum: And you're almost stunned by how even we have one landing page with information about the course, for example. And it's really not that much information, and it's really very basic. And you still get people opening the chat and asking exactly what's already on the page. So, it makes so much sense, just like, why would a human do that? Going back to that idea of the toil in your work. That's not fun for anyone to answer those kinds of questions. Hit me up with something more challenging. 

Sherry Jiang Peek: I'm convinced that actually the function of AI is actually to make people interact with technology in a way that's actually much more natural because we created a GUI graphical interface because it was the main way that people could interact in a complex way with computers.

But AI is removing that. And to be honest, yeah, it is normal. I can be very lazy sometimes. I don't want to click through a bunch of things, and you might not think it's a lot of time, but it is energy. Every click is you depleting your own cognitive resources. So, at a certain point, yeah, people just want to type in chat with something.

I do it too, right? I use the chat function now for a whole bunch of different services that I'm trying to learn more information about. I just don't want to navigate it. I'll just be like, Can you just tell me what the pricing plan is? Just spit it out right here or tell me where to navigate.

I think that's where AI does do well, but when it comes to analysis, like the next layer of interpretation beyond this is actually what's there, I think it struggles a little bit.

An example would be okay. Now that you know this about someone's financial picture, what are the risks that they should be aware of? What are the questions that you need to find out more about them to determine whether or not their portfolio actually aligns with their goals?

There's a lot more steps that are there that you need to make sure the fidelity is good because AI doesn't have human-level judgment, like we don't need to have mountains of papers and information in front of us to advise or talk to somebody about something that we've been experts and are super familiar with.

We just know the answer.

That's not necessarily the case with AI. It sounds a little bit like someone's on their first day of work in a specific job. They're very good at absorbing all of that content right away. But the answer sounds like it's like a junior employee when I want to be talking to the expert. That's how I feel. I think it's not

Daan van Rossum: They sound like a very confident junior employee. And I think that's why KC Arvind, the CHRO of Roblox on the last episode, said don't mistake confidence for competence. It may sound very confident, but it's actually not giving you what 20 or 25 years of experience in financial planning would have given you.

But who knows, Strawberry is coming. _____ is coming. It may, by the time that we release this, not be true. Let's see. 

Sherry Jiang Peek: No, you're exactly right. And that's why I tell people, Whatever we're building with Peek, treat it like your AI financial analyst. It is not like your 30 years of experience know exactly what you should do. No, but it can help you do the math. It can present data to you and give you research options, but it's a junior employee right at the end of the day.

Daan van Rossum: That's already a huge help. So I think all of this has been super fascinating. I have three quick lightning round final questions. The first one is, What is personally your favorite AI tool? 

Sherry Jiang Peek: I love Cursor. Can't live without it.

Daan van Rossum: What does cursor do?

Sherry Jiang Peek: It's basically like a sublime text or any other text editor that you can use to code, but it has AI built into it, and so it's got a really cool feature called composer, which actually helps orchestrate AI across all of your files.

So before Cursor only worked within a single file, and sometimes when you're creating an app, you have multiple files for different components, saved in different places. So, it's almost like a composer, like an orchestrator, or like you feel like your music producer. You get to like control so much. It's super freaking cool.

Daan van Rossum: That sounds very cool. One unique use case for AI in your own daily routine. 

Sherry Jiang Peek: Sometimes, I'm super lazy when it comes to buying groceries. So I'm just like, Hey, I want something healthy. These are like the ingredients I already have in my fridge. Give me some recipe ideas and what I should buy, and I'll just go buy it because I think I'm a fundamentally lazy person sometimes in these aspects of my life where I'm like, okay, I'm very unpicky about eating on a day-to-day basis. So that's one thing that I do.

Daan van Rossum: So Bill Gates was right. You need to find lazy people because they will find really clever ways to do the work faster and smarter.

Sherry Jiang Peek: Yeah, they will work extra hard to continue to be lazy versus trying to work a little bit each time. So yeah, that's totally me.

Daan van Rossum: That's amazing. And then finally, what's one thing that business leaders should do to get everyone, including themselves, their teams, and their organizations, to get the benefits of AI?

Sherry Jiang Peek: Tell them if they finish their work faster, they get three extra days of holiday or whatever time they have saved. I'm joking. But again, make sure the incentives are there. And it really does. You will obviously meet people who are very motivated by things greater than themselves.

You might get that from folks that really care about the mission or the product or the company, but you have to assume that the vast majority of people are like infinitely more incentivized by individual outcomes. So figure out what that is for that person.

It's just like any other product building. Know who's your ICP; who's the user? What do they care about? What problems do they have? And then position it that way. And you just got to do it that way versus AI, which is this important thing and concept that everybody should do.

Daan van Rossum: Absolutely. And we can quickly spin up a GPT or artifact that lets you link. Okay. If this is the person's own motivation, how could I sell AI to them? 

Sherry Jiang Peek: This is this person's like Myers-Briggs or whatever personality type; this is what they're currently doing. This is their hobby. Like now, give me a GTM plan to convince them to use AI.

Daan van Rossum: Okay, we'll quickly build that after the episode. Sherry, thanks so much for being on. This is really fun. 

Sherry Jiang Peek: Yeah, for sure. I had a great time. Thank you.

🎧 Listen Now:

In today’s episode, we speak to Sherry Jiang, CEO of Peek, about how non-technical leaders can successfully build AI-first companies.

Sherry Jiang is an entrepreneur and tech innovator. After honing her skills as a Product Marketing Manager for Google Pay, Sherry set her sights on revolutionizing personal financial management. 

Now, as the Co-Founder and CEO of Peek, she's leveraging cutting-edge AI technology to help individuals track, manage, and grow their net worth with unprecedented ease and insight.

In this episode, you'll learn how Sherry leverages AI to bridge the gap between technical and non-technical roles, rethinks traditional SaaS models, and tackles the challenges of AI adoption in larger organizations. We’ll also dive into how AI is helping her optimize team productivity and streamline innovation.

Key Insights from Sherry Jiang

Here are the actionable key takeaways from the conversation:

1. Non-technical Founders Are Supercharged with AI

Sherry, a non-technical founder, used AI tools to bridge the gap between her non-technical background and the demands of developing a technical product.
She discussed how AI tools like Cursor and Streamlit enabled her to build functional prototypes quickly and collaborate more effectively with technical teams while also giving them the space to work on the core roadmap.
AI can level the playing field for non-technical founders, enabling them to participate actively in the technical aspects of product development and more effectively bring their visions to life.
And I think this applies more broadly – all of us now have the tools and abilities to develop ideas and maybe even apps.

2. The Future of SaaS May Shift from Centralized Platforms to DIY Solutions

This ability to quickly spin up apps also makes Sherry believe that demand for traditional SaaS products, particularly those with low barriers to entry, could decline.
She suggested that, thanks to AI's capabilities, founders and small teams might increasingly build their own tools rather than subscribing to a multitude of SaaS products.
Businesses offering SaaS products should be aware of this potential shift and consider how they can add value beyond what can be easily replicated with AI, perhaps by focusing on specialized services, customer support, or integrations.

For everyone else, look for opportunities to cut your software bills! 

3. Resistance to AI Adoption in Large Organizations Stems from Incentive Misalignment

Sherry noted the resistance to AI adoption in larger organizations, even in tech companies, and even among younger employees who should theoretically be more open to it.
She discussed how employees might fear that automating their tasks could lead to job loss, and how the lack of immediate, personal incentives diminishes the appeal of AI tools like ChatGPT.
Contrast this with Sherry, who’s highly motivated and can inspire the team to feel the same way.
For AI to be embraced in larger organizations, leaders must align incentives with AI adoption, ensuring employees see clear personal and professional benefits from increasing their productivity with AI.
Leaders should also follow Sherry’s leading by example, and asking employees to submit work to AI before they send it to their manager.

4. AI's Current Limitations: Confidence vs. Competence

Sherry pointed out the limitations of AI, especially in tasks requiring deep analysis and nuanced judgment.
She compared AI’s outputs to those of a confident yet inexperienced junior employee, noting that AI often lacks the depth of understanding needed for complex decision-making.
While AI is powerful for automating routine tasks, leaders should exercise caution when relying on it for high-stakes decisions requiring expert judgment.
Finally, Sherry also underscored what Matt Kropp from BCG X said: businesses should focus on using AI to remove toil, and increase joy at work.

Learn More and Lead with AI

Subscribe now to Lead with AI and be the first to hear my conversation with another AI leader who's transforming their organization.

Want to level up your AI skills today?

Want to get your whole team involved? Check out our offerings for customized AI courses for leadership teams.

If you have any other questions or feedback, or would like to be considered for the podcast, just send me an email: daan@flexos.work

🔔 Available on:

Transcript:

Daan van Rossum: This is super exciting because you're in the Lead with AI community, and you're probably one of our best examples of someone who's really using AI in the Workplace in every facet of at least their working life, as far as I've seen.

So, I'm really curious, as people just heard from your bio that you're a non-technical founder, or maybe a semi-technical founder. How did you approach getting into building a technical product? And I know that AI played some role in that. 

Sherry Jiang Peek: I have to say it started out with motivation. Because anytime you want to learn something new, it takes a lot to get past the learning curve because it is completely different. I'm a gross and consumer marketing person with a background. The motivation for me was really, I want to help, speed up product development, and build the best possible product and service for our users.

And how can I do it in a way that doesn't require necessarily shifting around the core roadmap that my engineering team is working on? It really started out as, Can I create these like mini apps, spin them up in a couple hours in a day, get some validation, and see how that can maybe be embedded into the core product later on?

So that started out as the motivation for getting more technical. Now, how did I do it? I spent some time actually talking to people that I thought were good; they're technical, but they're a good blend of technical people that have been teaching other non-technical folks how to get started. So somebody adversely shared with me how to build products, actually using a software called Cursor, which recently I think has taken Twitter by a storm.

I actually want to say I'm an early user of Cursor actually. I used it before some of those really loud developers on Twitter. But I started using it back in the May/June period to start building my own app. So I was using Cursor on Python, which actually is the most logical language for me because I've done SQL before. I've done Excel.

So, again, it's very readable. It's much more understandable for me to debug and figure out what's going on. (See also our guide to Preventing AI Hallucinations.) So, I was using Cursor in Python, and then I was using Streamlit, which is essentially like a library where you can basically deploy front-end apps using Python. So it doesn't require you to dig into things like JavaScript or JavaScript-based languages that are maybe a little bit harder for somebody from zero to be able to learn. (See here for the best generative AI courses.)

So, then secondly, I found the easiest way to get started, which is let's start with a language that I think I can be familiar with and with they'll, a way to deploy it that's much simpler than building like a full-like front-end app.

Daan van Rossum: That makes a lot of sense. And I love that you started with the motivation because you somehow felt that it would help you in your role in this case as a founder to have some of these technical capabilities.

How did the technical people in your company respond to it? Did they welcome you with open arms? Oh, that's great. We have another technical person or semi-technical person, or was there also some hesitancy of, okay, now the founder, who usually is just focused on the big picture, is now going to worry about the product. Like, how was it received? 

Sherry Jiang Peek: I think generally it's been well-received because now we can basically run in parallel. There's something what I call an 80:20. Because 80% of what our team should be doing is executing on the core stuff related to the product.

So with Peek, we need to make sure that core tracking works. We track PNL, we make sure the API integrations work well, that I will probably not touch because I think far beyond my capabilities right now as an engineer, but there's a 20% that helps us test out the ideas, maybe two, three steps. (See all AI Statistics.)

That's beyond that 80% core. And that's actually been really nice for me to liberate that from the team so that I can get started with testing and validating some stuff. So maybe I'll give you an example. We hear a lot from our beta testers that we want to be able to project our net worth or see what our fire number is for those who are not familiar.

Fires is known for and called Financial Independence Retire Early, which a lot of people on Reddit like because it allows you to envision a different life for yourself and work 9 to 5 and retire at 65. So, people are really into this idea of projecting out your net worth.

So a lot of times when people ask for these things, I'm always like, do I really want to spend one or two months of effort building out a really nice feature to see if they really want it? Probably not. So, what's a cheap way to test it?

I actually built an app in Streamlit that uses static data. There's no connectivity with Peek, which is your real-time portfolio data. Static data, but it projects out with some simulations of Monte Carlo as well, what your range of outcomes can be for your portfolio that you manually enter. And tested two things to see if this is something we actually want to build into the product.

Number one, I waited to see how many people complained that they couldn't integrate their portfolio into this tool, and I got a few; they were like, I really liked this, but it doesn't integrate. It's not real time. I'm like, I know.

And then I had other people essentially vote on it in this like community chat that I also run with our beta testers. And I was like, 1 to 5, like, how important is this to actually integrate with your data? And a lot of people voted 4 or 5, which to me is again, not super scientific.

Again, you always want to look at product analytics, but at least an indication that, okay, this might actually be a feature that we want to have some level of persistence and build into the core app versus just a standalone app.

I guess in general, it's been a welcome thing for my team, but I'm also very mindful to touch the things that wouldn't necessarily break things at the foundational level, because I am aware that my skillset is still much less than probably a senior engineer.

I'm still learning everything, but I want to make sure that I'm not the reason why Peek goes down in a given day's deployment.

Daan van Rossum: Maybe an idea that you have. Maybe a product feature or new app that you are thinking of launching, you can now, instead of just telling people about it, say, What do you think?

And it's like it's very intangible. People don't know what they're or what you're talking about. Now you can create a demo version of it by yourself and actually show it to people, and people can interact with it. 

So it's a way advanced version of a mockup, basically for people to actually feel it and then say, Oh, I want this. I love these ideas: if you wait for people to complain that it's not pulling in the real-time data, then we've got to actually develop it for real. 

And at the same time, I could actually see this as a benefit for the technical team, because now they're not getting bothered by all these little ideas that the founder has of why don't we do this? Why don't we do that? Because you're doing that on your own time. And they get to focus on building out the bigger roadmap and making sure the core of the platform works well. I think that's a fascinating combination.

I know you mentioned somewhere that, maybe now with all these tools, you can almost be your own CTO. Do you think that eventually there will just be one central person in a startup doing both the product market side and the technical side, or will we always need that truly technical person?

Sherry Jiang Peek: I think that's going to depend on the skill level of the founder or the single founder and also the product area as well. So I would say for companies that are deeply technical, where your edge is something very technical.

Let's say you're actually working on a few layers above the foundational AI models. If those types of co-founders have been like PhDs for many years and use their academic research to form the basis of that product, then I think it's a lot harder for a non-technical founder to take it upon themselves to do that by themselves.

I'm not saying it's impossible, but I do think that there are products where the technical aspect is the edge. There's a product edge there, but I do feel like for a lot of products where that might not be the case, like I don't necessarily think you need to actually start out having a CTO.

And I think there's a lot of examples actually of Indie Hackers or Solopreneurs on Twitter that build really good SaaS products for a very niche audience. And they're not necessarily technically challenging, right? So they could be like new services for email automation or small standalone products. I don't think they require you to necessarily have a full CTO co-founder, right?

So again, it also depends on the person. I like doing this stuff. I like rolling up my sleeves and figuring out how to do this. Other people might not like it, and they'll always want to have a CTO. And in that case, that's up to personal preference.

Daan van Rossum: Yeah, absolutely. But what you're saying about these Indie Hackers that you're seeing on Twitter, for example, is that it does now enable way more people to “build products.”

I also think that a lot of the time we're overstating the importance of software or technology. And we've heard some voices around maybe the end of enterprise software because everyone can just build apps on the fly. 

Sometimes you just need something pretty small and simple to solve an actual problem rather than building a whole platform. So it sounds like definitely there's going to be changes in how the product is going to get developed, how it's going to get built, and how it's going to get marketed. 

Sherry Jiang Peek: Things that I think might happen in the future are fewer kinds of pure B2B SaaS products that are being sold to startups. And I say that because right now the tech stack for the startup is just, there's so many things that we subscribe to.

I subscribe to a to-do list. I subscribe to Calendly. I subscribe to so many things, and I didn't know, deep down, can I find some way to probably rebuild this? I think if I had the time, I could.

But I think in the future, this tech of building apps is going to get even better, where people are going to realize why am I paying a couple hundred dollars or maybe a thousand dollars a month for these like clonable products because now the barrier to AI adoption might be a lot lower and you can self-service to some extent.

So there are some apps that should be a little bit concerned, I would say. They feel like there's something that somebody can clone in a few hours; with AI, there's no moat anymore. Like so many people can either create ones for their own community that are very specialized, or they might just create it for themselves.

Daan van Rossum: It seemed like such a different world, the technical and product world. And now anyone can spin up a mini-app using a cloud artifact. I think when people discover that, when I think I saw an example of a tip calculator, you're in a restaurant, it's like, Oh, I need a tip calculator. Oh, I could find some apps on the app store. 

I can just literally prompt Claude (see also our prompt generator), and it will just make it for me. And I have it right there in front of me. So that is going to shift how people think about it. 

And I know you had a career as well in big tech before you started your own company. How about when we shift the frame from being a startup founder, where obviously you can do these things? You can build a product feature, throw it in a WhatsApp group, and let people vote on it. Even put it into the main app. This is obviously very different for people who work in larger companies and maybe even like non-product people. 

Do you think there's going to be changes on that side as well in the velocity of the work or how we work when we think about it outside of the startup world?

Sherry Jiang Peek: I think it's going to have to, but there is just so much resistance, Daan. I feel like sometimes I live in this startup founder bubble where I assume everybody is obsessed with how they can make their lives 10x more productive using AI, but I'm hit with reality when I actually talk to people who are still at big tech companies. And I just realized how big of a gulf there is.

So, I'll give you an example. My first team that I joined at Google was on the strategy and operations team that actually did the annual budgeting and planning for the global marketing organization. So what I worked on actually went up to the CMO of Google. Now, you could imagine that this team is like a bunch of ex-consultants for McKinsey, BCG, whatever.

A lot of the stuff, theoretically, can be automated with AI, like chart creation, dashboarding, analytics, and all of that. And so speaking to my former manager of that team, we were just catching up in Singapore. He happened to be here for a business trip, and he was telling me that there is literally an effort he is leading.

For his team and the rest of the organization, he called it migrating to AI Websites. They're trying to educate people about why it's important. And even the people who are like fresh out of school, who I thought would be much more motivated to learn this stuff. They're like young and hungry would be like, yeah, like I would love to save time and use Gemini or whatever to improve my workflows, but he's there's so much resistance.

There's so much resistance. And I was like,

Daan van Rossum: Where does it stem from?

Sherry Jiang Peek: I'm not in big tech anymore, but I do feel like there is some concern that if someone automates their job too easily, they automate themselves out of a job. I can understand why that fear would be there. So then somebody is Oh, here's a couple buttons you're pressing, and you have all these workflows done.

Like I'll just. Take it up and not pay for this headcount. So I think sometimes the incentives are not necessarily there for everybody. I mean, that's like, maybe on the more negative side. But I think on the other side of it is just that if you're in an organization that's quite large, like Google, how 10x your productivity might not necessarily impact how well your performance review is.

It's still very much tied to perception of the work that you do, so if there is not a change in the perception of the work that you do from utilizing these tools and just being so much faster at doing shit, you don't necessarily have the incentive.

I feel like it's an incentive thing, on both extreme ends. It could also be like an exposure thing as well because, again, I feel like we're surrounded by people that feel, wow, I just extended my runway 30% by doing this. It's so tangible.

That's so real for us, but for other people, there's a lot less of that kind of fervor, that kind of conversation, day to day. So it's not really as top of mind.

Daan van Rossum: No, I think you're totally right. I think that the idea of the incentive is that if no one around you is adopting it and you feel the pressure to also take it on, you're not going to be the first one in your team to suddenly work a lot faster.

You're just not going to do it. And, definitely not; if you don't see the vision, I think that some of the AI Change Management principles we've heard about in the season so far. It's you need to show people that here's where you're going to get going with AI. So if you use all these AI tools and principles, then this is what your role could look like.

This is what your career could look like. Show them the roadmap. Show them the picture of where they're heading. If they don't have that, of course, there would be no incentive for them to start using AI suddenly, at least not publicly, because you also see a lot of people are using it secretly.

Sherry Jiang Peek: I think the main motivation I can see people doing this secretly for is they don't have to work as much and they don't have to tell their manager. They just clock off at work earlier. That would probably be my guess of why some people would do that. It's not so much like I want to do more.

It's Hey, now I get to have more time with my hobbies. That would motivate; I can see that being a motivator.

Daan van Rossum: I know that you also spoke about it for your team. You're very purposefully trying to build an AI-first company and an AI-first team. So what does that look like? Are people as motivated as you in the company to use AI? Are they using it very publicly? Or what are some things that you're doing if it's necessary to get them to use the AI tools?

Sherry Jiang Peek: I think it definitely starts at the top, right? And so both my co-founder and I are like obsessed with it. Actually, she even before me, right? I think he's somebody who's just obsessed with productivity, even before AI became like a buzzword.

So for us, I guess there's a bit of that lead by example. I will just share sometimes with my team, like, Hey guys, I did this as a fun hackathon project and I built it in AI. So, it's okay, like she's doing it. She has time to learn this stuff. Maybe I can do it too. So there's that lead by example component.

But I also make sure I do relevant sharing because I think if you're just going to your team and be like, you must figure out how to use AI to do better. It just feels like too generic and not very actionable. But I'll point it out in specific spots. So I'll give...

Daan van Rossum: Like using the internet. This is okay,

Sherry Jiang Peek: Exactly, just Google it. I just use the internet. Let's make it practical. So I know I have one person on my team, and she'll basically ask me, Oh, like, how do I do this one thing in notion? And I'll be like, use GPT first. If you don't know the answer and it's not telling you what you need, then come talk to me. So I've gotten some team members into the habit of at least using AI to self-service, some of these questions. And a lot of times they just use it. I would just be like, just GPT it.

And it's very like in the situation. It's not like I'm saying it generically when they come to me, then I get them to think, Okay, I can ask AI first. And then another example is that just recently I've been really obsessed with a V0 development by Vercel. I think it's a Claude artifact on crack, actually.

Artifacts is great, by the way, but I think there's some of the ways that they like that Claude actually produces apps. It doesn't look like production-ready, right? It feels very mocky, which is fine, but when you use for sell, it actually gives you what can actually be quite close to production-ready websites that I've actually seen across the internet.

So, today I suggested to my designer; actually, I was like, Hey, I think that there's some ways that we can actually improve this particular page, but I want you to use the tool and see what it tells you versus me being like, Hey, this is what I think, right? So I sent her the tool, and then you know how to look at it as well.

I try to find like little ways to show in situations how it works, and then I also want to lead by example. It's not just hey guys like You all figure out how to use AI, but I'm just going to sit on my old habits and not do anything at all. That doesn't really work. So I think you have to have an actual tactical playbook as well as like your own example to get people to realize the full potential it has.

Daan van Rossum: No, I love that the people have to see that it's going to benefit them personally. Otherwise, again, there's no point going back to your example of when you're in a big company when you're in big tech and you don't see the immediate incentive for you to use it. It's obviously not going to work, and you showing that it works for you and then getting your team inspired to do it as well. Going back to motivation. That's very motivating.

At the same time, you're obviously seeing the benefit because your team will work faster. You will get more out of the time that they spend at work. So it works for everyone. 

Sherry Jiang Peek: Yeah, absolutely. And like you get to automate the stuff that's not as enjoyable sometimes. I think a lot of people enjoy, let's say, thinking of how to create a really comprehensive yet clear and friendly. Personal finance experience. But if there's ways to automate, like the pixel pushing of it or how to round out some of the edges, that may get you just 90% to 95%. If there's a way to get there without using all of your cognitive energy, I think people would like that. So then you reserve your energies for the creative stuff, the human stuff. That is enjoyable.

Daan van Rossum: That's what we heard from Matt Kropp from the BCG X first episode of the season. AI Implementation is really about looking at people's current jobs and just looking at how we can remove the toil. How can we increase the joy? Debbie Lovich at BCG talks about this a lot. This should be the opportunity of, like you said, if someone is just pushing a couple of buttons, then why are they doing that?

It really is not necessary anymore with AI because all that stuff where it's just these button pushing, we can remove it already. So there's that opportunity.

I also got to ask, Are there any examples of where you were trying to use AI either in the way that you're working to be more productive or actually in building the product where it didn't work out so well, where it wasn't as good? 

Sherry Jiang Peek: Oh man. I definitely have a lot of examples. I'll give one example on the writing side.

Daan van Rossum: I'm glad. I think otherwise. Everybody will feel like an imposter here. Good. 

Sherry Jiang Peek: I'll tell you where I've tried to use AI for writing. It's sometimes when you fill out those, like accelerator applications, or I just don't like filling out applications because it feels really repetitive.

It's very different than the content I put out on LinkedIn, which again, like you want heart and soul in it. And it needs to come from you. There's so much joy in writing that I don't feel the joy of filling out applications sometimes. I don't think anyone likes it.

So I've tried to see if AI can help with that. And I think it helps maybe in structuring thoughts, but still, I just don't think it's great at writing. And if there's anyone who's found a way to do it, like automating some of these, like writing tests without making it sound like AI, hit me up because I would be happy to be able to streamline that part of my workflow.

So that's an area where I think it's still quite difficult for AI to not sound like AI, which makes a lot of sense. It's like Sherry. Don't sound like Shirley when you're writing. I'd be like, Oh, I'll really try, but it's hard because I am sharing.

And then, by the way, this is not me saying it's not going to be in the future, but I think there's just limitations. I think that doing analysis is still a little bit difficult for AI. Anything that's factual, right? What has happened in the past? In our case, like what transaction someone did in the month of May, when it comes to finances, it's easy for AI to play a querying function, so you save yourself a couple steps of navigating something. That's super useful.

I find that using AI for help center support is actually really nice. You don't give it a lot of interpretive freedom. You tell it like this is just how this API integration works.

Daan van Rossum: And you're almost stunned by how even we have one landing page with information about the course, for example. And it's really not that much information, and it's really very basic. And you still get people opening the chat and asking exactly what's already on the page. So, it makes so much sense, just like, why would a human do that? Going back to that idea of the toil in your work. That's not fun for anyone to answer those kinds of questions. Hit me up with something more challenging. 

Sherry Jiang Peek: I'm convinced that actually the function of AI is actually to make people interact with technology in a way that's actually much more natural because we created a GUI graphical interface because it was the main way that people could interact in a complex way with computers.

But AI is removing that. And to be honest, yeah, it is normal. I can be very lazy sometimes. I don't want to click through a bunch of things, and you might not think it's a lot of time, but it is energy. Every click is you depleting your own cognitive resources. So, at a certain point, yeah, people just want to type in chat with something.

I do it too, right? I use the chat function now for a whole bunch of different services that I'm trying to learn more information about. I just don't want to navigate it. I'll just be like, Can you just tell me what the pricing plan is? Just spit it out right here or tell me where to navigate.

I think that's where AI does do well, but when it comes to analysis, like the next layer of interpretation beyond this is actually what's there, I think it struggles a little bit.

An example would be okay. Now that you know this about someone's financial picture, what are the risks that they should be aware of? What are the questions that you need to find out more about them to determine whether or not their portfolio actually aligns with their goals?

There's a lot more steps that are there that you need to make sure the fidelity is good because AI doesn't have human-level judgment, like we don't need to have mountains of papers and information in front of us to advise or talk to somebody about something that we've been experts and are super familiar with.

We just know the answer.

That's not necessarily the case with AI. It sounds a little bit like someone's on their first day of work in a specific job. They're very good at absorbing all of that content right away. But the answer sounds like it's like a junior employee when I want to be talking to the expert. That's how I feel. I think it's not

Daan van Rossum: They sound like a very confident junior employee. And I think that's why KC Arvind, the CHRO of Roblox on the last episode, said don't mistake confidence for competence. It may sound very confident, but it's actually not giving you what 20 or 25 years of experience in financial planning would have given you.

But who knows, Strawberry is coming. _____ is coming. It may, by the time that we release this, not be true. Let's see. 

Sherry Jiang Peek: No, you're exactly right. And that's why I tell people, Whatever we're building with Peek, treat it like your AI financial analyst. It is not like your 30 years of experience know exactly what you should do. No, but it can help you do the math. It can present data to you and give you research options, but it's a junior employee right at the end of the day.

Daan van Rossum: That's already a huge help. So I think all of this has been super fascinating. I have three quick lightning round final questions. The first one is, What is personally your favorite AI tool? 

Sherry Jiang Peek: I love Cursor. Can't live without it.

Daan van Rossum: What does cursor do?

Sherry Jiang Peek: It's basically like a sublime text or any other text editor that you can use to code, but it has AI built into it, and so it's got a really cool feature called composer, which actually helps orchestrate AI across all of your files.

So before Cursor only worked within a single file, and sometimes when you're creating an app, you have multiple files for different components, saved in different places. So, it's almost like a composer, like an orchestrator, or like you feel like your music producer. You get to like control so much. It's super freaking cool.

Daan van Rossum: That sounds very cool. One unique use case for AI in your own daily routine. 

Sherry Jiang Peek: Sometimes, I'm super lazy when it comes to buying groceries. So I'm just like, Hey, I want something healthy. These are like the ingredients I already have in my fridge. Give me some recipe ideas and what I should buy, and I'll just go buy it because I think I'm a fundamentally lazy person sometimes in these aspects of my life where I'm like, okay, I'm very unpicky about eating on a day-to-day basis. So that's one thing that I do.

Daan van Rossum: So Bill Gates was right. You need to find lazy people because they will find really clever ways to do the work faster and smarter.

Sherry Jiang Peek: Yeah, they will work extra hard to continue to be lazy versus trying to work a little bit each time. So yeah, that's totally me.

Daan van Rossum: That's amazing. And then finally, what's one thing that business leaders should do to get everyone, including themselves, their teams, and their organizations, to get the benefits of AI?

Sherry Jiang Peek: Tell them if they finish their work faster, they get three extra days of holiday or whatever time they have saved. I'm joking. But again, make sure the incentives are there. And it really does. You will obviously meet people who are very motivated by things greater than themselves.

You might get that from folks that really care about the mission or the product or the company, but you have to assume that the vast majority of people are like infinitely more incentivized by individual outcomes. So figure out what that is for that person.

It's just like any other product building. Know who's your ICP; who's the user? What do they care about? What problems do they have? And then position it that way. And you just got to do it that way versus AI, which is this important thing and concept that everybody should do.

Daan van Rossum: Absolutely. And we can quickly spin up a GPT or artifact that lets you link. Okay. If this is the person's own motivation, how could I sell AI to them? 

Sherry Jiang Peek: This is this person's like Myers-Briggs or whatever personality type; this is what they're currently doing. This is their hobby. Like now, give me a GTM plan to convince them to use AI.

Daan van Rossum: Okay, we'll quickly build that after the episode. Sherry, thanks so much for being on. This is really fun. 

Sherry Jiang Peek: Yeah, for sure. I had a great time. Thank you.

FlexOS | Future Work

Weekly Insights about the Future of Work

The world of work is changing faster than the time we have to understand it.
Sign up for my weekly newsletter for an easy-to-digest breakdown of the biggest stories.

Join over 42,000 people-centric, future-forward senior leaders at companies like Apple, Amazon, Gallup, HBR, Atlassian, Microsoft, Google, and more.

Unsubscribe anytime. No spam guaranteed.
FlexOS - Stay Ahead - Logo SVG

Stay Ahead in the Future of Work

Get AI-powered tips and tools in your inbox to work smarter, not harder.

Get the insider scoop to increase productivity, streamline workflows, and stay ahead of trends shaping the future of work.

Join over 42,000 people-centric, future-forward senior leaders at companies like Apple, Amazon, Gallup, HBR, Atlassian, Microsoft, Google, and more.

Unsubscribe anytime. No spam guaranteed.
FlexOS | Future Work

Weekly Insights about the Future of Work

The world of work is changing faster than the time we have to understand it.
Sign up for my weekly newsletter for an easy-to-digest breakdown of the biggest stories.

Join over 42,000 people-centric, future-forward senior leaders at companies like Apple, Amazon, Gallup, HBR, Atlassian, Microsoft, Google, and more.

Unsubscribe anytime. No spam guaranteed.
FlexOS - Stay Ahead - Logo SVG

Stay Ahead in the Future of Work

Get AI-powered tips and tools in your inbox to work smarter, not harder.

Get the insider scoop to increase productivity, streamline workflows, and stay ahead of trends shaping the future of work.

Join over 42,000 people-centric, future-forward senior leaders at companies like Apple, Amazon, Gallup, HBR, Atlassian, Microsoft, Google, and more.

Unsubscribe anytime. No spam guaranteed.

Our latest articles

FlexOS helps you stay ahead in the future of work.

New Data: Amazon Employees Ready to Quit.

New Data: Amazon Employees Ready to Quit.

October 3, 2024
4
min read
New Data: Amazon Employees Ready to Quit.

New Data: Amazon Employees Ready to Quit.

Is your organization prepared for the risks of a rigid return-to-office policy? New data shows the risks of mandating a 9-to-5 in the office.
Sep 16 2024
5
min read
How to Address “Coffee Badging” in Your Workplace

How to Address “Coffee Badging” in Your Workplace

Coined last year as a new buzzword to address the realities of hybrid work, “coffee badging” has become a common vocabulary. This guide explores practical ways to address and make your hybrid model work.
Sep 16 2024
5
min read
What you need to know about Meta AI + 10 use cases with ChatGPT Voice Mode

What you need to know about Meta AI + 10 use cases with ChatGPT Voice Mode

Meta AI new updates, AR glasses, ideas to try with ChatGPT Voice Mode, 10 daily AI use cases for business leaders, and more.
Sep 16 2024
5
min read
Report: Forget the Headlines. Hybrid Work is Growing

Report: Forget the Headlines. Hybrid Work is Growing

Latest data shows that hybrid and remote work continue to rise, with hybrid workers preferring to spend three days in the office. More key findings are below.
Sep 16 2024
5
min read
Amazon's RTO Push: Are We Headed Back to the Office for Good?

Amazon's RTO Push: Are We Headed Back to the Office for Good?

Amazon's strict office return in 2025: Is this the end of hybrid work or a risky move that could trigger mass resignations?
Sep 16 2024
5
min read
Google NotebookLM is Your New Must-Have AI Tool

Google NotebookLM is Your New Must-Have AI Tool

Ideas to try with Google’s NotebookLM, AI assistant, OpenAI’s investors, and more.
Sep 16 2024
5
min read
Q & A with Tiffany Owyang Lam: Inside Grammarly’s Transition to Hybrid Work

Q & A with Tiffany Owyang Lam: Inside Grammarly’s Transition to Hybrid Work

Grammarly’s North America Workplace Experience Manager shared her insights on how the company has been implementing one of the most robust hybrid work programs for years.
Sep 16 2024
5
min read
Gartner Hype Cycle: What’s HR’s Role in The Future of Work? 

Gartner Hype Cycle: What’s HR’s Role in The Future of Work? 

AI avatars and no HR in Gartner’s Future of Work report? Time for HR to step up or get left behind in the digital age.
Sep 16 2024
5
min read