AI For Better Career Resilience: A Tech Optimist's Perspective

Hey, everybody, welcome back to the Futurist Society, where as always, we are talking in the present, but talking about the future. We have a really special guest today. Her name is Joan Palmater-Bajorek, and she's an author, bestselling author, and a thought leader in the AI space.

And there's so many different things that are happening in AI, but we kind of lose sight of the fact that how is it going to affect us? And that's one of the reasons why I wanted to talk with Joan. You know, it's going to be very interesting to see what AI does for the world in general, but for just like the human level of Joan, I think you have a really unique perspective.

So tell us a little bit about your journey and how you got to this space. And, you know, I know you've had some ups and downs that made you write the book in general. So I'd love to hear how AI has affected you personally and what made you write the book.

Yeah, absolutely. Well, delighted to be here and excited to talk to folks. Greetings from Seattle. My name's Joan. I am traditionally, my background, I'm an ex-academic, so I'm a linguist. I'm a researcher. Back in the days when people laughed at linguists, are you going to get a job? We're worried about you.

Now, today, they're like, whoa, you're a linguist. I've got about 10 years background in data, building infrastructures of the back ends of Alexa and Siri devices. So my PhD is in conversational AI, speech language tech.

But I'm also one of those people that's lived the news. I've been in two large scale layoffs. It's so ironic to be building AI infrastructure and be in these AI layoffs, 30% of the company let go in a day, right? Like really high volatility. And so good news. I think the first time I was laid off, how many months later, I closed my first six figure contract after it.

So as far as entrepreneurship is something I recommend to folks. I look forward to getting into things today. But the ability to close my own contracts, build income streams, I've been very resilient and people have seen my career.

I've now more and more followers online. People are like, how did you do it? This resilience thing that we see you doing. My DMs are full of people from around the world being like, help me. I want a job in AI or I'm worried about losing my job. Joan, give me the advice.

So my book is called Your AI Roadmap. And it's all about modern careers, resilience, and then also modern finances. So that's in a nutshell, it's like, you don't make money on books. I don't know if you have one, but I make like $2 to $3 per book sold. So it's really a love letter giving back to my community and people around the world.

So what do you mean when you're talking about, let's say entrepreneurship, right? And you're saying you closed your first contract after you were, what does that mean? Nuts and bolts, explain that to me.

Who were you talking to? How did you transition the skills that you had in your linguist career to a job? And I think that that's one of the things that I feel like people should wrap their head around. Because for me, I'm a surgeon and it's a very clear path of how to get money. I have to work for a hospital, get a salary. So I don't really think of closing contracts.

After, let's say AI comes in and they build a surgical robot that takes away my ability to operate, how is that something that I'm going to be able to transition my skills into income for my family?

Well, totally. And I think in the United States, we have this framework of your job is your income and your medical insurance. And that's a very dangerous game sometimes.

So back the first time I got laid off, I had a side hustle. I was building a community. I was building a following. And when I got laid off, I had a lot more time for that side hustle. But also there have been other opportunities that had been nearby that I was like, I don't have bandwidth for that. I don't have time for that. 

And so they wanted to do an event together. And I was like, there are dollar signs on that. And we agreed on it. And Google was our first corporate sponsor at $100,000. So as far as, I mean, any contract, it's a relationship building thing. It's a what's the deliverable.

It's putting a dollar sign, people signing off. I mean, setting up bank accounts, like nuts and bolts. I think people think it's like, oh my gosh, it's so scary.

But if you like, here's the goal, and then step by step, how are you going to get there? Almost anything is achievable, I believe. So side hustles, different income streams that at the time, I didn't need to be making money on the side.

 I was making good money. Later, I wanted to be scaling things. I wanted to be paying myself.

So that shifted. And I think that's really the thing is we talk or depending on how much you think about your finances, people think about diversified, bless you. People think about their finances typically at a diversified portfolio, right?

You don't shove all of it in crypto. You want different things to be playing out. But we have this one income stream typically on a W2. And if you lose that overnight, you could be in a really dangerous place. So I have 22 income streams. I share all in my book. 

I write them all out. But I recommend to folks, you know, I'm not a wealth planner. My lawyer would say, hey, you know, disclaimer, disclaimer, two to six income streams could be a really great place for a lot of folks.

And are you leveraging AI to do that? Is that one of the, is that what the whole gist of your messages is that AI can help you get those 22 income streams? Because to me, that seems like a little overpowered.

You can't have 22 things that you're juggling like that. I mean, I don't know if I can juggle my one job, you know? But anyway, just tell me, how are you doing it?

Yeah. I don't think anyone needs 22 income streams. I'll be very honest with you. When I wrote it out, I was shocked how many I had, frankly. I'd say a quarter to a third are passive. So there's things I set up that just keep rolling, like a high yield savings account.

They just pay me hundreds of dollars to leave my emergency fund with them. I love that. I love that.

Setting up income streams, I use a rebalancing tool that's consistently making my portfolio the right percentages I want. So automated AI, I think I definitely certainly hear that in my DMs how do I use AI to get rich is a question I absolutely get. We can build AI and sell it, but I don't… any like AI to money schemes, I can't get behind.

Yeah. So I guess my question is when you say it's like your AI roadmap, is this a function of like, AI is coming and you need to protect yourself or is it more like AI is coming, you need to protect yourself and we're going to use AI to make these revenue streams and to make you financially independent?

It's more the resilience through the age of AI. It's a companion to my podcast, which is a very technical and career-based thing. Instead of just what's the future, like people on stages who don't build.

I have my friends from Microsoft, Google, Adobe saying, here's what we're building. Here's how hard it can be nuts and bolts. And also here's how I got in the door to my career to build these things.

So this is a companion to that of the modern careers and resilience in the age of AI. If you lose your job, future-proofing your career, future-proofing your income with the help of AI. But it was wild to me doing the research for this book and being like, tripling down on humanness for the modern careers is what the research bears out.

What do you mean by that?

Yeah. So during COVID, when there are huge COVID layoffs, would you want to guess the percentage of people who got their next job based on introduction?

50%? 

70%. So people think that sending my CV in, blah, blah, blah. If 70% of people are getting their next job based on introduction, it's not the side way. It's the main way.

So robustly working on your network, your personal brand, going to those events, following up with people. I recommend a spreadsheet. Follow up with all the people.

But that humanness, that Lisa, Sandeep, leveraging your network is how you might land the next job statistically. So I think tripling down on that human piece instead of polishing your CV is the way.

What about the skills that are going to be necessary in the future with AI? I know that people are talking a lot more about how it's much more human-driven and AI is a collaborator. Explain to people what you think that they're going to need for realistically when AI comes, I think everybody says that work is going to be very different.

And so I'd love to hear your perspective on what you think that those skills are going to be, that very human-centered, but that's a general term. What are some things that people are going to need to make sure that they're competitive?

Yeah. Well, and it's hard to make a pan like everybody needs to X. So I don't think I can do that.

What I do see in the shape of teams in the marketplace is really a bifurcation of entry-level jobs, internships. There are so few of them. It's disappearing.

And more like really senior technical research and development jobs, R&D, are specializing, niching, my friends in cybersecurity, computer vision, etc. Hard jobs, very well-paying jobs. So that's changing.

To anyone listening, at least this is a pan recommendation, upskilling in AI for your field, I think is a really crucial thing. There's that joke that is going around of human versus human who knows how to use AI, who's going to be kept on the team. Shocker.

So upskilling in whatever for finance, upskilling in whatever that is for surgery. I've definitely worked on projects where they're like, here are these augmented pieces that might be there for doctors in the future. Being open to that, being ready for that, testing it out.

Yeah. There's so many cool things in beta that are coming out soon. How to debug code.

Because we have generative AI, you've seen it in the marketplace, these images that are terrible, or text that you're like, this is decent. A human needs to make changes. Right?

And that's what I was just coding yesterday. I was like, this is pretty decent. I need to make changes to still make it correct coding and spit out the right thing.

So augmented partner translating your skill set. Yeah, I recommend.

So how much of this is going to be, when you say interacting with AI, or people who know how to use AI, how much of this is going to be coding and how much of it is just going to be a general curiosity to what's coming down the pipeline? How technical is this going to be?

Less technical than you think. I really think, especially products like Chachapiti, you can just ask it for what you want more and more. There's been a joke that English is now the coding language, is a phrase I've heard in my community.

Yeah. But the thing is, we often are in echo chambers. I work in tech, I'm here in Seattle.

Research from Pew Research Center says that US adults, 65% have heard of ChatGPT. Do you want to guess what percent has tried ChatGPT? 14%.

Oh, really?

14%.

Seems super low.

Seems super low.

Maybe I live in a bubble, but I feel like everybody's using ChatGPT these days.

I hear you. But maybe big family events start talking to other people and doing a little more math. But a lot of people barely have heard of it, to actually using it, to potentially using an API to build with it.

So there's a huge opportunity for people who want to be prototyping, testing with it, to be on the cusp of new things. I'm seeing products... Or yeah, I can pause there.

Well, my question to you is, when you're talking about building and using an API to build, what do you mean by that? What would be the line of somebody who's dabbling in this versus somebody who's actually building something with this?

Sure. So building with... And there's so many large language models, right? 

I'm using ChatGPT, but we could be talking about Perplexity, we could be talking about Anthropic products, just to clarify that. When I'm talking about being an end user of ChatGPT, maybe I'm using the free version, maybe I'm using the plus version, paying 20 bucks a month, maybe writing emails, cleaning up some things, advice on an awkward thing at work. My coworker said this, what do I do type of thing. 

Most of opening eyes money is coming from working directly with companies. An API is kind of like when you plug in a computer to the wall for electricity, we're connecting different pieces. So an API call is, we need this dataset.

We're plugging into a huge dataset, we connect those pieces through what's called an API, and we're calling different things back and forth. So that enterprise tier of using it. Now, people other than enterprises can use it.

Most US people can be using it. But as you mentioned, do you need coding? How much coding do you need to make these custom GPTs, for example?

Not much. If you put in six to 10 hours, you could be in a great place. The learning curve, I don't know. Maybe people on this podcast are like, I don't want to touch Python. No R. But I think it's more accessible today than any other day, is what I'd say.

Yeah. I think that I've heard of that, that it's more accessible. I personally have not dabbled in it.

I use a chat GPT and a few of the different image generating algorithms for presentations, for writing emails. It's so much easier to write an email, just giving them the gist and then shooting it off. But I know that people are building stuff.

And I feel like it's kind of slighted towards large companies, that enterprise level. A guy in their garage is not going to be, in my opinion, building something with this. But that's just my perspective.

And I look at Silicon Valley, like in the past, where people were building companies in their garages, and then they grew to multi-billion dollar companies. Is that still happening? Or is AI kind of a, this is a skill that I need to learn so that I can get a job in AI?

Or is it something that you can use it? If you own a food truck, is that going to expand your business? Is that going to bring value to your business other than just this thing that makes life a little easier?

 Yes. Whoa. Okay. Multi-tier question. You're like throwing them at me.

Sorry.

No, all good. I think for the opening eyes of the world with chat GPT, right now, I think it's like 90% of the Fortune 100 are prototyping and seeing if there's good value out of this. Whether they see a good return on investment, I have question marks about, because it can be very expensive.

Your second question, can you help me go back to that one?

Yeah. Is it going to be able to be extracting value for people who are on the ground level? Is it just companies that are going to be able to become wealthy through AI?

Or is it like, if I had a food truck, am I going to use AI in such a way that this would be able to take me from let's say $100,000 in sales a year to $200,000 in sales a year?

Yes. I believe just like we use the word computer, any company can leverage AI to be more profitable to understand. So the food truck, for example.

So I'm an investor myself. When I think about different companies that are working to scale that say, we have high ambitions, we want to do these type of things in revenue, for example, do they know the average order value? Are they doing upsell and cross-sell effectively?

Are there products that are losing them money that they would be wise to remove from their menu? Is there something they can niche down and triple down and be like, it's actually the sauce. The sauce is what's going to get us to that huge revenue goal.

So just having the data and leveraging AI to analyze the data, provide insights could be game changing for that truck or the location of it. I don't know if you eat at food trucks, but if it's parked in one place versus another place, they can see tons of food traffic differences. So instead of just shooting from the hip and be like, I hope it works out, I hope we do some food truck stuff.

Typically, restaurant and food margins are very slim. So if you do have those ambitious goals, maximizing each piece of variables I just mentioned, I'm not even in the food business. There could be other variables.

But if you were to optimize 10% of each of those, I could see robust fiscal changes.

Yeah. I mean, I hope so. Some of the terms that you threw out there, what the difference between, sorry, you said cross-selling and what was the other thing?

Upsell and cross-sell?

Yeah, upselling and cross-selling. That's not something that I know because business is not what I'm into. But when I look at the internet, for example, the internet allowed that one food truck that might only be known locally to expand to a much larger area.

 You could have your website on Google. You could have social media promoting your stuff. So you could very easily take something that was really locally known and make it at least regionally known very easily.

And it seemed a lot more egalitarian and accessible because I don't think it was very difficult for at least the last 20 years or so to get on the internet, have an internet presence and all this other stuff. But is AI going to be like that? I feel like when I look at it right now, it's this nuanced tool that makes us a little bit more productive.

And then you have this big leap to people who have tons of data, who have tons of manpower, who have tons of resources. And they're going to extract a lot of data. But there's such a big gap between those two things.

One is kind of like a novelty. And then the other is like people are actually making money on this thing. And that's just my perspective and it's the outsider's.

So I'd love to hear what you think.

Yeah. I love that perspective of saying like, can the little guy see those returns? I think is kind of the question I'm hearing.

And the answer is definitely yes. I think that's also what investors are really excited about is a team of three to 10 people with a multi-million annual recurring revenue. I don't know if you, for example...

Yeah. Can you give me some examples of that?

Yeah. There's a post. I wish I had that ready for this podcast.

I can look into it. But for example, Aragon.ai, I don't know if you've heard of them. They are a high quality portrait tool where you put in like 20 selfies of yourself and it spits out really pretty good imagery.

And so the ability to not have to buy a new outfit, hire a photographer, get it edited, like these headshots, I think you can get like 40 for 35 bucks, something like this. I was showing a family member and they were like, most of those look like you. Like I tried out the tool for myself.

I think their team is less than 10 people and their annual recurring revenue is in the multi-millions. So that's just one example. But I think if you hear that, you say, here's this cool technology.

Here's a really specific use case. Here's how we'd use it. I can articulate to you being like, here's what this tool does.

Go try it out. It's a price point that's relatively accessible. And they're building out for doctors, for lawyers, for real estate agents that they're specializing.

But that type of profitability in a small team is very attractive to investors. So it's possible. That's one of my favorite examples.

I've used the product. But yeah, Aragon.ai, they didn't pay me to say that. But that's just an example of these really small teams making robust revenue, selling the products that they build with AI.

Mm-hmm. Interesting. Yeah.

No, I'd love to hear more about those examples because I just feel like those are not what you hear about in the news. What you hear about in the news is that Microsoft is spending $80 billion in AI. Technically, I'm an owner of a small business.

I'm in private practice also. And so we have a team of 200 people within all of the offices that we work on. And people will come to us.

Vendors will come to us and be like, hey, we have this AI solution for you that will do all of your marketing or something like that. And I just feel like it's me spending money to pay somebody else. And they're getting the value out of it.

Whereas I feel like as someone that may not be as well-versed in AI, it would be awesome for it to be accessible enough for us to have a team. And it sounds like that's something that you could do. I just don't know.

How skilled do you have to be? Is this something that you have to be an AI researcher like yourself with all of these significant decades of experience in AI? Or is it something like you were saying, in 6 to 10 hours, can a small business gain value from this?

Yeah. I mean, these are big questions. I think it's really interesting that you're like, does it need to be external?

Do we bring it in-house? A lot of companies are asking these questions. Do we need to hire a PhD?

All these different things. Can we get someone who just can upskill internally? I've seen literally all of these scenarios we're talking about play out in different companies that I've talked to.

I mean, my B2B business to business company, Clarity AI, we do workshops for companies to say, is there low-hanging fruit you don't even know about that is just sitting in your repo that we could leverage and make you more money? That's a conversation we absolutely have with different companies. And then what we build, I think what's, at least my business, what we love to do is partner with other companies that their domain expertise is wildly outside of AI, but they have ambitious goals like we're talking about.

So my biggest customer is an agriculture company. They have deep domain expertise in vertical farming, which is just you stack the soil basically. It's pen and paper today.

That's where we're talking about. So for the ambitious goals they have, building their software, AI, data infrastructure, I could talk about data lakes, digital twins, but to partner with those domain expertise to leverage the best of what we know and hit their goals, their investor goals, are some of my favorite current projects. I think for the example you're talking about with doctor's offices, you could bring it in-house.

I don't know that six to 10 hours is going to get you the type of results my team can get you, no offense, or just the different play. But I think it's that, do we see, and as you mentioned, I'm hiring these randos, are they actually going to help me get value? I get it, totally.

I think that's why a strategic thing to say, are there pieces of the marketing stack? Is there something in customer service that we could automate? I'm all about the fiscals.

If we don't see a return, why are we building it kind of land? So I'm not trying to pitch you my company. I'm just saying how I think about what is hopeful, where we see it moving the needle for end customers as well as the business.

Yeah. I think that the idea of even just like data cleanup, right? I mean, there's so much data that's out there.

You kind of mentioned a few of those terms that this pen and paper company is actually going into. Are you building the software yourself or are you bringing people in for that? Or do they have existing software?

What is it like for a company like that? Because that's how I see the majority of companies, right? I mean, other than large healthcare organizations, most private practices are just like mom and pop shops, right?

And so that's my perspective on business in general. But I think that amongst my friends and stuff like that, the majority of small businesses are like that, right? So would you say like the first step is creating a system where you can get good data? 

Is that something that kind of like you're going into a situation with or is it just kind of... How are you addressing the majority of like these pen and paper businesses?

Yeah. Oh, there's so many different ways to do it. I mean, for this customer, I mean, he was like, I want AI on it, like yesterday.

Like, let's do it. Let's go. Which is not what I usually meet.

Stakeholders are like, I'm reticent. Is there money for this? I think, or any of your listeners as well as you, you need quality data.

Data is the foundation of the house that you're building the AI on. If you start with bad data, not enough data, like we are building extrapolations and models that are not good. So I always want a really, as you mentioned, data cleaning, that is a skill that will not go away anytime soon as far as the nasty repos I see.

So collecting their data, organizing their data, basic nuts and bolts need to be set up in the business, period. We are doing a custom build based on they want intellectual property in that domain. They would like to look at patenting and having, or that's a piece of the company.

I'm allowed to share these things. My customer, Agria, he loves me to talk about on podcasts, but their investors one day see the data set that we are building about productivity and yields potentially be more valuable than the produce they build in a three to five year time span. So it is an investment of the business that they see.

Do you know what I mean? Like today is the but into the future, this is core intellectual IP for the company. And I think that perspective may be different from a mom and pop shop that says data repos, Joan, I don't care.

I need to care about my fiscal ability to stay in business the next two years. It's a different conversation.

But I do see just huge opportunities to prototype if you say my biggest expense for the business is X. Is there a way we can increase revenue? Is there a way we can cut down on some of these expenses overhead driving me bananas?

Like I love looking at a specific project and seeing if we can move the needle on those different pieces because based on the projects I've done for enterprises in the last 10 years, worked with Fortune 500 customers, I could mention Verizon and Boeing, for example, moving the needle two to 10% on different pieces is absolutely possible. And that can be multimillion dollar changes to the business. Again, that's back to the enterprise.

I know you just said mom and pops, but as you mentioned, like, is there a huge marketing play? Is by hiring someone who is an expert at TikTok can increase our traffic 400%? That might be a cost to the business that's worth it for a test.

I mean, my cousin was begging my aunt to go to this like shop, I think it was a burger shop recently because he'd seen it all over TikTok. He was like, we have to go. My aunt was like, the burgers weren't even good, but like that desire to go and that, you know, whatever team had built at that TikTok, that burger shop was hopping, right?

So what was the cost to that of the social media to the business? And what was the return? That's a, I mean, we're talking more business stuff than we're talking to AI here.

But- No, I think it's important to realize that like that's realistically what AI is, right? It's just a business case, you know? I mean, I do wanna talk with you about, you know, what the lifestyle is gonna look like when we get to, especially like super intelligent, generative AI, super intelligent, general intelligence, you know, I think that that's gonna be a very interesting scenario.

Like, is that gonna, in your opinion, replace employees? Like, what do you think that's gonna look like? Because that's something that, you know, I think that most people are afraid of, right?

Most people are afraid of this robot that's gonna come in and take their job. But I don't know if I necessarily see this, as my listeners know, like I'm very much a techno-optimist. Like, I think that it's gonna make our lives easier, but what's your opinion about that?

Yeah, I think it depends, or it depends. I'm a wild optimist. I think there are opportunities that are open for basically anybody with Wi-Fi, time, and a computer that did not exist five to 10 years ago.

Like, the ability to build a side hustle, build an income stream, have the life you want for yourself is absolutely here today. So I'm very optimistic about that. I think there's a huge shift in what skillsets are needed, people being laid off, and are they prepared for the jobs that are burgeoning?

I've got big question marks around that. Who gets laid off? All that jazz.

I think what I'm experiencing, at least, is I have AI in my business, right? My team is leveraging these tools. We are so much faster.

We can fill our calendars, and that productivity, quote-unquote, can lead to burnout. It's a little bit scary. I have to put boundaries, or just I can be done by 2 p.m. and be like, do I wanna keep grinding? Do I want to go outside with my dog and work out? Like, I actually have to think about that far more carefully because, quote-unquote, being done with those tasks takes a different amount of time, and is it a high quality regardless? So, it's just a different way to think about my life and the structure of time, and that's a privilege.

I totally get that, but I think that is a question, I think, that's gonna keep coming up for people who are implementing AI, who are leveraging it as a huge efficiency ability.

Do you think that's gonna get better or worse when we have artificial general intelligence that's as smart or smarter than a human? Oh, sick. Like, what is our, what is, because I feel like our lives are gonna get easier.

I think we're gonna have more time to go to the gym and walk our dog and do all that stuff because it's gonna be a tool that does a lot of the mundane tasks. I mean, there's so many mundane tasks in being a human in 2025. Outside of even work, right?

Like, just driver's license renewals and registration and taxes and all this other stuff. And if there was somebody else that was doing that stuff within reason, I think that would be really great. The way that I look at it is, for example, radiologists right now, they have the ability to have an AI that looks at the radiographs and basically you just approve it or disapprove it.

So you could still do your job, but you can do it at a higher capacity and maybe it might take you eight hours to do 10 reads now and you could do that in four hours so it'd give you more opportunity to do the things that you like to do, like talking to patients, like that kind of stuff. Is that the way that you see it or do you think it's gonna be a little bit different than that?

I definitely see that for certain things. So exactly as you're talking about, I'm more on the same page. If it's giving you these insights and being like, here are the things you need to review that we're worried about, whatever pathogens and we need a human in the loop to review it, that highly skilled person, maybe or maybe not, but that bifurcation of, we don't need as many entry-level people to review X number of charts potentially or whatever visualization.

I'm still worried that people are gonna burn themselves out and that like, oh, just keep grinding hamster wheel. And the reason I say that is because when the advent of the dryer and washer for clothes and the, for plates, washing, whatever, these machines that help us, they were meant to automate and like give people back time. Did people actually rest?

No, a lot of people just filled their time with things. And so I think that's, maybe that's a mindset, maybe that's the hustle culture that I participate in, whatnot. But for me, I feel the excitement and I see the output so fast.

I'm like, whoa, I do this. Whoa, it's already coming back to me. Let me just keep doing it.

And so if I'm experiencing that and I go to therapy, workout and try to set those boundaries, I'm a little bit nervous that other people are just gonna start grinding away as well. And as you mentioned, those efficiencies, I think we just need to be really mindful is kind of what I see.

Yeah, yeah. People are gonna be using it as a tool to get ahead as opposed to a tool to get time back. Is it, but that's a cultural thing.

I think that that's something that if you really wanted to, you could stop in your own life, but you just have to be mindful of it, similar to what you're saying. Yes. But what else do you think is gonna look different with the artificial general intelligence piece?

Because everybody's talking about it. Everybody's worried that it's like two years away and here we are in 2025, let's say it comes in 2027. What do you think is gonna be from your perspective, like something that is gonna be very, very groundbreaking?

Very, very groundbreaking.

Like how is the world gonna look different when artificial general intelligence is just like our cell phones, you know, just everywhere?

I mean, arguably it is in many ways already. I have been working on a lot of opening eye demos. I do technical courses for Pluralsight, which is a company that helps big companies upskill internally their employees.

So I make very technical courses. What opening eye says it can do. And when I do the demos and I'm like, this is glitchy, like what you say publicly and what it can do today for an end user is not the same.

We still need to figure out what to do with it and how to use it. You know, is it gonna help us with that admin task? You do still have to set it up to run whatever you need it to do.

AGI, oh, it is slower and less savvy than anyone will tell you. I think there's one thing to be like, wow, these products. As a builder, I'm like, this dataset is not good.

Is the output really strong? I still have question marks. We think we've solved things like voice recognition and it still performs horribly.

So I'm, we've got to set benchmarks for these things, but I really, AGI, groundbreaking. I think agents and setting up agents effectively for, as we mentioned with that photo example, a specific task to continue to do for you. I mean, I already have, so I use SaneBox.

It organizes my email every day before I even get up. Really important emails are in one folder, customer emails, family stuff, right? Goes, that's how I set it up, top of my inbox.

Like other things, less important things are in different folders. And it says that it saves me six hours a week of time with my email stuff. It's encrypted.

It doesn't even read the emails. It sees my behavior of how I've chosen to move those emails and given me six hours back of my time. We could call that an agent.

We can call that AI. We can just call it an automation that I've set up for my email. But like basic low lift stuff about Joan organizing her email is really, really helpful for me.

And I could have paid a virtual assistant to do that. I think it's like a hundred bucks a month. No, a hundred bucks a year.

It's super cheap. I think it's a hundred bucks a year. But like, as I'm saying, instead of like, oh, it's coming, I'm like, pieces of it are already here.

And I think it's knowing which task and what you want help with. I don't know if you've heard of Sandman Box or these inbox automation.

No, I mean, I've definitely seen like some of the automation coming from Google, like just because I'm very much in the whole like Gmail, Google Drive ecosystem. But, you know, I just like, I think that like, the lowest hanging fruit is like the executive assistant piece, like right now for me to book a vacation, right? Like I'm looking for very specific things.

I'm looking for like a nice hotel that has childcare, that has a gym and also, hey, it would be nice to have a sauna also, right? And so I'm going individually through all of those websites and checking that myself. I'm sure that AI right now can do it.

Like ChatGPT has that capability, right? Expand that to my haircut, expand that to, you know, my dental appointment, expand that to making my schedule in such a way that it's like visible to like all my family. I can do those individual pieces right now, but if there was like one, you know, executive assistant to rule them all, right?

Like that would be the piece that, you know, I mean, I feel like the secretary piece of like, you know, any sort of like business person or, you know, that is something that like everybody wants. Everybody wants to have that person that is kind of like a buffer between them and admin work, right? Or like, you know, basic admin work.

I think that's gonna be really awesome for people, you know, for somebody to just say to their AI, oh, hey, can you schedule a doctor's appointment for my daughter because she's sick and they don't have to sit on the phone and wait and get their health insurance information. Like that's gonna be huge for people. So I'm really looking forward to that feature.

Like that's something that I think is gonna be really cool. I just wonder, you know, what are people who are in the industry thinking about this kind of stuff? Like what is your perspective on it?

But I think that we're both optimists, you know, like what are you most excited about with artificial intelligence? I mean, it's just changing so rapidly. I can tell you that mine is the executive assistant piece.

And also humanoid robots. I think that when they're able to like, you know, do stuff that you talked about, like fold our clothes and, you know, wash our dishes independently, that's gonna be awesome for me specifically in my household. But what about yourself?

Like what are you most optimistic about that's coming down the pipeline?

Yeah, I mean, those robots, Fei-Fei Li was just talking about they're coming sooner than not. If you've watched Factor, that robot company that they're working on that, I mean, I think it's a exponential compounding effect because different teams are working on different pieces of the puzzle, right? I'm really excited about different healthcare stuff that I see some huge, you know, potentially game-changing things for, you know, breast cancer or otherwise.

Again, back to a specific project, specific pain point problem. What am I most excited about? I mean, I'll say at CES in 2020, I was on stage being like basic productivity stuff.

That's what I want. If you can help me with my email, I'm down for that. Just basic stuff that we're grinding on.

The problems you mentioned related to the phone calls is a relatively complex one. So Google Duplex, I don't know if you've heard of that product. There's, if you go online, you can go on YouTube and it like does a phone call to phone call.

It's still a pretty sophisticated back and forth sometimes. So tasks that humans can do quickly and easily, most humans, competent adults, can still be very difficult to set up. I'm just gonna put that out there.

But I think the chance for robots to talk to other robots, it may be better to fill those slots of restaurant, you know, eating preferences, time, location, number of people, like these are basic slots you have to do for a restaurant reservation, for example. Other things I'm really excited about. I mean, I'll say I'm really excited about my customer's work.

Like whenever I see a problem, so like food security, especially with climate change coming, I look at the data and I'm like, if we don't figure out food, we're in a really bad place. Like supply chain's failing, you know, temperature, volatility, et cetera. So I'm really excited for some of those like moonshot opportunities that we need a lot of different horses in the game to be able to see the unicorn's output at the end.

And basic things like food, basic things like water around the world is a huge pain point. Some of the diseases. I mean, just, we can talk about any one of these, but like, I really love under, I think it's, or at least if any listener is listening to me like nuts and bolts, like big idea, who's actually step by step, you know, looking at that big vision and making, doing the build.

It is vision, you know, talk is cheap, execution is very expensive. And to be really, really doing the work day in and day out year over year is where I believe we'll see the biggest returns. So that's, I'm excited to be working with people doing the work.

Yeah, no, it sounds like an exciting field to be in right now. Everybody's talking about it and the changes are happening so rapidly and you know, who knows what the potential buys? Like that's such an interesting place to be, I think.

So, you know, kudos to you for giving your insight and getting people inspired about this stuff. I think that that's something that I just really want people to take home. But you know, there's so much negativity and fear around AI and I feel like it's just gonna make our lives better.

I mean, that's just my opinion, but I do think that it's something that will come to fruition. But we're getting close to the end of our time. I'd love to ask you just the general questions that I ask all of my guests.

You know, first off is, you know, growing up, what made you go into this field? Where did you draw inspiration from?

I loved languages. I still do. My family is one of those families that we didn't big, big TVs.

We went traveling. Like that's kind of the ethos. So loving languages got me into linguistics.

Linguistics is a gateway drug into other things, which is how I find myself in tech and AI.

Do you think we're gonna have a universal translation anytime soon? Like in Star Trek when you just put something in the ear and you just talk to people?

It's getting better and better, but we are still struggling with many benchmarks. But speech to speech, text to text. I know startups working on this specifically.

I mean, Amazon like has been crazy with it. You know, they're coming out with this new dubbing feature where you can dub like any sort of video from our intellectual property from another country. So like I watched like a lot of South Korean shows and you know, I'm always reading the subs.

It's such a significant barrier for entry for my wife, but like if they could get it to be dubbed in English, that'd be huge, right? So hopefully that'll be something that comes down the pipeline. But you know, wondered your opinion on it, especially with linguistics.

I mean, I know that there's a lot of sort of intricacies when it comes to language, but do you think that that's something that maybe in 10 years, maybe in 20 years that we'll have?

If people are working on it, yeah. I mean, language is so rich and complex a data set. There's a reason why these large language models have language because it's really rich.

I mean, if one language to one language is lovely, most people are in the world are multilingual. A lot of people are doing code switching. So English, for example, right?

Hindi part of the sentence, English part of the sentence. Right now computers are terrible at this, terrible. You switch languages partway through and then back, don't do that to us, yikes.

So the data sets needed for this, how well they work. I mean, we just need teams working only on that instead of what Amazon frequently does is here are the 30 languages, benchmark it good enough and deploy. So it's just a different way to think about the problem and how to execute it.

Yeah, interesting, interesting. I never thought about the code switching part. That would be a pretty significant barrier.

But next question I think kind of leads into that is like where do you see AI in 10 years specifically in regards to the barriers, right? Because I feel like everybody's talking about all of the energy that's necessary for it, all of the computing power that's necessary for it. But then recently, I'm sure you know about this, but China came out with a model that is significantly less compute necessary, significantly less power that's necessary.

So what do you think is gonna happen? Do you think it's just gonna get cheaper and more accessible? Or is it gonna be like I was telling you in those big industries that people were able to pay for these big capital expenditures to run the AI?

Yeah, I mean, just like any big utility, I see decreasing dramatically because the costs are so high. I mean, I've seen early self-cooling GPUs, for example. I mean, the physical energy of these systems, of these data centers is very high expenditure.

The more people want to be using them, right? People are hitting the maximums on usages. So I think it's really, we're gonna see a big curve, I believe, and efficiencies, there's just, my friend always tells me, she works in climate tech, she's like, if there's a big dollar sign, people will go after it.

So if efficiency has a ginormous dollar sign, there are teams going after those dollar signs. Capitalism is still well, well happy. So if we are seeing it infiltrate into AI being in almost everything, people need access to it.

And so models, data, GPUs will continue. And I think companies like OpenAI I'm worried are very overbloated. So if I were to make my bets on which companies we'll see in five years that we're talking about on podcasts like this, OpenAI might not make that list for me.

Which I tell to some of my family members who are investing too much in Microsoft, not gonna name names. But just really thinking about, as you mentioned, efficiency. Do we see a return on investment?

And yes, most companies will have this as a plugin, as something they're using Anthropic, they're using Perplexity for the future.

That's cool. And that is interesting to hear your perspective on that. Last question, you're in AI, outside of AI, right?

What is the technological breakthrough that, or what kind of technology are you personally interested in that you just like can't get enough of? Like I was telling you for me, for me it's like AI and robots, right? Like this is something that I'm in medicine and I see a robot video coming out from Boston Dynamics and I have to click on it.

I just like, that's what I'm curious about. What about for yourself? Like what is like something that you find yourself kind of gravitating towards?

Gosh, just one? Oh my.

I mean, you could talk about the food, the food issue. Like is vertical farming gonna be the next big thing? What about like medicine, longevity, anything?

Gosh. Okay, yeah, two I'd mentioned off the top of my head. So my customer, I mean, organic is actually a benchmark of amounts of pesticides, not no pesticides.

Sad thing I learned recently. So they use no pesticides. It's beyond organic, high quality, high output produce.

Delicious. They validated 170 types of produce. If the apocalypse happens, you want one of these systems.

You know what I mean? So they're actively scaling. Agri is a really cool company.

Another company I think is super, super cool is my friend working on industrial graphite. So they are literally taking carbon from the air to make, and not graphic like, or graphite like for pencils. I'm talking about industrial graphite, which is a whole other sector.

But we need lots of different ways for climate tech to be working on the problem. Like I think you and I wanna make it past 2060 on this planet, probably. And so we need lots of different approaches.

I saw a startup that's doing seaweed plastics. You know, like, anyway, my friend literally lives 30 minutes away. He's working on industrial graphite.

I'm really excited for the opportunities for his company to literally be taking carbon dioxide from the air.

That's awesome. Yeah, climate tech is so, so different than the majority of technology. Like it's because there's, I feel like a time limit on it, you know?

It's like, you know, a lot of this stuff we're doing just for our own, you know, personal benefit. But like those people are actually trying to get us past the 2060 mark that you're talking about. So super interesting to talk to you.

Thank you for giving us your perspective, Joan. And I really wanna just thank everybody who listens on a regular basis. If you don't mind hitting the like and subscribe button, we'd really appreciate it.

And for those of you guys who are, you know, subscribing already, as always, we'll see you in the future. Thanks, Joan. Thanks, everybody.

Have a great evening.