Shownotes
Fabricio Costa
Bio
Fabricio F. Costa, PhD, MBA is a GenAI, AI/ML, and Data Science expert with over 18 years of global leadership experience across technology, healthcare, and life sciences. He has held senior roles at Apple, Amazon, and Accenture, co-founded and exited two AI-driven startups, and led large-scale programs generating over U$100M in client revenue. At Apple, he managed 200+ professionals and delivered 300+ Apps, tripling market share in key regions. As CEO of a startup named Datagenno, he helped build a platform for rare disease diagnosis that generated U$20M+ in revenue and was acquired for a 5x return. He has formed partnerships with major tech firms and is a trusted advisor on AI & GenAI strategy. A Harvard-trained scientist with 100+ publications and multiple patents, Fabricio was honored at the White House and on AI and innovation.
Intro
Fabricio Costa elucidates the myriad of challenges and missteps encountered during his entrepreneurial journey, emphasizing the critical lesson of refraining from raising capital prematurely. Within the discourse, he poignantly recounts how this initial error culminated in significant equity dilution, ultimately resulting in financial outcomes that fell short of expectations upon the sale of his company. Furthermore, Fabricio articulates the necessity of aligning product development with genuine market demands, advocating for a flexible approach that embraces pivoting when customer feedback indicates a misalignment with their needs. Our conversation traverses his remarkable trajectory from academia to significant roles in esteemed corporations such as Apple and Amazon, culminating in a rich discussion on the implications of artificial intelligence in contemporary business landscapes. Join us as we delve into these insights, framed by Fabricio's experiences that illuminate the intersection of innovation, perseverance, and the evolving dynamics of the tech industry.
Conversation
The dialogue presented in this episode encapsulates a profound exploration of the tumultuous journey of entrepreneurship, underscored by the insights shared by Fabricio Costa, a seasoned entrepreneur and academic. Fabricio elucidates the myriad mistakes he encountered during his inaugural venture, particularly the perils of soliciting capital prematurely, which resulted in significant dilution of equity. He articulates the consequential disappointment experienced upon the eventual sale of the company, where the anticipated financial return starkly contrasted with reality. This discussion serves as a cautionary tale for burgeoning entrepreneurs, emphasizing the critical importance of market analysis before product development. Fabricio's narrative is interspersed with reflections on the necessity of adaptability in the face of customer feedback, illustrating the iterative nature of the startup ecosystem. Moreover, he delineates the vital distinction between 'good stubbornness'—the perseverance to pursue meaningful ideas—and 'bad stubbornness,' which is characterized by an inflexible attachment to one's creations. This episode ultimately provides an invaluable blueprint for aspiring founders, highlighting the significance of strategic decision-making and adaptability in the dynamic landscape of startups.
Takeaways
- Fabricio Costa emphasizes the importance of not raising capital too early in a startup's lifecycle to avoid unnecessary dilution of equity.
- He recounts the numerous mistakes made during his first startup, highlighting the necessity of understanding market demand before product development.
- The conversation reveals how persistence and resilience are crucial traits for entrepreneurs, enabling them to navigate challenges and learn from failures.
- Fabricio discusses the evolving role of AI in the workplace, suggesting that while it may displace some jobs, it also creates new opportunities that require enhanced AI literacy.
- He recounts how his academic journey informed his entrepreneurial endeavors, illustrating the interconnectedness of research and practical application in business.
- The podcast concludes with a call to action for aspiring founders to leverage AI tools to enhance productivity and foster innovation in their ventures.
Transcripts
Hello.
Speaker A:Please meet today's guest, Fabricio Costa.
Speaker A:What kind of mistakes did you make in the first startup?
Speaker B:Well, lots of mistakes.
Speaker B:The first one was raise capital too early, dilute my shares in equity and from the other co founder also.
Speaker B:And then we, we got venture capitalist, and in the end, our equity was so low that when the company was sold, we got much less than we expected.
Speaker B:And then when building the company, the mistakes are you, you, you build something, and that's how you learn.
Speaker B:You build something that you think before you analyze if the market really wants that solution or the product, and you think it's amazing.
Speaker B:And then when you start talking to potential customers, they say, well, well, that's not really what we needed.
Speaker B:We needed X, Y and Z because they have already a mindset.
Speaker B:And then you have to pivot a lot.
Speaker A:What happens when a Brazilian scientist with a PhD discovers something groundbreaking that nobody believes in?
Speaker A:He gets 30 rejections, pivots multiple times, and eventually publishes work that helps lay the foundation for RNA vaccines.
Speaker A:Then he takes that same relentless persistence and applies it to startups, big tech and AI consulting.
Speaker A:Today, I'm talking with Fabricio Costa, whose career reads like a masterclass in reinvention.
Speaker A:From Harvard postdoc to startup founder to executive at Apple, Amazon, and Accenture, he's made every mistake in the startup playbook, from raising money too early to falling in love with his own ideas.
Speaker A:But he's also built something that helped diagnose rare diseases in months instead of years.
Speaker A:In this episode, we dive deep into the AI hype cycle.
Speaker A:Why most job displacement fears miss the point and the difference between good stubborn and bad stubborn.
Speaker A:Plus, Fabricio shares the research story that taught him persistence beats pedigree every single time.
Speaker A:This is designing successful startups.
Speaker A:I'm Jothi Rosenberg.
Speaker A:Let's dive in and hello, Fabricio.
Speaker A:Welcome to the podcast.
Speaker B:Hi, Jyoti.
Speaker B:It's a pleasure to be here.
Speaker B:Thank you for inviting me.
Speaker A:I'm excited to have a chance to talk to you, and I think people will be interested in your very varied background.
Speaker A:But first, let me ask you this.
Speaker A:Where are you originally from and where do you live now?
Speaker B:I was born and raised in Brazil from Italian roots, so I have an Italian citizenship.
Speaker B:I lived in Brazil probably until my mid-20s, 20s, and I came to the US to do a postdoctoral training at Harvard University in Boston.
Speaker B:I think that's where you live right now.
Speaker B:And that's how I started my American, I would say professional career and personal journey.
Speaker A:Yeah.
Speaker A:So speaking of that, you have this interesting journey and what I found when we first met was that we don't exactly parallel each other but boy, we've done a lot of similar steps along the way.
Speaker A:Let's see, you've done academia, you've done startups, you've done big tech tech and.
Speaker B:Then you've, you've done consulting.
Speaker A:Consulting and then you have had this path of going from being a researcher to a lot of work in AI, a lot of work in gen AI.
Speaker A:This is probably a good place to start.
Speaker A:You're just giving us a little bit of a run through of that whole path of you have a PhD.
Speaker A:Oh, I was going to mention that I, when I read that you had a PhD and an MBA, I thought that was a really cool combination because PhDs can sometimes they love being down in the weeds, we love it.
Speaker A:And then the MBA tends to operate more at the strategy level, at a higher level.
Speaker A:And so the MBA is pulling you out of the weeds, but the PhD is grounding the MBA back down in what's real, what's working.
Speaker A:I, I think that's a cool combination.
Speaker B:You got it.
Speaker B:So the PhD and postdoc, after the PhD you are looking at a microscopic problem and you are very narrow and you are very specialized.
Speaker B:You don't know about business.
Speaker B:That's how I got my business acumen and was a long time after I finished my PhD that I did this and I was starting my first, the first company that got a good exit first startup.
Speaker B:And we got into an MBA learning by doing program that this accelerator got us in and I was the CEO of the company.
Speaker B:So I got to learn a lot of things about how to raise money, how to do market analysis, how do your product do, product fit analysis and what analysis.
Speaker B:All of those things that you have to do in business and that was like open your mind to the general things of business and how to the word works and how you get customers and all of that.
Speaker B:So two different words.
Speaker B:But in the end I think the same problems that you have doing a PhD, the frustrations and things are very similar.
Speaker A:The, the problem with learning by doing, I should start by saying learning by doing is great.
Speaker A:It's the way most people who go into startups have to because you, most of us didn't get any training and I was making so many mistakes and stupid mistakes because I just didn't know I came straight out of being a professor and was doing a startup and I suddenly supposed to know what I'm doing.
Speaker A:What kind of mistakes did you make in the first startup Lots of mistakes.
Speaker B:The first one was raise capital too early, dilute my shares in equity and from the other co founder also.
Speaker B:And then we got venture capitalists and in the end our equity was so low that when the company was sold we got much less than we expected.
Speaker B:But and then when building the company, the mistakes are you build something and that's how you learn.
Speaker B:You build something that you think before you analyze if the market really wants that solution or the product and you think it's amazing.
Speaker B:And then when you start talking to potential customers, they say that's not really what we need, that we need the X, Y and Z because they have already a mindset.
Speaker B:And then you have to pivot a lot.
Speaker B:And then there's a lot of work behind the scenes to make something that the customers are going to buy for, are going to license or going to give you money to do pilots or things like that.
Speaker B:So the beginning is really tough.
Speaker B:And I think for entrepreneurs that are starting, I think the beginning first lesson is learn how to raise money.
Speaker B:If you don't need in the beginning, if you have already customers or if you are bootstrapping, don't do the mistake of diluting your equity.
Speaker B:And when you do that, the cap table and all of that, be very alert and be smart to negotiate.
Speaker B:If you really think the idea, you don't need to have customers or have cash flow already.
Speaker B:But I think you can be mission and vision driven that when you are talking to those people, they are not eating you alive.
Speaker B:If you understand what I said.
Speaker B:That's one.
Speaker B:The second one is no, don't attach too much to your idea or to what you are building because otherwise you, if you have to pivot or people criticize it, you take it personally.
Speaker B:So it's difficult to do that when you were young because everything that you are building you are very proud of.
Speaker B:Right?
Speaker B:So yeah, you have to be able to say yeah, this is not well or this is not working well, so let's change it.
Speaker B:So you have to be able to pivot in a very flex.
Speaker B:You have to be very flexible to pivot several times.
Speaker A:Okay, so that's a little bit of an introduction about your startups building something from scratch part of your career.
Speaker A:Let's compare and contrast that to.
Speaker A:And here's an impressive.
Speaker A:To me, this is an impressive list of companies that start with A, that you've worked at.
Speaker A:Apple, Amazon, Accenture, I do you only work, by the way, do you only apply for jobs at companies that start with the letter A?
Speaker B:No.
Speaker B:So Apple the international training program I joined when I sold the first company that had a nice exit.
Speaker B:And since I was exposed to a lot of, I would say a lot of conferences that were investor relations and all of those high tech companies like Apple, Google, Microsoft, Amazon, they were all there listening and I presented that Google, I presented that I have some videos on YouTube.
Speaker B:It's a one minute pitch, which is crazy, but they were like 50 companies presenting at the same time.
Speaker B:So you have one minute to each and all of the investors from Google Ventures and all of those, they have a venture side.
Speaker B:All of those companies, not just the starting with the aa, but all of the high techs and consulting companies, they have a venture side and they are looking for the best thing that is coming up.
Speaker B:At that time in healthcare, they didn't have anything.
Speaker B:Google Health was dying and Microsoft was trying to get into healthcare and life sciences.
Speaker B:All of them are always trying to get into this.
Speaker B:And we came up with a very nice and interesting idea to identify fast patients that have rare genetic diseases.
Speaker B:They are very rare diseases, but in and together they affect a lot of people.
Speaker B:And pharma companies, they are not developing therapies and drugs because they think it's not business.
Speaker B:They are not getting a chunk of revenue.
Speaker B:They prefer to focus in blockbusters or diseases that are complex like diabetes, cancer and I think regenerative diseases, they are called orphan diseases because nobody cares about those.
Speaker B:But the patients suffer a lot.
Speaker B:I think the relative suffer a lot.
Speaker B:I think we built something to accelerate like we what would take 10 years, a decade to help diagnose a disease from a patient in three to six months.
Speaker B:And we were the middleman between life sciences and pharma companies, biotech companies and the patients, the medical doctors that had the patients, the healthcare professionals.
Speaker B:And that was something that today chatgpt if you type signs and symptoms and genes associated, it can give you the response really quickly, maybe hallucinate.
Speaker B: ut it's better than we did in: Speaker B:But at that time those companies were listening to startups that were coming up with ideas like that.
Speaker B:That was Apple, I think Accenture.
Speaker B:Before the pandemic started, I moved to California.
Speaker B:I was living in New York, the east coast.
Speaker B:And I decided to do consultancy full time, build my own company.
Speaker B:And then the pandemic hit.
Speaker B:Everybody was closing their budgets like nobody was spending.
Speaker B:And then I had to find a full time job.
Speaker B:And I knew a lot of people from those consulting companies because of my tracker record of talking to A lot of people growing my LinkedIn network today I have 30,000 connections, which is the maximum of LinkedIn.
Speaker B:And I joined Accenture because of that.
Speaker B:I stayed there almost two years under what they call industry X. I moved to San Francisco and then Amazon was something that I've done with them because they are, they were building, they're still building their Amazon pharmacy business, delivering medications at people's doors, which is, there's a lot of setbacks and privacy concerns with medical records.
Speaker B:And I was helping them in those, with those issues.
Speaker A:So when you were at Apple, you mentioned to me when we first met that this.
Speaker A:You were in a developing country when you were working for them.
Speaker A:Which country was that?
Speaker A:Was Brazil.
Speaker B:So South America.
Speaker A:Oh, it was Brazil, yeah.
Speaker B: e Apple developers Academy in: Speaker B:They wanted to increase the apps in the App Store that were focusing in problems, quote, unquote, that affected those companies.
Speaker B:Was a very social problem.
Speaker B:Wasn't the Table and I diversity and inclusion program that Apple has.
Speaker B:And they were starting something that lots of those high tech companies do.
Speaker B:So bringing back to the community something.
Speaker B:And they needed a leader that spoke Portuguese and English and that could help start the program in developing country like Brazil.
Speaker B:And I started this with one site in the capital of Brazil, Brasilia.
Speaker B:And when I left in, I stayed for five years.
Speaker B:When I left they had 12 sites all over the country and today they have sites in the Philippines, Indonesia, Italy and Europe.
Speaker B:It's not just developing countries anymore in South Africa and Detroit in the US.
Speaker A:So they have that, that's funny.
Speaker A:They have it because I'm from Detroit and you just called it a developing country.
Speaker B:Yeah, no, the beginning was dead.
Speaker B:But then they have, they had the deal with GM and they developed.
Speaker B:They wanted a site close to GMC and that's how the.
Speaker B:I think the University of Michigan, I don't know exactly got a site.
Speaker B:So all of those sites were associated to either college or university.
Speaker B:So the students that were doing any course could come to the program, not just computer science, one year program, learn how to code in iOS develop, they use something called challenge based learning, CPL or PBL.
Speaker B:Problem based learning.
Speaker B:You have a problem.
Speaker B:So the solution is going to be an app.
Speaker B:And they could be three to five students each year developing an app.
Speaker B:And in the end, in five years, I think we, we deployed, we delivered like more than 300 apps.
Speaker B:Because some of those projects will die, would die in the process because of the selection.
Speaker B:And some of Them would pivot and vote.
Speaker B:It's like starting a company, building an app, but then you can pivot along the way.
Speaker B:And I would deal with developers, data scientists.
Speaker B:At that time we had some AI machine learning, AI ML designers, which I love.
Speaker B:I love design.
Speaker B:I think it's amazing to develop the app logos and they would design the app config like the setup, the back end, the front end and then the software engineers and the developers would develop the back end of the app.
Speaker B:And there was fun.
Speaker B:A lot of apps that are still used today, they were developed probably of.
Speaker A:All the people I've talked to, you've got to be the most experienced and advanced in AI research and whatnot.
Speaker A:Might be good to have a short part of our conversation here be about AI.
Speaker A:IT as we all know, it's moving very fast.
Speaker A:I could no longer do what I'm trying to do in my work without my AI assistant in particular.
Speaker A:I'm developing this.
Speaker B:Interesting.
Speaker A:Yeah, no, I'm just, I just paid for one.
Speaker A:Okay.
Speaker A:I just, I.
Speaker A:You.
Speaker A:I have.
Speaker A:I chose to use Claude AI.
Speaker A:I really like that one and yeah, anthropic and we are doing what I'm.
Speaker B:Doing advertising for the companies.
Speaker B:But I don't know.
Speaker A:Yeah, I know.
Speaker A:Hopefully they're happy that I'm saying this anyway that what I'm doing that I need such good strong AI support for and then I want to hear your all your thoughts is I'm building this set of online courses for startup founders, wannabe founders.
Speaker A:I'm calling it.
Speaker A:Who says you can't start up.
Speaker A:It is a set of 60 lessons.
Speaker A:I think I shared with you a little bit about it.
Speaker B:You did, yeah.
Speaker B:And as I feedback about that, but.
Speaker A:Oh good.
Speaker A:Anyway, so I've got the first 45 lessons done.
Speaker A:There's going to be a total of 60.
Speaker A:Each time I'm working on a lesson I get Claude to read my script, correct any errors when I'm trying to create some nice visuals.
Speaker A:Claude is extremely good at generating SVG based images and so they're just, they're great and I'm able to it.
Speaker A:I'm not that good with graphics myself anyway so it would really take me a long time to create these things and so it's just wonderful.
Speaker A:So my point is just to give a personal shout out to how AI is, is speeding up what I'm doing and simplifying my work.
Speaker A:Tell us what you think is being hyped versus what's real and where the opportunities are and I also want to hear your thoughts about are you, do you have any like long term, when you look way out, do you have any fears or concerns about where AI might go?
Speaker A:Hi.
Speaker A:The podcast you are listening to is a companion to my recent book Tech startup Toolkit how to Launch Strong and Exit Big.
Speaker A:This is the book I wish I'd had as I was founding and running eight startups over 35 years.
Speaker A:I tell the unvarnished truth about what went right and especially about what went wrong.
Speaker A:You could get it from all the usual booksellers.
Speaker A:I hope you like it.
Speaker A:It's a true labor of love.
Speaker A:Now back to the show.
Speaker B:That's a great question.
Speaker B:I think everybody's talking about how AGI, right, is going to affect all of us and displace jobs.
Speaker B:It's happening already, but I don't think we are there yet.
Speaker B:I think there is a lot of hype like you said, because all of those companies, and I won't mention names, they want to raise a lot of money.
Speaker B:Let me tell you why.
Speaker B:I read a lot.
Speaker B:So this last year and this mid last year and this year I took half of my day to do a quote unquote sabbatical to learn because I saw this wave coming.
Speaker B:I was using ChatGPT since they launched.
Speaker B:I follow all of the conferences that all of those big techs are doing about what they are releasing, what is hyped, what is usable.
Speaker B:And I think all of those companies, what they fear most is that they need a lot of processing power using GPUs, NPUs.
Speaker B:And now we have, you have TPUs also, not just CPUs, core processing units.
Speaker B:You need a lot of graphics like you said, to be able to reach something like AGI.
Speaker B:So we are talking about trillions of dollars that they need to raise to build data centers.
Speaker B:And they come to developing countries because the water is cheap, the land is cheap.
Speaker B:And I was reading a book called Empire of AI that tells a lot of those stories.
Speaker B:And I think the hype that they do is to raise money and affects the public like you and me.
Speaker B:It's displacing jobs.
Speaker B:Yes.
Speaker B:But all of this user focused AI that we see, I'm sending a prompt to chatgpt so I need to write an email or I need to build an image or I need to build a logo.
Speaker B:That's fine.
Speaker B:Everybody's using that.
Speaker B:Oh, I need to do my homework.
Speaker B:A student, they use all of that.
Speaker B:They can use other tools too.
Speaker B:You mentioned cloud from Anthropic.
Speaker B:There are others people that came out of Chat OpenAI chatgpt that built those companies and all of them are raising big amounts of the money and I think the user focused like AI or Genai generative AI with retrieval, augmented generation that you use prompts.
Speaker B:I think there's a lot of hype.
Speaker B:I don't think AGI is closed, it's displacing a lot of jobs.
Speaker B:But you mentioned AI assistant so that's how I think that this big impact is going to be.
Speaker B:So AI agents you mentioned like using an AI assistant for you but enterprise, they can decrease the cost and of workflows and they can use AI agents orchestrating or autonomous agents.
Speaker B:They are already using that.
Speaker B:I see Salesforce using Agent Force to help inside and they are 70% of what they do is being done by an agent, one agent or multi agents.
Speaker B:So I think what is going to be displacing jobs really are those agents and not necessarily they use large language models which ChatGPT uses, Claude uses.
Speaker B:They can use smaller amounts of data that are inside companies to help the company in some processes or workloads.
Speaker B:That is displacing jobs.
Speaker A:Yes.
Speaker B:So I think there's a big impact in the future of the companies.
Speaker B:People are scared because low level or like easy to do jobs are going to be displaced.
Speaker B:They are being displaced.
Speaker B:That's the fear.
Speaker B:But I don't think like writers from journals or writers from books are going to be displaced.
Speaker B:I think there is always the creativity that humans still have.
Speaker B:I don't think we are there yet.
Speaker B:So my bet personally after reading a lot and seeing all of those presentations is the that AI agents are going to be something that is going to be pursued and it's going to be sticking around.
Speaker B:But I don't have an idea on how those big corporations like OpenAI, it's not a corporation, it's still a nonprofit because they're struggling to transition.
Speaker B:I think Elon Musk is suing them.
Speaker B:There's a big fight.
Speaker B:But I think anthropic with cloud and others.
Speaker B:I think they are betting the end user which is everybody that it's using.
Speaker B:And I think Gemini from Google for example in their phones, Google Pixel.
Speaker B:I saw the release last week and they are good but they are just for us to write emails and to do things on in our phones or to select images and trying to figure out or is this are those clothes good at me if I'm buying them at a specific store?
Speaker B:Yeah, there's a lot of displacement of jobs, there's a lot of changes.
Speaker B:But I think justifying the hype, it's all about raising money because of the energy that they have to use to process the prompts and whatever you are asking the water to cool the systems they are talking about using nuclear energy.
Speaker B:Right.
Speaker B:And the data centers, they have to build a lot of data Centers, not with CPUs but with GPUs.
Speaker B:And because it's all about parallel computing, it's not just core processing units anymore.
Speaker B:I don't know if I was too technical, but that's how I see it.
Speaker B:The hype is all about, oh, we need to raise money, a bunch of money, billions of dollars to make this thing going.
Speaker B:But they don't really know where this is going.
Speaker A:I think there are if I'd like to read some analysis of it.
Speaker A:But I do understand how the job displacement of low level jobs in another area that you didn't mention of low level jobs that are being displaced all across Silicon Valley is programmers.
Speaker A:Because these things are very good at doing the kinds of things that entry level programmers would do.
Speaker A:Yes, junior programmers, but they're not able to do design and architecture of good software systems.
Speaker A:They're far away from that.
Speaker A:I think that like most times when there's a big disruption in the industry and society in the, in the economy, there's a shift that there's still the same number of jobs needed or people needed to fill jobs, but they're being moved from one area to another.
Speaker A:So it will hurt some people.
Speaker A:Yes, but it's not.
Speaker A:So in other words, what I'm trying to say is that AI is also creating jobs and.
Speaker B:Yeah, and what's at the same time that is displacing some jobs.
Speaker B:And it's good that you mentioned that software engineers that out of college they are struggling to find jobs, jobs that won't spay half million dollars a year for coding like they are not there anymore.
Speaker B:AI can do that like quite easily and without doing mistakes.
Speaker B:It's better.
Speaker B:It can even correct old codes that people did.
Speaker B:That's exactly what you said.
Speaker B:But I think I agree.
Speaker B:I think the job market is going to shift and AI literacy, not you don't need to know backend or how it's working.
Speaker B:Literacy is like how to use this to make my workload fast better, cost effective, time effective and things like that.
Speaker B:I think that's where we are moving.
Speaker B:And then people can do creative things on the side.
Speaker B:They have more time to do other things.
Speaker B:That's how I see it.
Speaker A:Yeah, I agree with you.
Speaker A:And I think that just, just on our own Little tiny microcosm of my family.
Speaker A:Suddenly people, everyone in the family is discovering one of the two, ChatGPT or Claude and it's enabling them to do things they couldn't do before, including things that are making them better at their job and do more at their job.
Speaker A:And so that's really creating opportunities because somebody else can come in and say oh now I couldn't have done this kind of thing before but now I can.
Speaker A:And, and so it's pretty interesting.
Speaker B:Interesting.
Speaker B:And it's scary because I presented the Chat GPT thing to my dad and he was like at first he didn't understand and now he's using and he likes it because oh, I don't need to do nothing against Google.
Speaker B:I don't need to do a Google search.
Speaker B:I can try to search whatever I need or a recipe or something using the tool that Gen AI is offering.
Speaker B:But at the same time, like I said, there's still figuring out subscription alone.
Speaker B:It's not going to make them money forever.
Speaker B:I think the business model is struggling still.
Speaker B:I don't know.
Speaker A:Yeah, business models are very hard to come up with for new kinds of products.
Speaker A:I'll tell you though, and it's interesting you said that because my wife has pretty much stopped using just plain Google searches and she only finds things using ChatGPT and you're right, we ate a dinner last night based on a ChatGPT recipe.
Speaker B:Recipe, yeah, yeah, no, but tell her that Google has Gemini which comes from Google DeepMind that they bought.
Speaker B:It's pretty good.
Speaker B:So it's embed on search.
Speaker B:So if you do a search now the first thing that appears is like a description what of what you are asking.
Speaker B:Not just random pages and then they have the index of random pages so they are getting better at it.
Speaker B:And Pixel 10 that was just released, I saw the presentation Jimmy Fellow was doing the presentation was a little funny in New York and you can do a lot of things with that phone And I'm an Apple user.
Speaker B:Everything that I have, it's Mac based and I think Apple is lagging behind because they are not catching up.
Speaker B:So there is a struggle and there's a, I would say a competition that is good for the customers.
Speaker A:It doesn't help that Mark Zuckerberg is paying millions of dollars for in salary to hi to hire AI experts.
Speaker B:But most of it is stock options.
Speaker B:It's not cash.
Speaker B:Yeah, of course there is millions in cash if they bring people in.
Speaker B:But they have a vesting time.
Speaker B:They have to stay to get A hundred million dollars, sometimes 200.
Speaker B:It's a lot.
Speaker B:Some companies that you sell don't even get that.
Speaker A:I know, I have, I wanted to ask this question.
Speaker A:So it's my, it's my wrap up question.
Speaker A:Okay, so you've, you've done a lot of things.
Speaker A:You've been a scientist, you've been an entrepreneur, you've been an executive, you've started things, you've sold things.
Speaker A:And of course consultants.
Speaker A:You always remind me that.
Speaker A:Don't leave out the consultant piece.
Speaker A:Almost all of those things are hard and take a lot of grit.
Speaker A:And I love asking people to tell us the story of where you think your grit came from because you can't do startups without it.
Speaker A:So you know, you, at least there you had to have it and you probably have had to have it in many of the things you've done.
Speaker B:That's a great question.
Speaker B:So I have a nice story in academia.
Speaker B:When I was starting my PhD advisor, she was a lady, very nice lady in Sao Paulo, Brazil and I was working in cancer research, which is something that I wanted because I think it's mission driven.
Speaker B:And I was thinking that I was going to be an academic professor, but then I saw that was much harder and there was a lot of politics, but that's another story.
Speaker B:So at that time I did something in my PhD.
Speaker B:The PhD, you have to find something that nobody found in that specific field and then publish, right?
Speaker B:So you can defend your thesis and then you have a board that is going to say yes, passed or no.
Speaker B:So I did all of that.
Speaker B:But in the process of doing that we found something that.
Speaker B:I don't know how to put it technically, but in science there was something in genomics that was emerging.
Speaker B:The Human Genome project was happening at the time and they were finishing it and they said that the human genome has 25, around 25,000 protein coding genes, which are proteins that are the building blocks.
Speaker B:But the human genome has 3 billion base pairs of nucleotides, which doesn't make sense.
Speaker B:So it's too much.
Speaker B:They call the junk DNA to produce just a few proteins.
Speaker B:So they started to see areas in the genome that were transcribed, they would become rna.
Speaker B:And probably, you know, because of the RNNA vaccines against Covid, they would become RNA that wouldn't become proteins and those RNAs would call non coding RNA because they didn't code for proteins, they were just RNAs.
Speaker B:And at that time I got really curious about that, what's going on and I started reading a lot.
Speaker B:I finished my PhD, I came to the US with this idea.
Speaker B:And I told my PhD advisor, there is something here that we found that doesn't make sense.
Speaker B:We are not publishing it.
Speaker B:We are putting the drawer, but I want to understand what's going on.
Speaker B:So the grip came when I started eating, reading and reading papers about this because there was nothing, there was not a lot of papers.
Speaker B:And we are talking about a very huge constellation of scientific publications that are being published every day or every month or every week.
Speaker B:So I grabbed everything, I started printing and reading and I saw something and I told my advisor, I'm writing a review about this because I think there's something here and I have enough information to build on a hypothesis not based in bench lab, but based on what I saw people describing.
Speaker B:And she told me, she said, reviews are not for everybody.
Speaker B:You have to be a professor, you have to be well known in the field.
Speaker B:You have to be X, Y and Z.
Speaker B:You're not publishing anything.
Speaker B:I'm sorry, I'm not participating.
Speaker B:I said, I'm going to try, right?
Speaker B:So I wrote the review.
Speaker B:My English wasn't very good at that time and I had to rewrite several times, like pivoting a lot of times because constantly new publications were coming out.
Speaker B:And I came to the US and the guy on Harvard that was my advisor in the postdoc, he said the same thing.
Speaker B:I said, no, I want to do this.
Speaker B:Can you, can you help me out?
Speaker B:And he said, I'm going to read to you, but I don't think this is going to go anywhere.
Speaker B:So what I've done was I wrote the review.
Speaker B:This was when I was 26 years.
Speaker B:And I said, I'm going to publish this anywhere because I think it's important.
Speaker B:So I started with the top journals like Nature Genetics.
Speaker B:I sent a letter to the editor, I did all by myself.
Speaker B:I got like more than 30 rejections.
Speaker B:Like, I started from the top to the bottom.
Speaker B:And then once a journal called Chin, which has.
Speaker B:It's a very respected journal, but doesn't have a lot of impact in the scientific community.
Speaker B:A Japanese editor said, send me the manuscript, I'm interested.
Speaker B:Then I said, that's the first one that asked for the manuscript.
Speaker B:So I sent it.
Speaker B:And after a week he sent.
Speaker B:He replied me back and said, I think it needs some changes, but I like it.
Speaker B:We are going to send it to review, but you have to change X, Y and Z.
Speaker B:And then I did the changes I sent.
Speaker B:And the revision process was very laborious because all of the reviewers, they were like this is going nowhere.
Speaker B: it got published on June, in: Speaker B: Sorry,: Speaker B:A long time ago now.
Speaker B:Remember when I got the acceptance letter, I said persistence is something that you have to have anywhere.
Speaker B:And this was when I was still in academia.
Speaker B:I think you mentioned grips, and then nobody thought that I could do that.
Speaker B:It wasn't a big impact.
Speaker B:But then after that, a lot of papers came out.
Speaker B:Was a bunch of papers, even Nature, discussing the same thing that I was telling.
Speaker B:I'm not saying I'm gonna win a Nobel Prize because I'm not in academia anymore, but I'm proud that I got into something that a few people were just scratching the surface.
Speaker B:Even though I didn't do any lab work.
Speaker B:I published a review that was combining everything that they have at the time.
Speaker B:And I still follow and I still write reviews today about that because I think it's something that is changing.
Speaker B:RNA vaccines came from this micrornas and non coding RNAs.
Speaker B:So that's something that I'm proud of.
Speaker B:But it didn't get recognized by the scientific community.
Speaker B:But some other people are getting a lot of recognition because of that.
Speaker B:I think that a nice story.
Speaker A:I think it's a great story.
Speaker A:And you said something I totally agree with, which is that persistence is definitely a component of grit.
Speaker A:But another thing that you exhibited throughout this whole story was, I guess the most friendly word for it would be determination.
Speaker A:But another word for it is stubborn.
Speaker A:And those are good.
Speaker A:And those are both good traits as well.
Speaker B:There is good stubborn and bad stubborn.
Speaker A:I know, and I was referring to it.
Speaker A:That's why I said the word determination first.
Speaker A:Because it's good stubborn.
Speaker B:It's the good stubborn.
Speaker B:Yeah, I agree.
Speaker A:Let's do this, Fabricio.
Speaker A:Let's end on that wonderful story.
Speaker A:I want to thank you for being here and this has been a fantastic 40 minutes that I think people are going to love.
Speaker B:No, I didn't even see the time going.
Speaker B:I appreciate and I thank you very much, Jyoti, for the opportunity.
Speaker A:You're welcome.
Speaker A:And thank you.
Speaker A:Here's your founder's toolkit from today's conversation with Fabricio.
Speaker A:First, don't raise capital until you absolutely need it.
Speaker A:Fabricio's biggest regret was diluting equity too early.
Speaker A:Bootstrap as long as possible.
Speaker A:Get paying customers first, then raise from a position of strength rather than desperation.
Speaker A:Second, practice.
Speaker A:Good stubborn, not bad stubborn.
Speaker A:Good stubborn means persisting when you believe in something important, like Fabricios 30 journal rejections.
Speaker A:Bad stubborn means refusing to pivot when customers tell you they don't want what you're building.
Speaker A:Know the difference?
Speaker A:Third, AI won't replace creativity, but AI literacy will become essential for every role.
Speaker A:The jobs being displaced are routine and predictable.
Speaker A:The opportunities being created reward people who can use AI to amplify their unique human capabilities.
Speaker A:Learn the tools.
Speaker A:Don't fear them.
Speaker A:That's your toolkit.
Speaker A:Now go build something that matters.
Speaker A:That's our show with Fabricio.
Speaker A:The show notes contain useful resources and links.
Speaker A:Please follow and rate us@podchaser.com designing successful startups.
Speaker A:Also, please share and like us on your social media channels.
Speaker A:This is Jothi Rosenberg saying TTFN Tata for now.
