People.ai CEO Oleg Rogynskyy

"I think sales as we see today is not going to exist. In reality, we're seeing this with kind of PLG and Stripe where more and more or bigger and bigger sales transactions happen literally through APIs."

"While in the U.S., the outlook on this is the 10-year thing, two-year thing, make peace, and off we go to the next election. In Ukraine, it's like, this is the one time when we are actually successfully holding them back."

Oleg Rogynskyy (00:00.304)

Yeah,

Auren Hoffman (00:01.4)

We're getting sk - Okay. Okay, that's what I thought. The two Y's kind of messed me up though, but okay.

Oleg Rogynskyy (00:08.188)

I had to use ChartGPT to see if anybody else had three Ys in their last name and I'm one of like four people globally.

Auren Hoffman (00:15.564)

All right, here we go.

Hello, fellow data nerds. My guest today is Oleg Ruginsky. Oleg is the CEO of, sorry, I just got all messed up here. My bad. Let me just, okay, here we

Oleg Rogynskyy (00:30.384)

No worries.

Auren Hoffman (00:34.222)

All right. Let's try it again. Hello, fellow data. My guess is Oleg Griginsky. Oleg is the founder and CEO of People .ai, a AI powered revenue platform. He's also an advisor to the Ministry of Digital Transformation for the country of Ukraine. Oleg, welcome to World of Dats.

Oleg Rogynskyy (00:52.58)

Alright, thanks for having

Auren Hoffman (00:54.19)

I'm really excited. Now, People .ai was founded like eight years ago, 2016, way before this kind of current wave of AI hype that's kicked off. What did you think AI capabilities would be in 2024 when you started the

Oleg Rogynskyy (01:10.972)

So that's interesting. There's a couple of things people don't realize about people AI. First of all, we were in the same YC batch in a couple months after incorporation as scale AI and open AI. So there was definitely something happening in summer 2016 around understanding of what AI capabilities are going be in the future. Second piece, I just had an interesting discovery related to podcasts. I recorded

a podcast within recent Horowitz in a summer 2018 where we talked about back then we didn't call it language models, but these AI capabilities that will be able to transformers, they'll be able to translate large amounts of unstructured data into human -like outputs and will completely change the interfaces that we are interacting with to chatbot style interfaces. So back then, was the vision. The vision was

Auren Hoffman (01:51.694)

Transformers.

Oleg Rogynskyy (02:10.428)

activity data of salespeople contains the signal to making salespeople more productive and effective because it literally is a step -by -step telemetry of how you've done every deal in the past, successfully and unsuccessfully. And therefore, you should be able to use AI to augment humans initially and eventually kind of do full automatic selling process. So that was the vision all along. It got there.

It got there pretty fast, much faster than I expected. I actually have one of my fundraisers index, think Sirius Biva, say we're going to get there by 2030. And then so that's that's that was the situation.

Auren Hoffman (02:49.136)

huh.

What and when you and if you had to like predict out like eight years from now, is it just like things kind of get better and better at some sort of linear or exponential rate or is there some other like capability that we can't really think of that you think is going to be happening eight years from

Oleg Rogynskyy (03:10.534)

To be honest, I think the world is going to change completely, especially when comes to sales in eight years. In fact, I'll throw in something controversial there. I think sales as we see today is not going to exist. In reality, we're seeing this with kind PLG and Stripe where more and more or bigger and bigger sales transactions happen literally through APIs. And so if you think about this, humans,

Auren Hoffman (03:35.81)

Yeah. Yep.

Oleg Rogynskyy (03:40.844)

service translation layers on the buyer side as someone who understands requirements of your organization, summarizes the requirements of your organization to what needs to be acquired to improve the capabilities of your work. And on the seller side, there's a human's other translation layer between what your product can do and what the buyer wants. We've seen PLG, good product marketing, more and more of kind of self -serve capabilities.

Auren Hoffman (04:02.221)

Yeah.

Oleg Rogynskyy (04:09.584)

be able to translate more and more complex requirements on the buyer side and offerings on the seller side to kind of happen without.

Auren Hoffman (04:15.808)

these are like, some ways attacking the classic inside sales, right? Like they're maybe lower dollar amounts, those types of things.

Oleg Rogynskyy (04:20.932)

Exactly.

Oleg Rogynskyy (04:25.962)

Yeah, so that has been happening even before AI. You could buy more and more expensive things online up until you could buy a Tesla without talking to a

Auren Hoffman (04:34.082)

Yeah, I bought a car without talking to anybody and it just kind of appeared in my driveway a few weeks later. It was amazing.

Oleg Rogynskyy (04:41.222)

There we go. what language models brought into that equation is they completely like you don't need humans translating requirements into a conversation or capabilities into a conversation anymore. It can be done with language models because it'll take us a couple of years to get there fully. But I do see the world in three to five years where humans only do the human to human part, like go for dinner once in a while, etc.

The rest of the work is being done fully with a language model, talking to a language model. It's going to be that simple.

Auren Hoffman (05:17.688)

Got it. So you're saying not only is AI going to replace the salesperson, AI is going replace the buyer.

Oleg Rogynskyy (05:24.122)

Yes, and AI is enabling us to move more and more of the sophisticated large transactions into fully digital kind of like API level handshake

Auren Hoffman (05:38.668)

Yeah. Some of the.

Oleg Rogynskyy (05:40.592)

You will be able to buy enterprise software or an airplane or whatnot the way you bought a Tesla.

Auren Hoffman (05:50.85)

Yeah. The thing when I bought a Tesla, though, I was very familiar with the Tesla. had a I've seen my friends Teslas. I understood what it was. I understood the brand pretty well. I had a sense of the dimensions and whether it would fit my garage like all the like I had a I kind of understood those things like how how when you're buying something in a company like there's all these weird

Like, well, does it even integrate with certain things? I don't even maybe even know to ask that question until like I try it out or like there's a look and feel like a UI of like a user's using it though. Maybe, maybe users won't use software as much in the future. like how is that going to

Oleg Rogynskyy (06:39.868)

So, and that's a great transition to the conversation about data companies versus AI companies here. What's going to happen is you need very large training data sets or fine tuning data sets, the language model, the intuition around how every possible feature of a much more complex sales process works. Just like, mean, Tesla has about 50 features that are variable. Let's say ERP software has 5 ,000 features that are variable.

then it becomes just a factor of how much data you had in training set, how many times you've seen this transaction happen for the language model to be able to do the same job as you have with your Tesla, but just with 10 times more features or a thousand times more features.

Auren Hoffman (07:24.366)

Yep. And I imagine there could be like an agent as a buyer to help you buy. When you're selling, you're selling many, many times. And so you're constantly selling, you're constantly honing. When you're buying a piece of ERP software, you might be buying once every 10 years. And so you don't have that knowledge in your organization. And so if you worked with like a specific agent that buys all the time,

That could be pretty valuable for you. Can you imagine those types of things happening?

Oleg Rogynskyy (07:56.24)

which leads me to cross -customer data sets. Because if I were even as a seller, I'll sell ERP to the same company once with this company's specific priorities, et cetera. So having a cross -customer data set where a number of customers pull together their selling experience without, a number of sellers pull together their selling experience across

millions of transactions versus thousands of transactions they see now in a way that doesn't violate anybody's privacy, security, confidentiality, all that stuff is already happening. And then the buyer side, mean, Koopas of the world or the next generation of Koopa software will be pulling together the procurement data across many buyers. mean, that already exists like Vantas of the world and whatnot, other companies that do SaaS buying for their companies, for their customers.

pulling together buyer data, understanding the full end -end market of how that works. And then giving buyers network -based leverage is already happening manually. It will definitely be happening with AI. So now you will have an AI -based buyer that has seen you sell your thing in like 80 % of your customers. An AI -based seller on your side has seen

Auren Hoffman (09:20.578)

Yep.

Oleg Rogynskyy (09:23.376)

how you sold to all customers and how other similar companies have sold to the same customer and other similar customers. It becomes an arms race of data sets at that

Auren Hoffman (09:38.112)

Interesting.

Auren Hoffman (09:41.863)

In this world, are there other types of agents that help? Like, are there other types of external data sets that come in or are there other types of like, how does it, how does, or is it just like one company to rule them all? Like what happens? Is it more federated? It becomes more centralized? Like, because I can imagine if like one company is like in the middle of all these transactions, like that could be a good thing. Like maybe it becomes very efficient, but also like that company may

objectives that aren't the same objectives as the buyers or the sellers.

Oleg Rogynskyy (10:13.82)

Yeah, I mean, first of all, that company, like we already have a company like that that is doing that on the lower end of the market, it's Stripe. And Stripe is literally putting in a percentage fee on every transaction. Yeah, and Google in some ways too. So there will be a Stripe of enterprise sales where it's going to be Stripe or someone else is unclear, but someone, think it's most likely to be Microsoft, will insert their fee, their cut in every enterprise transaction going back and forth.

Auren Hoffman (10:21.974)

Yeah. And Google in some ways,

Auren Hoffman (10:32.845)

Yeah.

Oleg Rogynskyy (10:42.78)

which leads me to your initial question of how does it work? In reality, the hardest thing to get in that case is not the data sets, it's the human attention, the human eyeballs. And there are four companies that already have all the human eyeballs, like Microsoft with their Office 365 and business applications and Xbox and whatnot, already has probably 30, 40 % of your attention spent daily.

Auren Hoffman (10:57.507)

Yeah.

Oleg Rogynskyy (11:11.088)

So for them to start inserting additional utility that allows you to do bigger and bigger transactions fully automatically into that span is a matter of time. And so the way I look, I see this market unfolding is that you'll have one or several platforms that already have human attention built in. it's Microsoft, it's Google, it's a little bit of Salesforce, things like that. TikTok of the world.

Auren Hoffman (11:12.493)

Yeah.

Oleg Rogynskyy (11:39.74)

And then they will be running some kind of language model based universal prompt or inference engines. And then it's all about how many data sets you plug in into your inference engine to give an edge to your clients versus the other side. And so that's where this podcast was very interesting for me besides you, Aron, is

People who are listening to you are the folks who are operating those data sets. And eventually it's going to be an arms race of getting access to unique data sets that will give the advantage to my seller or your buyer over the other

Auren Hoffman (12:08.151)

You

Auren Hoffman (12:26.318)

And how does that work when, are, because we've already started to see some of these AI companies try to make proprietary deals to get access to data. But for most of these, most companies that sell data, still very small percentage of their sales is to AI companies for companies that maybe don't sell data like traditionally like a Reddit, it might be 100 % of their sales.

Like how do you see that kind of market evolving over

Oleg Rogynskyy (12:56.06)

So that's an interesting one because I think the longer the companies do not sell their data to you, the more leverage they have. For as long as you have a unique data set, nobody else can get.

Auren Hoffman (13:09.632)

Yeah, which most people don't have. Their data set isn't completely unique usually. Usually there's alternatives to their data.

Oleg Rogynskyy (13:16.321)

That's why you see media companies doing all these lawsuits because they actually have unique data that is highly curated, human -proved, proved by MBAs and PhDs. That's very hard to do. And so I think data rights is what will drive massive differentiation for AI models. And those AI models that have better differentiated data sets will

drive much higher leverage and upside for the market. this is kind of like whatever data rights are being acquired today for unique data sets is times infinite when the AI starts using those data sets for at scale high leverage transactions.

Auren Hoffman (14:01.848)

So if you're advising a data company, they could sell to multiple AI companies. They could sell to one as an exclusive. They could get a fixed fee. They could try to negotiate a variable fleet, though I don't know how they would even do that. They could wait and not sell right now and sell in two years. What would be your advice to them?

Oleg Rogynskyy (14:22.246)

First figure out how unique your data set is because the more unique data set, and ideally the data set disappears after you consume it. That nobody else can have the same data set is very important.

Auren Hoffman (14:27.448)

Yeah.

Auren Hoffman (14:35.02)

Yep. And some of them are temporal too, right? So the temporal ones are more valuable than the non -temporal ones.

Oleg Rogynskyy (14:41.452)

Exactly. So first figure out how defensible you are. And then once you figure out that you have a very defensible temporal data set that contains the signal to something that drives economics, like in our case, sales activity data drives deals. Then the next question is, what is your strategy to maximize the economic value of the data set? And so if it's temporal, you can be selling it or leasing

to AI companies for a specific temporum, so specific period of time, or for specific, let's say, last years, this year, whatever it is. I mean, there is many ways to do it, but the uniqueness of data set and inaccessibility of the data set through public data or other sources is what dictates your leverage.

Auren Hoffman (15:33.23)

Now let's say you're not a data company, but you're a company that has interesting data as an exhaust. So let's say you're Hertz or Avis rental car, and you've got this really cool data about like all this different stuff about when people rent cars and where people go with those cars and how people use them and probably all this other like really interesting data that probably could be valuable. How would you help those companies develop a data strategy?

Oleg Rogynskyy (16:00.412)

That's a good question. So we knew this from the very beginning at People .ai. And while we have massive data sets, probably some of the biggest data sets about how sales are done on the planet and who they're done with, we never sold the data and will never sell the data. We knew all along that this data is going to be extremely valuable for fine tuning the models for the benefit of all customers. Same story

Hertz and Amos, et cetera. I don't think they first have thought through yet all of the data collection they could be doing. Like fleet telemetry data.

Auren Hoffman (16:40.482)

Yeah, they're probably only collecting like a very small percentage of the data that they could be. Yeah, because they could get even like the breaking and they could have cameras to see what's going on. And there's probably a lot of other really great data they could be doing. Yeah. Holy. Yeah.

Oleg Rogynskyy (16:43.472)

Yeah, the biggest asset.

Oleg Rogynskyy (16:50.384)

They could see every pothole. They could see every pothole. could see economics. So why do people rent cars? Because they go somewhere with a purpose. They could see there's a lot of things they could do. In fact, what makes Tesla is not the car. It's the data set they collect every time you take over your steering wheel when you want to take over and the auto drive is messing up. And so the other self -driving cars are begging for that kind of data right now and large fleets and

Auren Hoffman (17:00.589)

Yeah.

Auren Hoffman (17:08.078)

That's right.

Oleg Rogynskyy (17:20.452)

understanding of the physical world through cars eyes or sensors. I don't understand why Hertz or Avis have not outfitted every car they have with a ton of sensors and haven't become data companies

Auren Hoffman (17:35.212)

It's a good point. Yeah, you're right. Because for a very small amount of money, they could and they could even store that data locally. And then when you return it, usually return it within a week or so, then they could upload it to the Wi Fi. So you don't even have to worry about the bandwidth or anything. They probably don't need that data in real time. And that could be really super valuable.

Oleg Rogynskyy (17:50.342)

Mm -hmm.

If airplanes do that, we work with one of the large airplane manufacturers and they basically are, they used to the fact that there is millions of sensors on every airframe and every time they land in the airport, there is a big cable if you know they plug in. That's a really big USB cable that downloads like couple terabytes of data in a minute in between flights. So that then it goes into a massive cloud compute.

that processes all the data, teaches airplanes how to fly better every freaking time, but also does preventive maintenance. they know that something is going to break six flights in advance. So they know that they make sure the part is waiting at the right airport at the right time.

Auren Hoffman (18:29.313)

super cool.

Auren Hoffman (18:41.262)

Yeah, yeah, yeah, that's super cool. in your big company, let's say your big airline company or some other type of company that's out there, you always have this kind of discussion internally, like should we be buying AI capabilities off the shelf and maybe tweaking them for our business? Should we be building our own niche models? Maybe we don't even have the talent internally to go do that. Like how should a large Fortune 500 company be approaching this?

Oleg Rogynskyy (19:10.364)

So that's a very fun topic to talk about. let's assume, so where are We're getting with people AI and I'll use sales as an example, but this is applicable in every other domain. Let's assume what we started the call with is that we have AI based seller. Let's say today. What that means is that I can put out on the street infinite amount of sales capacity. And what I mean by that, it's not just a chat bot.

Like you saw the deep fake technologies and hey, whatever that company is that does, does avatar video avatars. There's a couple other ones we are working with one partner right there. This could be not me on this video right now. This could be my avatar that actually has studied everything I've ever said, both my personal email, but also my work stuff and all my blogs and everything.

Auren Hoffman (19:58.862)

And it makes you 30 % funnier and more charismatic and, et cetera.

Oleg Rogynskyy (20:01.436)

Exactly, exactly. could just my problem could have been like make or and laugh and make a joke about this and done and off you go and by the way, I could be on 50 or 500 or 5 ,000 of these podcasts right now. So when we can put out infinite sales capacity on the street, the economics of this planet completely change. It's kind of scary because imagine I'm a Hertz or Avis.

Auren Hoffman (20:08.039)

You're doing a good job. Yeah.

Auren Hoffman (20:25.514)

Absolutely. Yeah.

Oleg Rogynskyy (20:30.842)

What if I could call on every customer that I have identified of mine, but also of my competitors, because it's not hard to buy data on every person who's ever rented any car or who has a license or, I mean, you get the point. And if I could use that data, travel patterns, whatnot, to identify who is most likely to travel and rent a car in the next year,

Auren Hoffman (20:46.626)

Yep, yep, yep, yep.

Oleg Rogynskyy (20:58.138)

I know exactly where to go right now to get five of those APIs plugged in together and get an understanding. The bottleneck is not identifying who could buy my stuff. The bottleneck is having humans in a call center today, dialing those people and reading the script of the screen to get them to rent from Hertz, not Avis. Now, if I can have a perfectly educated, perfectly enabled

AI based seller that talks, walks and looks like me. But has human intelligence of 10 ,000 of me and people like me or 10 million of me and people like me, having had those conversations, call on every Avis customer and hurts before noon tomorrow. And by call on, mean not just phone, but retargeting ads, outreach, email, text message, you name it.

and also have some kind of temporal element to it of these people are more likely to buy than those people. I could saturate the town and win the market against another public company literally overnight. The only limit to that is how many GPUs Amazon is

And so if the other company has not been on the same strategy doing similar offensive but also defensive measures. Like someone calling on public companies, 30 % customers with a 10 % conversion ratio in 24 hours and then doing it again for 24 hours and doing it again for 24 hours nonstop. So that in 15 days you lost 30 % of your market share and there's nothing you can do about

completely changes how companies exist.

Auren Hoffman (22:51.308)

It's really interesting because like historically there's been kind of two types of successful salespeople. There's been that relationship sales person who builds a deep personal relationship, takes you to the ball game.

has that type of, and then there's been like more of the product oriented sales person relationship oriented sales person maybe isn't as good as follow up. Maybe don't know the product as well, et cetera. And then there's more of the product oriented who's like got a whole process. And, and I would say in the last like 25 years, we've moved from where like the relationship oriented sales person was the dominant kind

breadwinner in a company to the product oriented salesperson being the dominant breadwinner. if AI takes over, it seems like the AI will be better at doing the product oriented stuff faster than the relationship stuff. And we might see a resurgence of that relationship oriented salesperson come

Oleg Rogynskyy (23:50.34)

I wouldn't say it's a resurgent, there'll be less of them, but they'll be extremely more valuable. I literally, when we shipped our first fully end -to -end AI autonomous agent last December, I was at a Christmas party in Toronto with one of my top salespeople. And she just went through enablement on how to sell these AI capabilities that scale infinitely, basically. She comes up to me and she's like, Oleg, is this the end of my career?

Auren Hoffman (23:55.116)

Yeah. Yeah, okay, that's fair.

Oleg Rogynskyy (24:20.604)

I'm like, no, not yet. In fact, let me give you an analogy. We've had autopilot on airplanes for 30, 40 years. And today on small Cessnas and Ciruses, autopilot, problem, get on and they'll be the first planes to completely take off and land fully automatically. I small drones are that. Yet when you're flying a big super Jamba plane, you'll still have two and eventually maybe only one pilot

Auren Hoffman (24:30.189)

Yeah.

Oleg Rogynskyy (24:50.65)

because you want human being there as a copilot to a full autopilot in case the autopilot messes up and you have so much responsibility on the

Auren Hoffman (24:57.654)

Yeah, yeah, something goes wrong. Like every, every five flights, something goes wrong and you have to do something or whatever. Yep. Yep.

Oleg Rogynskyy (25:02.17)

Yeah, yeah, you need to have a human there. But let me give you an analogy. I was talking to an airplane manufacturing customer, their chief customer commercial officer, and I'm like, what's the biggest leverage for a business? And they're like, the biggest leverage for a business for our business is using AI to make sure there is only one pilot in the cockpit, not

Auren Hoffman (25:24.206)

instead of two. Yep.

Oleg Rogynskyy (25:24.42)

Why? Because he's like, we can make all the airplanes in the world. There's not enough pilots. By just putting AI in a cockpit to replay in the core pilot and they were human, they'd be for safety compliance, et cetera. We literally double our time and we drop the travel prices on the planet in half because now there is twice as many flights. So to your point, going back to the sales person.

Auren Hoffman (25:47.362)

Yep. Yep.

Auren Hoffman (25:51.488)

A lot of times why a reason a flight gets canceled is, you know, one of the pilots gets sick or you have all these other things that could happen there.

Oleg Rogynskyy (25:55.088)

Yeah. Yeah. And AI doesn't sleep, doesn't go on vacation, doesn't join unions, all that stuff. mean, for unions, we will see if that happens. But to the original question you asked, I think the strategic most experienced sellers will be the only type of sellers remaining. And there is downsides to that as well, because one, AI is very good at replicating

Auren Hoffman (26:02.338)

Yep.

Oleg Rogynskyy (26:23.292)

the playbook that's already established when you collect the telemetry. So there'll be in a big company has now five, 10 ,000 salespeople. There'll be 50 salespeople who are the most high end sellers on the planet. They will think they're selling their job is to sell. In reality, their job is to teach AI how to sell. And so humans will still be figuring out the playbook. AI will be replicating and scaling it

Auren Hoffman (26:40.333)

Yeah.

Auren Hoffman (26:48.716)

Yeah, yeah. If you're a tiny company, which still will search for product market fit, you still need to have that like then it's the AIs can be a little bit. It's a little bit harder to use AI to replace that.

Oleg Rogynskyy (26:57.54)

Yeah. And so the interesting part is that while humans will be coaches to AI, we will look, we'll see a much bigger problem over time, which is who's going to coach humans. Just like in engineering right now with GitHub Copilot, there's no more need for junior engineers. Like the AI acts as a junior engineer right now and just helps really experience people do more, faster. And that's it. Same story will happen in sales where

Auren Hoffman (27:25.326)

That's right.

Oleg Rogynskyy (27:26.844)

AI is the perfect sidekick. don't need inside sales, junior salespeople, etc. If 50 high end sellers can do the

Auren Hoffman (27:34.318)

Yeah. And it seems like the SDR seems very automatable, like most of the stuff in SDR does. In fact, a good SDR today is just like using like, it's just coordinating 10 tools. They're just like a coordinator of the tools, but they're not doing as much of the outreach. And it seems like we'll see way more of those things start to happen over

Oleg Rogynskyy (27:48.124)

Yeah.

Yeah.

Oleg Rogynskyy (27:56.656)

Yeah, I mean, to be honest, like SDR automation, the only reason why SDR automation hasn't happened is the way fully is that the cost structure of automating SDR scales not with number of users, but with number of contexts or accounts. So like the only reason why SDR is not automated is the LLM inference cost. We're talking now about automating SMB mid -market and even lower enterprise sales teams.

Auren Hoffman (28:14.275)

Yeah.

Oleg Rogynskyy (28:23.772)

initially in the copilot mode but eventually you won't be needing a steering wheel in your car either.

Auren Hoffman (28:33.506)

Yeah. And while that like amazing sell on your team probably won't get replaced anytime soon for the sellers that are worried about it, like how should they be thinking about their own careers?

Oleg Rogynskyy (28:47.036)

lean into AI and learn how to become the most AI enabled high end seller you can be. That's the best job security.

Auren Hoffman (28:55.458)

Just like being a software developer, leaning in, leaning into Copilot or leaning into whatever the tools that exist.

Oleg Rogynskyy (28:59.484)

Yeah. In fact, I believe that 2022 was the year with the most salespeople on the planet. So all the layoffs, et cetera, obviously, a lot of it was performance based or headcount based, whatever it is. But the same number of headcount is not getting rehired. By the time companies are back to full steam hiring, you'll be hiring AI agents, not

Auren Hoffman (29:26.22)

Yeah, interesting.

Auren Hoffman (29:33.58)

random aside, but you have this amazing domain name people, people .ai. How did you get that? Is there a story behind it or?

Oleg Rogynskyy (29:40.732)

We also have sales .ai. Yeah, we have BDR .ai and a whole bunch of other ones. The story there

Auren Hoffman (29:46.532)

I didn't know that. Amazing. That's another amazing one. wow. Okay. wow. Okay. So you're just like, you're just like on the secondary market collecting these like great domain names or

Oleg Rogynskyy (29:56.764)

No, actually simpler. I worked for this company called H2O .ai before starting People .ai. And one morning when I decided to start a company, I was sitting down in a diner in Menlo Park with a couple of friends. And we were like, well, let's take a look. And there was only one domain name, a registrar doing .ai domain names in February 3rd, 2016. I remember that. was called onlydomains .com. And we just went.

Auren Hoffman (30:23.575)

Okay.

Oleg Rogynskyy (30:25.724)

heavy at it and like I ended up buying like 60 or 70 primary like first word dot AI domain names.

Auren Hoffman (30:33.102)

my gosh. just, they just, so, so in 2016, the dot, the dot AI domain name wasn't as well. Cause I feel like I thought I had seen it, but maybe you're right. Maybe it was like 2018 before I started seeing a lot of the dot AI domain names.

Oleg Rogynskyy (30:35.004)

Everything was available in February 2016.

Oleg Rogynskyy (30:46.876)

Or by that summer, it was all gone. Like I was literally like probably in the last moment when it was available. We bought like customer success .ai, like inside sales .ai, account executive .ai. We went heavy on that and spent probably like three, $4 ,000 on a portfolio that's worth much

Auren Hoffman (31:04.972)

Got it. So and I believe dot AI is this like country of Anguilla is the country that that has it. And I think it's

British protectorate or something. So I guess there's probably some kind of rule of law. They're not going to just all of a sudden just say like, hey, tomorrow you need to, you know, I can imagine like a more extracting type of thing. Like I believe like, I think it was one of the countries was like one of the domains that was popular like 15 years ago was like owned by Libya or something. And so you could see how they could like be a little bit more extracting on your domain name.

Oleg Rogynskyy (31:36.998)

Yeah.

I mean, I would be very curious to find out what is the percentage of Angular's GDP that is coming from registering .ai domain names now, because I'm sure they're getting quite significant inflow there.

Auren Hoffman (31:53.964)

Yeah, yeah, exactly. Exactly. It's really interesting. That's great. That's amazing that you're able to do that. I mean, I think those those those brands really are important. They might not be important to selling to AIs in the future, but they're very important to selling to people like people do when they see something like horses, people dot AI or sales dot AI. That means a

somebody rather than you can imagine like people underscore sales dot AI is just not as compelling of a name. Now you personally describe yourself as a born seller and what would be your advice for CEOs that never really saw themselves in a selling role? Is it just like not going to be as important in the future or?

Oleg Rogynskyy (32:20.794)

Yeah. Yeah. Yeah, for sure.

Oleg Rogynskyy (32:37.91)

you

Well, I mean, right now you gotta know how to sell because it's fundraising is selling, recruiting is selling, building your

Auren Hoffman (32:52.728)

Can you imagine the fundraising side and recruiting side? Like you could imagine having like an agent and then it would be a little weird. Like then in the middle of it, someone's going to be like, okay, am I really talking to the CEO or am I talking to the CEO's agent? Right. And then the, okay, I'm not, you're not talking to this, the real CEO right now, but like, if you pass this round, then the next round, you'll actually talk to him live or talk to her

Oleg Rogynskyy (33:14.236)

Well, I mean, have a couple things about that. One is if SBF was able to raise money while playing computer games, like, and he raised billions of dollars that way, like, I don't think agents will be that much worse than SBF spending 10 % of his attention span on fundraising in real time. But second piece there is that I actually think VC job is safe for now.

Auren Hoffman (33:23.074)

Yes.

Auren Hoffman (33:29.388)

Yeah, that's right. That's a point.

Oleg Rogynskyy (33:42.972)

Just because the number of reps or transactions that are happening is so low versus mid -market enterprise selling. Exactly. collecting enough telemetry data on venture transactions to build a meaningful agent experience of that is actually really hard. only even the biggest firms, your Andres and your Sequoia probably don't have even 1 % of transaction volume that's needed.

Auren Hoffman (33:49.526)

Yeah, there's a lot of gut feel and

Oleg Rogynskyy (34:11.494)

to make a VC agent work really well. And same story on the founder side.

Auren Hoffman (34:15.246)

Well, but let me push back on that a little bit. Like a lot of times you're you get an inbound and it's it's hard to know if it's worth your call if you're a VC because it's 30 minutes. You got to schedule it and you know, so so oftentimes you don't even take the call, but you're it could be 1 % of the time of things you don't take are just like an amazing company, but you just didn't you didn't you didn't go around to actually taking the call.

Oleg Rogynskyy (34:25.36)

Mm -hmm.

Auren Hoffman (34:41.674)

Maybe you hire some super junior people to take as many calls as possible. That's kind of a similar thing where they could have like, you, probably have kind of a core list of things of information you want to get from them. You probably have a few kind of questions that are, and yeah, maybe it's not as good as you personally, but it like, could get there and you could even be honest with the company. Hey, you didn't, you didn't make it to the bar of me personally getting on the phone with you, but you made it to the bar of my agent getting to the phone with you.

Oleg Rogynskyy (34:52.604)

Mm -hmm.

Auren Hoffman (35:11.542)

And they'll have a kind of a quick conversation with you don't want to waste your time. then, and then I can get on if it gets to the next

Oleg Rogynskyy (35:15.26)

Yeah. venture capital associates. I wouldn't want to be in that job anytime soon just because I think agent on agent initial triage, definitely, 100%.

Auren Hoffman (35:29.644)

Yeah, the CEO could have an agent and the venture capital firm could have an agent and you could just be figuring out right away if there's like a fit to then get a little bit more deeper.

Oleg Rogynskyy (35:37.828)

Yeah, yeah, so that's that will be I would be surprised if that's not happening,

Auren Hoffman (35:44.11)

Yeah. Okay. Yeah. Recruiting what I imagine is a similar thing where like one of the biggest bottlenecks to me to hiring is just doing the calls with all these people. it takes a lot of time to, to recruit people, to talk to them, to have meaningful conversations, even if you optimize it to a 20 minute call or something. you know, it's, and you just need to schedule and you say, you need to

break up your day, et cetera. So it's very hard to do and engineers hate it. Engineers hate recruit, having to recruit and go talk to all these people. So you can imagine at least the early things. And sometimes you already do that. Like you may put a test at the front of recruiting or you do some, you can imagine some of the more early stuff being done with AI.

Oleg Rogynskyy (36:30.156)

So, you know, I was with a CEO of a top three manufacturing company, industrial manufacturing company on the planet this Tuesday in DC. And she told me something really interesting. She's like, all this AI white collar automation thing. She's like, we've been doing it for decades. It's called digital twins. We've been building digital twins of wind turbines and generators and trains and planes and whatnot. And we would

Auren Hoffman (36:54.796)

Yeah, that's right.

Oleg Rogynskyy (36:58.096)

test how things behave with an environment, simulated environment, or how the two digital twins operate with each other. So what we're talking here about is creating digital twins of humans in certain roles. From that perspective, it's not that different. We just needed a better interface, which is language models.

Auren Hoffman (37:12.631)

Yes.

Auren Hoffman (37:18.198)

Yeah, that's really interesting. Now, a few questions about Ukraine. You're really involved in Ukraine. First of all, I know that you guys had a lot of employees in Ukraine and you were very kind of proactive of moving out of the country even like months before the invasion. How did you have a sense that that invasion was coming? Was it just obvious because there's all these like tanks at the border or how do you have like a good sense that that was coming?

Oleg Rogynskyy (37:43.548)

Well, there was a number of things there. There was thanks on the border, but also just a simple pattern matching when Russian leader, dictator, or whatever you want to call him now, started saying things and drawing red lines that that obviously were leading to war. Like I just assumed that people of that caliber do not change.

their mind or don't step back from their word easily.

Auren Hoffman (38:16.13)

Yeah. It's interesting. There's kind of like different types of people out there. And I'm actually like you, like when a leader says something, I just believe them. Like whatever they say, I believe if they say they're going to do something, I believe that's what they do. And then there's like people who are like much smarter than me. And when a leader says something, well, he's not really going to do that. You know, and so there's like, there's different ways of like analyzing people.

Oleg Rogynskyy (38:19.556)

Yeah. Yeah.

Oleg Rogynskyy (38:24.368)

Yeah, I

Oleg Rogynskyy (38:38.392)

whether they're good or bad, to become a leader of a large entity like a country, you got to do what you said you're to do consistently. Therefore, I remember it was November before the Ukraine war started where I was like, whoa, things have been said that cannot be walked back. Yes, things have been said that cannot be walked back. In fact, I remember I went and took my parents to Cabo for vacation.

Auren Hoffman (38:48.972)

Yep, yes, yep, that's right.

Auren Hoffman (38:56.27)

This is November, 2021. Yeah.

Oleg Rogynskyy (39:05.916)

in November 2021 over Thanksgiving. And I was actually spending part of my vacation scouting for where we would bring our team from Ukraine to Mexico. Similar time zone. Doesn't need a visa.

Auren Hoffman (39:17.646)

I because they could get Mexico. Mexico is easy visa from Ukraine, whereas the US would be very difficult.

Oleg Rogynskyy (39:21.308)

Yeah, same time zone and easy for Ukrainians, no visa needed. You can live there, lots of hotels available, etc. So I had a conviction that it's going to happen back in November. And then I have a lot of friends in Ukrainian government. I'm from there. I'm from the same town as Zelensky and a whole bunch of people. And so they were not kidding. Like things were grim and people are preparing. And so that's when I came back early December 2021.

Auren Hoffman (39:30.316)

Yep. Yep.

Oleg Rogynskyy (39:50.776)

and told my team, like, hey, we need contingency plans. So we built out all the way to printed Google Maps for our team. forced everybody to get their passports. We forced everybody to vaccinate their dogs and cats and get all the COVID paperwork, because back then it was still COVID, so that we could get people mobile quickly. And then in January, we literally

Auren Hoffman (40:09.655)

Yeah.

Oleg Rogynskyy (40:15.44)

paid people to go on vacation in Europe kind of thing. recall calling one of my first engineers who was on his honeymoon in somewhere and say, hey, what if we double the length of your honeymoon? Just stay there. Don't come back. And we literally did that. And so that's how a lot of the team ended up being outside of the

Auren Hoffman (40:37.922)

Now, one of interesting things that you and I had talked in the past about is when Zelensky first ran for the election for president, one of his big supporters were these software developers. Why were software developers such a powerful block in Ukraine?

Oleg Rogynskyy (40:51.26)

Mm -hmm.

Oleg Rogynskyy (40:56.764)

So that's a very interesting story. So 2014, when Russia annexed Crimea and attacked Eastern Ukraine, Obama administration did not come to help. They kind of like, Russia, you're bad, some lightweight sanctions, but they didn't even give Ukraine money to do anything. So Ukrainian government was forced to print money. And so when they printed money,

I think the Ukrainian currency devalued from like 8 to 1 to like 25 to 1 in like one year to one dollar.

Auren Hoffman (41:33.996)

So it lost lost two thirds of its value.

Oleg Rogynskyy (41:35.992)

lost two thirds of its value. There was one group of people in Ukraine, unlike Russia, that was making money in US dollars directly. And these were all the software developers because the Ukrainian market is too small.

Auren Hoffman (41:46.618)

Software developers. Yeah, you're paying them you're paying them in dollars because the currency risk was probably to you know, be a whatever PayPal back then or something you're sending them whatever the monthly fee is Yeah

Oleg Rogynskyy (41:50.81)

Yep.

Oleg Rogynskyy (41:55.196)

Exactly. Payoneer, all those things. And so the difference in Russia and Ukraine was really big because in Russian market, it's big enough. You have Yandex, have your own Uber, your own eBay, your own whatever, your own Google. They would all work for local companies for internal market. They would be paid in rubles. Exactly. In Ukraine, market is too small. Everybody was working for foreign companies paid in dollars. So when currency

Auren Hoffman (42:15.692)

Yeah, and get paid in rubles. Yep.

Oleg Rogynskyy (42:24.568)

Ukrainian engineers ended up actually increasing their relative income versus the rest of the Ukrainian population because

Auren Hoffman (42:33.486)

Because they can now can they can afford apartment three times bigger, they could go to the best restaurants, they can go to the all these different things. Because relative like on a PPP basis, they they went from like middle class to wealthy or something.

Oleg Rogynskyy (42:37.528)

nicer car

Oleg Rogynskyy (42:45.103)

Yep.

And then combine that with 2016, 17, 18 is when Silicon Valley started hiring people anywhere because there was not enough engineers. So demand globally went up and there even dollar salaries start growing really quickly. suddenly

Auren Hoffman (43:05.614)

Yeah, I remember like people in Ukraine were like, you know, fourth of a US engineer and then all of a sudden they become half the US engineer and all of a sudden they become 60 70 % of US engineer like their salaries start rising pretty quickly.

Oleg Rogynskyy (43:15.345)

Mm -hmm.

Exactly. Three times the PVP income versus Ukraine, suddenly the engineering jobs started making you into like top 1 % earner in the country. And then Ukrainian government, of course Zelensky did was really good. They extended the regime where engineers could have a subsidized 5 % income tax in Ukraine. It's like indirect. There's a whole thing there, but the government supports it and engineers pay 5 % income tax

Auren Hoffman (43:30.978)

Yep. Yep.

Oleg Rogynskyy (43:47.748)

And so, crazy.

Auren Hoffman (43:50.196)

Why is that? it just because otherwise they would just avoid taxes so they weren't getting any? Why is the tax rate so low?

Oleg Rogynskyy (43:55.325)

That's one thing, but I actually think in the long run, that was a very smart governmental policy.

Auren Hoffman (44:01.774)

It's just encouraging more people to become software developers.

Oleg Rogynskyy (44:03.42)

Suddenly, if you look at enrollment in Ukrainian engineering schools, it went from like 10 ,000 to like 150 ,000 people a year in 2016. Suddenly, all the pop bands started seeing like, it's cool to be a developer, I wanna marry a developer, like all that stuff you would go

Auren Hoffman (44:13.434)

my gosh.

Auren Hoffman (44:23.638)

Yeah, correct. Those are the people going to their concerts. So the only ones who could afford to go to the concerts.

Oleg Rogynskyy (44:27.132)

Exactly. You go to a bar, you go to a club. People who are spending money were engineers. People who were traveling to Europe were engineers. In fact, here's a story. We tried to relocate during that time a few engineers back to Canada or US because of time zones. And people were like, why would I go to where on PPP Parody I'm a middle class and I have to have 50 % tax.

Auren Hoffman (44:49.134)

Right, even if you pay me, yeah, even if you pay me 30, 50 % more, I'm losing massively, yeah.

Oleg Rogynskyy (44:55.484)

Yeah, so people just would not do that. And then the next thing that happens is this happens like this increase in everybody wants to be an engineer, not a doctor or a lawyer happens in 2015 -16. Four years later, 2020, the first big cohorts of engineers start being released, graduating. 150 ,000 in 2020, COVID is taking

Auren Hoffman (45:18.189)

okay. And that's right in time for COVID. Yeah.

Oleg Rogynskyy (45:22.216)

Now it suddenly doesn't matter where engineers are, salaries become on par. Companies like Deal come out and make hiring engineer in Ukraine one click. Suddenly that is taking off even further. They don't want to travel, they make a lot of money, they like being in Ukraine.

Auren Hoffman (45:32.524)

Yeah.

Auren Hoffman (45:37.526)

In they weren't even allowed to travel in those years.

Oleg Rogynskyy (45:39.45)

Yeah, but then in COVID Kiev had really liberal COVID policies. So all of Europe. So.

Auren Hoffman (45:45.942)

All right, so people started moving there, right? Like, I know people who like move there from London even to move to Kiev.

Oleg Rogynskyy (45:50.396)

Yes, so Ukraine became like the Florida of Europe at the time during COVID. So suddenly we're here, we're engineers, we're making a ton of money and Europeans are now coming in to hang out with us and open Michelin star restaurants over here. It's a cool spot. Yeah. And so that happened under Zelensky. And so why would you not support the guy who is giving you 5 % tax rate and is making your country a cool place with policies and

Auren Hoffman (46:04.832)

Yeah, it's like the cool place. Yeah.

Oleg Rogynskyy (46:19.738)

young and hip and all that stuff. And then the war happens and these people, unfortunately, or fortunately, were not as strong of believers in Russia attacking and all that stuff. So they stayed in the country. Now, that is one of the reasons why Ukraine is doing well, relatively well in the war is

Auren Hoffman (46:40.75)

because they've got a tech talent base as well.

Oleg Rogynskyy (46:41.296)

tech talent stayed and that tech talent went to the front lines, got the drones working, got the innovation going, software, all that Starlink, all that stuff that we all know about

Auren Hoffman (46:53.106)

One interesting thing about Ukraine that I didn't really know about until I met you was that it has kind of a super app where citizens can do almost anything related to interacting with the government, getting a driver's license, paying your taxes, you know, all this kind of thing. And a few countries have this, it seems like what Ukraine is certainly at the forefront. How did that evolve?

Oleg Rogynskyy (47:14.044)

So that's, as a Ukrainian, that's something I'm really proud of. The current Minister of Digital Transformation, Michael Fodorov, had a thesis, which is like, corruption happens when SQL databases are not joined. Which, if you think about it, it's true. Like, when you have your driver license and driver test databases not joined, and when you insert a record for someone not doing a driver test,

Auren Hoffman (47:31.682)

Yes, in the shadows essentially, right? Yeah.

Oleg Rogynskyy (47:43.58)

And then that's not going to kick off cascade effects in 15 other databases. You can do it in a shadow and take money for inserting rows in a database. But if everything is joined and suddenly you do this and your medical record updates and your voting record updates and everything, suddenly you cannot do corruption anymore. So the thesis Ukrainian government had is let's join everything we can. And they started with a mobile app. And first they start

Auren Hoffman (47:55.052)

Yes. Yeah.

Oleg Rogynskyy (48:11.002)

believe driver licenses and then driver licenses were joined to tax ID and then it was joined to your real estate transactions and then they have a processing system there and then suddenly you could pay taxes online and then they're like well let's just do medical records too and then they said well you can only do a COVID test and leave to leave the country during COVID but you can only do a COVID test through the government medical record system therefore go to your doctor and update your data.

Auren Hoffman (48:38.284)

Okay, so I encourage more people to go do it.

Oleg Rogynskyy (48:39.664)

And we're going to connect a passport data. So it's one click update. And now they're doing military service registry this way. So the whole country has joined every freaking database in a single data model. And it's kind of insane. And so the app itself, I think it's cool. It's easy, great UX. They copied all the Apple ask stuff there. Super easy to use. But the app is the tip of the iceberg.

Auren Hoffman (48:56.802)

Yep.

Oleg Rogynskyy (49:08.668)

for the country to have joined every possible database into a data fabric.

Auren Hoffman (49:18.542)

And so essentially they have all these microservices and APIs that go between all the different databases. Wow.

Oleg Rogynskyy (49:23.868)

100 % like US is super federated in comparison with Ukraine like Social Security and immigration don't talk to each other like you still have to do fax in between or something Yeah in Ukraine Social Security medical systems I don't know conscription systems military procurement and I don't know Streetlights and traffic fines are all connected

Auren Hoffman (49:29.774)

Correct. Yeah.

Correct. You have no idea. Yeah.

Auren Hoffman (49:50.99)

It's amazing because like

I don't know if this would be good to do in the US, but I'm confident that the US would not be able to pull it off. You have to get these super amazing talented people in there. And in the US, they'd probably have some Booz Allen created. And it would cost $30 billion. And then they would get out something super buggy that no one would want to use or something. I assume it wasn't like that. I assume it was a group of like a ragtag group of like,

five super sharp engineers who are working throughout the night to make something like that. how did that evolve?

Oleg Rogynskyy (50:26.544)

Yeah. So that started with, so the current ministry used to be a startup founder. He came in and they had to build a ministry from scratch. That was Zelensky's mandate. And so what they did that was really cool, they basically didn't build the ministry based on the textbook of how to build the ministry. Actually, there was no textbook of how to build a new ministry. Like, well, let's do it like a startup. So they basically built a ministry based

Auren Hoffman (50:34.728)

Okay, got

Oleg Rogynskyy (50:55.128)

Scrum teams with product managers, with project managers, with they run on Jira and Asana and the whole thing. And so it was like a hundred person ministry that was operating massive scale. like I remember an example of like they were doing budget acquisitions for a census, like a census is a multi -billion dollar freaking thing for any country. Right. And these guys were like, well, we have three major cell phone providers in the country. Everybody has a cell phone.

Some people have few, so we just need to compare GPS location data of two SIM cards to see if two SIM cards always travel together, it must be the same person. And so if we just resolve that duplication, we now can run a census literally through SIM IDs. And they got census done for like a fraction of a cost in two weeks.

Auren Hoffman (51:36.13)

Yeah.

Auren Hoffman (51:45.58)

Yeah, yeah, yeah, of course. Yeah, yeah. Yeah. Then you just have to figure out like all you have to do is figure out then the kids and the elderly from there and then you're done. Yeah. Yeah. Yeah.

Oleg Rogynskyy (51:53.776)

Yeah, there we go. But then in Ukraine, actually drove a program to, they were even at some point giving out free Android phones to elderly because when they got to this position and free classes, because when they got to this digital situation, they realized that government services through mobile app are like a thousand times cheaper. And so they could find massive leverage in headcount that doesn't need to be in the government.

And the only people who were using the head count were elderly. So therefore, let's educate elderly to use Android app. And you don't have to have as many people sitting in counter and accepting people.

Auren Hoffman (52:28.269)

Yep.

Auren Hoffman (52:33.164)

Yeah, that makes a lot of sense. Now, you've told me some crazy stories about helping Ukraine over the last two years. I'm sure most of them are way too secret for this podcast. But what are some anecdotes you can say publicly?

Oleg Rogynskyy (52:47.065)

That's a good one. So let me think which one is less classified.

Oleg Rogynskyy (52:57.724)

We'll talk about, can we cut this one out? Yeah,

Auren Hoffman (53:05.102)

Sure. Yeah. Okay. All right. We'll cut it out. right. Back to people AI. At least 90 % of the founders who go on this podcast have a co -founder. And I think you're the sole founder of people AI. I mean, also went through like YC in the early days like that.

Well, how did you think about it and how did that kind of come about? And you wish you had like a co -founder you could commiserate with. Would you suggest other people do something like that?

Oleg Rogynskyy (53:40.016)

To be honest, I wish I had a co -founder and I think there is, obviously there is a loneliness aspect here and no matter how good and how close you are with the executive team, it's not the same. How I get by is I have a group of friends, founders and we commiserate. part of YPO. It's been an incredible experience. Highly recommend for everybody who is a solo founder to go through YPO. Now I do feel

in hindsight that being a solar founder also reduces your chances for success for all obvious reasons. But also you have to hire executives earlier because you're just by yourself. And so you basically dilute yourself much more, raising more money earlier, hiring executives earlier, churning executives earlier because they came in too early and all that stuff because you're a solar founder.

Auren Hoffman (54:24.439)

Yep.

Auren Hoffman (54:37.186)

Yep. Weren't the right fit.

Oleg Rogynskyy (54:39.068)

And so in hindsight, I wish I'd spent a bit more time kind of really honing in on building an initial Fauni team and anything in the future I'm going to do, I'm not going to be

Auren Hoffman (54:52.598)

OK, interesting. I love it. Now, back to Ukraine, I think most Americans don't have as much awareness of Ukraine kind of pre -Russian evasion. They may have some idea that, obviously, at some point I was part of the Soviet Union and stuff. As someone who grew up there, what do you think most people in the US understood or knew about Ukraine?

Oleg Rogynskyy (55:01.02)

Mm -hmm.

Oleg Rogynskyy (55:14.94)

I'll go straight to the hot political button right now. There's multiple camps. Some of them are saying we've got to fight with Ukraine all the way to the end and supply them. Not going to get in the fight, but give them weapons. Another side says, hey, it's a local conflict. If we create a truce between Russia and Ukraine, we can kick the can down the road and stuff like that. Most of the West is looking at this as a two -year or maximum 10 -year

When you are in Ukraine, you look at this as a thousand year conflict. The Ukraine was started in Kiev, Rus in 1912, nine, twelve. So over a thousand years ago, Moscow was started three, four hundred years later. There is a popular meme of like the picture of Kiev, which is like in nine hundred, which is like.

looks like any medieval European city built out, like looks like Venice or Venice level of sophistication, churches, big buildings, stuff like that. And then there's the other side of the meme is like, meanwhile in Moscow, it's just like forest. But what I'm going with that is Russia had been coming to Ukraine and killing a bunch of Ukrainians for a thousand years, probably every 50 years. Every time Ukraine didn't fight

Auren Hoffman (56:25.686)

Yeah, yeah,

Oleg Rogynskyy (56:40.508)

bunch of Ukrainians died. Every time Ukrainians fought back, Russians would, and it was only two or three times when they did it successfully, Russians would retreat, but then wait for the opportune moment and come back again. And even more Ukrainians would die. So while in the U .S., the outlook on this is the 10 -year thing, two -year thing, make peace, and off we go to the next election. In Ukraine, it's like, this is the one time when we are actually successfully holding them back.

The first time since like 1648, I believe, when Ukrainian Cossacks were holding Russians off. And if we give up now, they'll come back at much more opportune US president, much more opportune European situation when the right -wing people empower or whatnot. And they will not forget. And so either we deal with it now or our kids deal with it at much worse odds.

Auren Hoffman (57:19.725)

Yeah.

Oleg Rogynskyy (57:39.024)

That's how Ukraine looks at

Auren Hoffman (57:41.006)

And why does the odds get worse over time? Like you think like certain demographics are in Ukraine's favor versus Russia. There's other types of things. So why is it worse for Ukraine 20 years from

Oleg Rogynskyy (57:54.22)

I think Russia's biggest mistake in this war was to start the war during the last US president who remembers Cold War. In five, seven years, the president will not have, had this war not happened in five years, the next US president or the one after that would not have had that DNA level fear of nuclear war, of Cold War, of all the things that Biden remembers.

Like even Obama did not have it on this level. so, so, I think next when Russia comes back, if the job is not finished now, they will come back during US president. They'll be much more thoughtful about who is in charge in the US and Europe to pick their moment when to come back. And Ukraine is going to be all alone again, like it was in 2014 or.

Auren Hoffman (58:50.062)

But any scenario, even if even Russia's quote unquote defeated now, I assume they're going to try again in 10 years, right? Or no.

Oleg Rogynskyy (58:56.252)

Yeah. Yeah. The question is, what is the defeated situation? Like there is from Russia rolling back to Russia. Like in 1991, Russia shed a third of its population into 20 other countries. Well, that could happen again.

Auren Hoffman (59:11.116)

Yep. Yep. All right. This is great. Two last questions we ask all of our guests. What is the conspiracy theory that you believe?

Oleg Rogynskyy (59:19.567)

that's an interesting one.

The conspiracy theory

Believe in.

Auren Hoffman (59:41.506)

Maybe it's a Russia one, I don't know. mean, I assume you're, you seem like a guy who's got some really good conspiracy theories.

Oleg Rogynskyy (59:51.342)

I'm trying to not share something very sensitive.

Oleg Rogynskyy (01:00:00.346)

My issue is that I have access to some classified information and it's very important not to share it. Yeah. Yeah.

Auren Hoffman (01:00:08.449)

Please don't, please don't. Yeah, we're being recorded.