top of page

Artificial Intelligence for Startups: Privacy, Web 3 & Angel Investing with Arthur Root

Updated: Apr 17, 2022

By: Thoughts of a Random Citizen Podcast



Today, I am joined by Arthur Root, an angel investor, CEO, and co-founder of a tech company called Nostra, that utilizes artificial intelligence.


Thank you. Yeah it's fun over here. Thank you for having me, I'm excited to dig in.


Sweet. I know if we go all the way back, your entrepreneurial journey starts somewhere with tie, dye, socks in middle school and tech companies in high school. Can you tell us a bit about the path leading up to becoming the CEO of Nostra in between or outside of the things I just mentioned?


That takes me back. Wow, you did your research, well done. Yeah, my first company I started, well, I wouldn't exactly call it a company. It was more of a venture if you will, with my, longest friend, Willie. He, and I just, tie dyed, some socks created an eBay account.


We were in fifth grade, maybe, I don't know how old we were, and we sold 'em on eBay. We packaged them, we probably made like a hundred bucks. Some small amount of money, but that was the start. I had lemonade stand, I got kicked out of the science lab a couple of times for trying to create some stuff, just who I was as a child and it's funny because that's what genuinely gives me happiness.


Being out there and trying to figure stuff out on my own, I always am curious. My parents must have thought, what is this guy doing? Then like, oh, well, it's probably good. I grew up in San Francisco. That was my middle school, which is not a normal middle school.


Then in high school, I worked at a technology company, a venture backed software business, and I'm a pretty lazy person generally so for the first two days, I had to do some stuff that was pretty manual, like manually scraping the internet. I was like, this is stupid. I worked with a buddy, and I friend and I just built a software that automated my internship. Then I was there for three years.


I am an awful engineer, transparently, like that is not my skillset. But I do love what software can do and what technology can do.


That was when I realized, wait, this is awesome. I had a full job where I was supposed to work 40 hours a week, and now I finished 12 weeks of work in three days. That's it and I'm going to go play ping pong and eat M & M's.


Wow. That's incredible, man. I wasn't even considering that in high school at that point, so props to you for that. Before Nostra, can you walk us through other jobs and roles that you had after that internship?


Sure. Yeah. So I started a company called Flash Trip in college. I graduated in a pandemic, so the company didn't do too well because it was a travel company and that was tough. I still had some money left over from the investors, so I spent six months just interviewing chief marketing officers.


I actually created a fake podcast, YouTube channel, and would say, "Hey, you want to come on and talk about yourself and talk about your problems?" Low and behold, these super, sophisticated, smart, CMOs would come in and tell me what to build. Eventually I signed a couple of them as clients brought on an engineering team raised some capital and built the product. So that was the story. Then fast forward a bit and that's where we are now.


Wow. Speaking about that, raising capital specifically, I know now you're currently involved in a few different angel investing ventures, Corsali, Ponto, and Glimpse to name a few. First off, I wanted to give you the chance to speak on those if you wanted to, but I just wanted to dive into what, the perspective of an angel investor is for anyone trying to raise capital.


I probably know more about this from an entrepreneur's perspective than an investor's perspective, but I've mainly just invested in people.


When I invested in Ponto, I legitimately did not understand what they were doing. Will is like one of my close friends and is just one of the best CEOs, best fundraisers, can tell a story, he just traps people and these stories. Then people just want to follow him to the end of the world, which is why, I invested, maybe a year or so ago.


The company's valued at probably 10 or 20 X, what we invested at, something crazy like that. It's just because Will, is such a good storyteller. At the end of the day when I'm investing it and again, I would invest in Will to do anything, it's more about, hey, will this person be able to build a business?


I don't really care what the business does. Let's be honest, it's the guy. The product hasn't launched yet. Who knows what it'll become. Then, you know what he did, he brought on just an incredible team, all stars all around and then he attracted future capital. This team is just, probably one of the best startup teams in the world. He originally got to build this great team, which then attracted more investors to build an even better team. They'll be launching a product soon, he's just an incredible CEO.


Then Anna is another close friend of mine, and she's just one of the smartest people, most considerate smart people in the world. She said, Hey, I'm starting a company, it was in the AI space, so I did understand what she did this time. She's just, she's one of those people that, there's almost no way that she loses, in my opinion. Will is probably the more likely person to build a trillion dollar business and Anna's going to build a very good business and it could be worth a ton of money, but one way or the other, she just, she will figure it out and build something. She's got an incredible team as well. I think that investment's doing well. She's also just very down to earth, listens very well and can do it.


Then Glimpse, another company that's doing incredibly well. Akash, the CEO is again, just a close friend. Just like Anna and Will and is just incredible. I know what he does, it's placing products in Airbnb’s, and it's a good business idea, but at the end of the day, I just believe that he's somebody who I would want to work with, and he will figure it out.


He'll be able to attract future investors. So, to sum it all up, I invest purely in people. I don't really care what people are doing. It's " is this person going to be able to attract future investors, is this person going to be able to attract an incredible team and customers" and that's about it.


So it comes back to that, not what you know, but who, kind of thing. Cool. Moving on to what you specifically do at Nostra. Can you tell us a bit about how or what Nostra is, first off, and then how you've integrated AI into your business?


Sure. We're an AI business so from day one, it's been, AI. What we do is based on who's visiting a website, we'll allow our customers to upload a bunch of different images, a bunch of different text variations for like descriptions calls to actions, headlines, customer testimonials. Then in real time, based on who's visiting a website, our artificial intelligence will determine, all right, this person came in from this Google app and has these attributes, okay, we should show image three, call to action four, headline seven, customer testimonial three. Then in real time, our AI will make that determination. And then our software will build an entire website that is optimized to that individual, so that they're most likely to do that desired action.


Nostra logo with Arthur Root, artificial intelligence website design

Wow. That's incredible. So, the websites change, depending on who's viewing it, like from the second that they click in on it? That's crazy. I'm curious, to know how the back end of AI works. Is it just super advanced coding? Am I completely wrong or can you elaborate on how that works in layman terms?


Sure. It's not as complicated as you think. Imagine, you can almost think about it like a stock chart. If you look at a stock chart or no, maybe a horse race is probably a better way to do it. We can see five horses start, and then let's say there are five different combinations about that.


Now we're doing this with millions, but let's just say it's five horses. Two of 'em start to pull away. When two of 'em start to pull away, our AI will say, "Hey, we should probably send more content to these two people that pulled away." What do they have in common? What do they not have in common? Okay. Out of those things that they have in common, they don't have in common, we'll start creating more and more, we'll start sending more and more traffic there.


If no horse is catching up, we start sending less and less track happen to those people. Now imagine if that horse races with, call it, 12 million different horses, it just gets better and better over time. It sees the similarities between all the different horses and in real time, it's making these determinations.


In theory, it's nothing that a human couldn't do if they just had unlimited hours and they were very good at stats. It's just in reality, humans can't compute millions of combinations of content in 30 milliseconds. So, it's mainly just the speed of learning.


That's what I gathered from it. I was reading one of the posts that you had and for the marketing aspect of things, how is AI different than A/B testing? I think that's what you're getting at. Then, if you could explain what A/B testing is for everyone.


Why don't we start with the basic of A/B testing. Some people call it split testing, AB testing, whatever it is testing if one page, for websites, but it also happens in ads and a lot of different places if one page performs better than the other. Essentially what you'll do is you'll send 50 per percent of your traffic to one page and 50% of traffic to your other page to simplify everything and you wait to see which one's better.


After 5,000 visitors, let's say Page A gets 2,500, Page B gets, 2,500. Page A converts, call it a thousand, Page B converts, a hundred, then what they'll say, Page A is the winner. We then scrap Page B and we only send people to Page A.


That's the basics of, of A/B testing and it's been an effective strategy, people have built huge businesses off it and there's nothing wrong with doing that. With that said there are significant downsides, its incredibly time consuming to wait or to build out each of those different pages, number one. Number two, it takes a lot of traffic and it's hard to just attract more traffic to get more answers. Number three is, what if in January Page A is better than Page B, but in February, page B is better than Page A, but you ran the test in January. How are you going to know?


There are a lot of like little things that when you dive in and you look, you take the shovel a little bit further down. Not that it's a bad system, it's just not the best system. Whereas with AI, what you can do is, you can upload all your content all at once, all your images, all your headlines, all of the stuff that you think a customer may want and then the AI can really make it so you don't have to do anything anymore. Pretty much you set it up and you press and go. Then the AI makes those bets. If it's a 51-49 bet the AI will place that 51-49 bet. If you place enough of those bets, it gets to 60, 40, then you keep on going and it gets better and better, your conversion rate.


So, with 5,000 visitors, you might be able to test out a hundred thousand or 500,000 different combinations of content. You just couldn't do that if you were running an A/B testing. Then when you add on top of that, that, visitor sentiment shifts, significantly, like if we look at the pandemic.


Can you imagine if you had, all these A/B tests that were running and then just three weeks later, like the entire years of testing, is now just all useless. Whereas AI is always testing. They pick up that sentiment is shifting, and they make those changes on the fly to the point where AI based websites performed 5-6X better than non AI based websites during the pandemic.


Nostra does all of this, I know that you even take some of the fonts and adjust it per user. Is it one of those things that the longer the user is on the website, the better results it'll get, or do you guys plug that back in, into, social media? Or how do you best promote to the individual user as the website changes as they use it.


As they use it, there are different paths that someone can take to convert to a customer. For example, someone might come in from Facebook ad one, go to read three different blog posts and then schedule a demo.

Or someone might come in from Facebook ad seven, and then immediately buy something. Who knows what it is, whatever it is, we will be able to start seeing what those paths are and telling you, "Hey, this is what's happening on your website" and then you can say, "oh wow, that's an interesting funnel". I didn't know that was happening. Maybe we should start diverting more traffic to that funnel because that's performing well.


It's an age old thing, which is like attribution and it's the hardest thing to do. AI is t not completely there yet . Transparently, it's still a work in progress, but it's getting better and better. We're all making progress toward getting toward that perfect attribution so that eventually all you have to do is, upload your or content once for your website, for your ads, and then, upload your budget and we'll start getting you customers, democratizing, hyper growth.


All these companies in Silicon Valley where I spend a lot of time are, spending millions and millions of dollars on growth, and our goal is to make it so that anyone can do that. You don't have to have a sophisticated growth engineering team. You can just use Nostra and pay a monthly subscription and grow your business without having to grow your marketing overhead as well.


That's like music to the ears of all of the small, web based, , entrepreneurs out there. Can you elaborate a bit on the differences between personalization and customization and how businesses need to be aware of that in an online world?


Yeah. When you're thinking about. Personalization. You're thinking more about Hey, how do we talk to this individual, individually and how do we do it in a way that is automated where they're not having to do much. Whereas customization is slightly different where it's okay, I need I as a customer am going to adjust these filters so that I can have my own custom experience.


If you think about it, personalization is more doing it behind the scenes, helping the customer without the customer necessarily knowing they're being helped. Whereas customization is very manual and it's like saying, "Hey, I wanna only look for flights after 7:00 PM at, flying from New York to San Francisco", something like that.


Okay. So does that mean that the very first time a customer comes onto a new site that Nostra is utilizing. Are they going to see from the start different content than would someone else potentially?


Very possibly. Yeah, we will show them based on what information we have on them and a lot of the times we don't have any information on someone. So we're like, Hey, we don't know who this is, but we think, with a question mark, this is the best, this is the most likely case that they'll convert.

So then how does your software build on people that it doesn't know anything about?


That's its own audience then. So, the unknown is a known. It's to say, Hey, we know that 10,000 people a month. We don't know anything about them, but if we show them this combination of content, that's the most likely that they'll convert. We don't know who they, why they are doing this. Whatever the hell going on, but this this is the most likely that they'll convert. And we go from there.

Okay. Cool.


With all data and privacy and everything like that, , does your company have like any policies or anything to implement and prevent polarization of perceptions or, how is this whole tech space approaching, this new conversation that's being had.


Yeah. So, we avoid, media companies the way I think of it. It's like people like to be sold to when it comes to clothing, when it comes to, B2B software’s, there's nothing, know, people want to be sold to and they want to get the right thing.


We avoid any political, things like that, just cuz it's we think it's a gray area that we just don't need to be in. If what I mean? it's just like, why, would we be there if there's so many websites with so much potential? Even if you look at Facebook, like we're talking about like maybe half a percent of their global revenue is like politics. If I were them, I just put why do I even allow this, all the headache that it causes? For us, it's just there's so much lower hanging fruit. We just avoid things, topics that are controversial.


Yeah, absolutely. I guess going back to that music for everyone's ears, being an entrepreneur, in building their business through their website. Who are you finding Nostra is helping the most and meaning like, what business' is your tech specifically applicable for at the moment?


We look at websites that are, know, have around 5,000 to 20,000 visitors to a page per month. Those are the, and actually I shouldn't say that there's some that get hundreds of thousands that we can help, know, millions actually, but over 5,000. visitors a month to a given page.


That's when we can really be effective. It's not to say that we can't for the smaller ones, sometimes it'll just take a little extra time for our AI to train . There's nothing wrong with that. It's just, know, you have to be a little bit more patient. We've got one customer, super nice people, great cut company, super noble mission. They get 300 visitors a month. It's gonna take us a number of months to actually provide value to them. So that's where we're focused. A lot of our customer base we're looking at, e-commerce businesses, software companies marketplaces, those sort of businesses.


Okay. So jumping back and forth again, back and forth, to the privacy a bit. I'm really interested in like the regulation side of things. Are you seeing it becoming more difficult for Nostra, in regards to just the way new laws are coming out? I just curious on my end, how, it's, all the time everywhere, how is a company like yours dealing with all of the stuff.


We're in a very fortunate position that we were built after the GDPR was released after the CCPA was released. And so we have been from day one, very privacy conscious What that means is, we can't, we, we don't use any third party bookies, which like technically right now are still allowed.

But the general way that the world's going is that they won't be allowed. We avoid, trying to work with data brokers and third party data sources that are from sketchy areas. It's just, when you're looking at a $20 million fine or 4% of global revenue, It's brutal, like a $20 million fine would crush us.

I don't know what I would do with that. It's pretty cut and dry. We are very conscious of what's going on and, quite frankly you know, I think that it's a good thing for, the world and the privacy regulations are going a long way to, move the business world in the right direction when it comes to data privacy.


I completely agree. Just to show that I have no idea about the privacy laws, GDPR, CCPA? Could you