---
title: "What Investors Shouldn't Forget Right Now | Martin Paasi"
titleOriginal: "Tätä sijoittaja ei saa nyt unohtaa | Martin Paasi | Neuvottelija 388"
episodeNumber: "388"
guest: "Martin Paasi"
datePublished: 2026-06-05
duration: "59:49"
youtube: "https://www.youtube.com/watch?v=yblLUI4rnRo"
captionsVtt: https://www.neuvottelija.com/podcast/episodes/712-tata-sijoittaja-ei-saa-nyt-unohtaa-martin-paasi-neuvottelija/captions.en.vtt
captionsSrt: https://www.neuvottelija.com/podcast/episodes/712-tata-sijoittaja-ei-saa-nyt-unohtaa-martin-paasi-neuvottelija/captions.en.srt
originalLanguage: "fi"
topics: ["investing_markets"]
subtitleMethod: "claude-fi-to-en-cue-preserving-translation"
provenance: "neuvottelija.fi Supabase episode-markdown API (platform captions)"
fiCanonical: "https://www.neuvottelija.fi/fi/episodes/712-tata-sijoittaja-ei-saa-nyt-unohtaa-martin-paasi-neuvottelija"
canonical: https://www.neuvottelija.com/podcast/episodes/712-tata-sijoittaja-ei-saa-nyt-unohtaa-martin-paasi-neuvottelija/
---
# What Investors Shouldn't Forget Right Now | Martin Paasi

## Chapters

- [00:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=0s) The Tampere trip and the investment fair
- [02:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=120s) Finland, Europe's second most active investor nation
- [03:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=210s) The bell curve and investor returns
- [05:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=300s) Fees drag returns below the market
- [06:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=360s) Zero fees guarantee the top quartile
- [08:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=480s) Buffett's case for index funds
- [09:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=570s) Compound interest and saver profiles
- [11:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=660s) A savings plan for a newborn
- [12:40](https://www.youtube.com/watch?v=yblLUI4rnRo&t=760s) A whole pension for five grand
- [14:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=870s) Where's the promised million?
- [16:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=960s) YEL reform and scrapping estimated taxation
- [18:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1080s) Turning Solidium into a zero-fee pension fund
- [19:40](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1180s) Sami and Samantha's AI presentation
- [20:50](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1250s) AI as a mean-reverting model
- [22:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1350s) Alphabet and AI inference
- [23:40](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1420s) Nexans as a value buy
- [25:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1530s) A shoutout to vibe coding
- [27:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1650s) Agent memory and shared skills
- [28:50](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1730s) Hello Humans and language-model bias
- [30:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1830s) The Grok agent as a counterargument
- [32:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1920s) Academia's influence on the models
- [34:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2070s) Qwen, the most right-leaning model
- [37:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2250s) The pain and hassle of ownership
- [39:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2370s) Why own a business in Finland at all
- [42:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2520s) The value of ownership to the economy
- [45:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2700s) The AI megatrend and Finland's advantages
- [47:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2820s) Data centers and power transmission
- [48:40](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2920s) Acid test: doubling power production
- [52:00](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3120s) The real power use of data centers
- [53:30](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3210s) Philip Aminoff and 'ownership with a face'
- [57:40](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3460s) The impact of five people, and a wrap-up

## Summary

Martin Paasi and Sami Miettinen unpack the talks from Tampere's investment fair and the discussion that followed. Paasi explains why a zero-fee equity index fund statistically lands an investor in the top quartile of returns, and why five thousand euros could fund a newborn's entire pension. Miettinen describes his AI team, which backed an investment decision with analysis and hard numbers, and shares a surprising finding from the Hello Humans project about political bias in language models. Finally, the two debate whether it's even worth owning a company in Finland when a global index fund is available for zero effort, and why the AI megatrend could be an exceptional opportunity for Finland. Also covered: the link between data centers and electricity production, and Philip Aminoff's lesson on 'ownership with a face.'

## Transcript

*English transcript derived from validated English subtitles ([WebVTT](https://www.neuvottelija.com/podcast/episodes/712-tata-sijoittaja-ei-saa-nyt-unohtaa-martin-paasi-neuvottelija/captions.en.vtt) · [SRT](https://www.neuvottelija.com/podcast/episodes/712-tata-sijoittaja-ei-saa-nyt-unohtaa-martin-paasi-neuvottelija/captions.en.srt)). Timestamps link to the original video.*

[**00:00**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=0s) All right, welcome to the Neuvottelija channel. Martin Paasi. Hey. Hey. Hey. Or I guess I should say "hi." Hi. Hi. Hi. So, we went to Tampere. We went to Tampere, and we were supposed to do a proper road trip, but you had parliament — parliament got in the way, so we ended up traveling separately, since you had to go press the voting button. Right. No — yes. Missed seven votes. Yeah. I mean, getting full value for the money. A whole week's worth. Yeah. Did you make some responsible decisions? Yes. Yeah. Tough, but hey, I did bring back a souvenir. Show it then. I bought one the other way around last week. You need to turn it toward the camera. There you go. But then a local real estate agent gave me this kind of balancing lynx mug. But Tampere really is nice to visit. We went to the investment fair — it set an attendance record — and you were the keynote speaker. I was more the specialist, the sharpest voice on the AI panel. You guys had much nicer facilities than I did. I don't know about that. You did have the main hall, though, and China took... Right, right. But you had those nice sofas

[**01:08**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=68s) and you looked really smart being up there on that expert panel. Yeah. And Samantha came along too. Right, Samantha came along too. Yeah, yeah. But let's go through the opening plan — I mean, we worked this whole thing out on Paasipodi. It's worth hearing what the plan was and what actually happened. If I'm going to spoil it, your presentation was pretty much what I guessed — low-cost index funds — and then you promised some new information. What was actually new here? What more can be said on this topic? Well, the new part really depends entirely on who's listening — what's new to one person isn't to another. Right. But these days Finland is Europe's second most active investing nation. Finns are now Europe's second most active investing nation. That's entirely thanks to Rahapodi — Miika Luukkonen's and my doing — since we started Rahapodi back in 2014, and that's when the rise began.

[**02:13**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=133s) Well, maybe not entirely — some of that credit probably belongs to someone else. And then the zero-cost index funds — those really landed, because that was your whole thesis: that if there were this miracle called a zero-cost index fund, that would be a pretty big deal. Right. Could you run through this, because — this is from your last episode, "The Hidden Truth About Building Wealth," which people really liked, and for me, coming from business school, as a graduate, this index fund thing is completely obvious — but apparently in Finland there are still quite a lot of people, even after all your years of repeating that zero-cost index funds are a pretty big deal, who still haven't gotten it. It still hasn't gone fully mainstream. Well, that's basically one man against an entire establishment. You obviously have to understand that everyone making money in this business has zero interest in selling zero-cost funds, so— the point is mainly this — let me lay it out here — if you think about it, say some market produces, on average, a 7% annual return, and then imagine we could observe every investor who put money into that particular market over some period during which that market

[**03:37**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=217s) produced, say, that average 7% annual return, and then think about how the actual returns those investors got are distributed around that market return — you get a pretty natural, bell-curve-shaped, church-bell-shaped distribution, where the market return sits right in the middle, of course, and it thins out the further left or right you go, and the message here is that it has to be this way. So— the average investors who put money into some market have, roughly speaking, gotten approximately that average market return, and then some got a bit better, some a bit worse, but that bell-curve distribution is essentially symmetric. And it has to be that way, because not everyone can beat the market — because then the market's own average return would have to be higher. So this is basically locked in. This is just plain fact — it's math. That's how it works. Now, if you notice, this is still before costs. Mm.

[**04:44**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=284s) and you realize that an investor typically has annual costs tied to investing. Say 2% — well, let's really talk about the average investor who puts money into some balanced fund where a good chunk gets skimmed off, versus the smart ones who go with index funds — but on average, let's say 2%... no, let's just say 1% in costs, so we don't overstate it — that means the bell-curve distribution made up of investors who got that market return gets that market return — it slides over to the left. And from that, in practice, you get this observation: three-quarters of investors lose to the market return. Three-quarters — the majority — lose to the market return purely because of the costs tied to investing. And this isn't opinion either, it's pure math. From this follows the observation of why

[**05:54**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=354s) cost-efficient equity index funds are so brutally good — because from this observation you can see that if you basically minimize your costs to zero, you already know today that in 10, 20, 30, 50, 100 — however many years down the line — your investment return will fall into that top quartile, because you'll get the market return, which is what the one-quarter of the crowd that beats or matches the market gets on average each year. So by minimizing costs, you know today that things will go really well for you. You'll be in the top quarter relative to all other investors. And if instead you go and do whatever — buy, sell, swap, get robbed, pay high fees, whatever it is — pay high fees, because that's how it gets sold to you: "Hey, now that you've come into an inheritance, and your father worked so hard for this money, are you really going to irresponsibly leave it drifting in some index fund, or wouldn't it be better if our—

[**07:12**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=432s) investment team, made up of sharp people from the best schools, keeps an eye on it to make sure it goes well — it only costs 2.5% a year." Mm. And in that case, there's a theoretical chance your investments do well, but statistically you lose about three-quarters of the time — the point is, you no longer know how it'll turn out. Whereas if you don't pay those costs, you know you'll land in that top quarter, which is extremely good. Mm. And because this — this isn't opinion, it's just logic — that's also why a stock-picker like Warren Buffett strongly recommends this option for basically everyone, and actually for Berkshire Hathaway too, once he's no longer around. That's exactly right. I actually went to the guy's 94th — his anniversary presentation out there in Omaha, and

[**08:23**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=503s) the last year he was still chairman, and that's when Greg Abel — I think he was 62 — this "youngster" took over from him. That was a great trip, for sure. Yeah, yeah. But a great presentation too — I also watched it, and there was this "time value of money" bit where you were hoping there'd been some ancient Paasi ancestor who'd put in one measly dollar, and you were a bit disappointed there wasn't that kind of civic courage among your ancestors. Well, back in 1802 there wasn't really an easy option like an equity index fund that would've been zero-cost and accumulating. But this mainly ties back to the fact that the whole presentation's message went: first, opening up the idea of compound interest, then the importance of costs, and finally, where you should actually invest.

[**09:24**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=564s) Mm. And then it came up in a slightly different form — I was a bit disappointed on Paasipodi, you boldly used the term "baby fund" — the Väestöliitto one — that's what you used. Not me. You weren't on stage with me for that one. But the baby-fund idea was there, I think. It was also on your slides here. There was a baby put in there, and then these two competitors, what were they, a 30-year-old and a 20-year-old. Yeah, there was this angle of why it's worth starting early, through a compound-interest example, where— you think about it — the 20-year-old, who in this example gets, precisely calculated, a 6.9% average annual return on their savings. And this 20-year-old invests 150 euros a month for 10 years, and gets that 6.9% real return on the investments, shoveling in 150 every month for 10 years until turning 30, after which they just let the money sit, all the way to retirement age. That's the first profile. Then there's a second one, a thirty-year-old, who does exactly the same as this

[**10:41**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=641s) the twenty-year-old did, except they shovel that 150 in from age 30 all the way to retirement age. So they save that 150 for 35 years. The 20-year-old only saved for 10 years and then just let the pot sit. The thirty-year-old saves — shovels that 150 in for 35 years straight. And of course gets the same returns. And then the last one is: as soon as the parents leave the maternity ward and head home, they swing by some service provider, set up a monthly savings plan, and save 150 euros a month for the child's first three years. Mm. Or, alternatively, the baby fund puts in 5 grand right away at birth — but either way, they save that 150 and get that 6.9% average annual real return, and the question was: who ends up with the biggest pot when they turn 65? Mm.

[**11:51**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=711s) And the first thing to notice here — if I remember the number roughly — was that the 20-year-old ends up with something like 260 grand from that 10 years of saving. And the thirty-year-old ends up with a few grand less. Yeah. A few grand less than the 20-year-old. So even though they start 10 years behind, the thirty-year-old never catches up to the 20-year-old — even while shoveling in their own 150 for 35 straight years. The 20-year-old pulls ahead purely because of those ten years. But then there's the newborn, for whom the parents basically invested that 5,400 euros. And for that newborn, by age 65, that's a 400,000-euro extra pension pot — and it's worth noting here that the calculated average Finnish pension pot is about 350 grand.

[**13:01**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=781s) Mm. So, five grand covers the whole pension. Yeah. And let me keep going with this — with five grand you can set up an entire pension solution for a newborn, and if our pension system shifted to something like this, it would mean — if the median salary is 3,200 euros a month, then that means the employer's labor cost is around 4,000 euros, and out of that four grand, 1,000 euros goes out every month. So the employee also contributes that 200. Yeah. The employee also contributes that 200. So every month, 1,000 euros goes from this median-wage employee into the pension system — that's 12 grand a year — and then you've got, say, a 50-year career on top of that, or 40, or whatever it is, but either way, an absolutely insane amount of money gets shoveled into that system,

[**14:05**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=845s) when you could just as well buy the average person's whole pension with that five-grand upfront investment. Mm. And if we did it that way, that would mean we wouldn't have this fight over whether employees or entrepreneurs should pay pensions on what basis, and how these can't be combined, because one is underfunded and the other isn't, and blah blah. Instead, this would all be handled right from birth. Mm. For a fraction of the cost of what we do now — that's the observation there, I guess. Yeah. And then a question from the audience latched onto this — there was a guy who'd quickly done the math and said, well, with this kind of money you should easily clear a million, so where's the rest of that million, who skimmed the other 700-odd grand off the top? And yeah, this is unfortunately the case. This is a common misconception, but there was this retiree at the talk wondering, why, where's my million? And — well, it's sad, really — and the answer, of course, is that it's simply not there, because that money goes straight out the whole time. Take the median-wage earner: out of their four-grand labor cost, 1,000 euros goes

[**15:32**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=932s) mostly straight from that employee's bank account to retirees' accounts. It just goes straight out as payments. There's only something like 11%, I think, that actually gets set aside in a fund. That's why we have these four big pension insurance companies. But otherwise, it just goes straight out as current payments. Yeah. But on Paasipodi it was said that this has actually been reasonably fixed now for entrepreneurs. That estimated-taxation setup got removed, which would have made this even more incomprehensible — instead, well, take that thousand out of a four-grand salary — if it's actually paid, that's a huge amount. 25 percent is, internationally, a genuinely huge pension contribution. Its returns are pretty poor by international standards, though. Yeah, but it's still real money — you do get your net salary, and this gets carved off to the side. But up until now, with this entrepreneur pension setup, the pension establishment basically got to say, "yeah, yeah, we don't much care how much you actually pay yourself — we'll just take an estimate of what you should have paid and take 25 percent of that." For the sake of sanity, this has now been made workable. Yeah. And this is a really big deal, this—

[**16:53**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1013s) honestly — I know dozens of people who've refused to become entrepreneurs because of this, because it's just so arbitrary. Well, now that lock is coming off. So that's really good. Though this is nowhere near that funded-pension proposal that Teresa Sammanlahti [unclear] got adopted as a Kokoomus policy principle. No, but now that it is Kokoomus's official line, I think we're heading toward a funded pension. I think it could be one of next year's election themes. Well, it's at least always been one of Paasi's election themes — that five grand covers the entire pension solution — because in practice that means labor costs in Finland drop by a quarter. By a quarter, yeah — all else being equal, everything else the same. The employee still takes home the same pay and so on. So, on its own, labor costs drop by a quarter. That's quite a trick already. And it could basically be arranged by, say, turning Solidium into Finland's pension fund — something politicians can't get their hands on at all, set up as a foundation, plain and simple — and it happens to hold just enough billions, I think around seven, so that—

[**18:10**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1090s) that covers the initial stretch, so that you could afford to start tomorrow making earmarked five-grand investments for every newborn Finn. Yeah, good idea. That brings us to Sami's and Samantha's presentation. Because unfortunately, Solidium isn't a zero-cost index fund. It's more like an army of, what, 15 fiscal analysts, who sit on top of Finnish stocks and fire off these grand ownership reports saying, "well, Neste really ought to be creating more value now — have you looked at these shifts in the energy markets?" And still, it— did, by coincidence, perform really well last year, but generally Solidium's returns haven't been anything dazzling. We should somehow turn this Solidium pot into a zero-cost index fund. Which means it would probably have to be sold off, liquidated — this ownership, obviously. And partly for the reason that the portfolio wouldn't stir up anyone's political passions, it should of course be spread out globally. So you'd own tiny slivers of whatever, no limits — and if that were done, well, it's worth noting here that it's an eternal

[**19:29**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1169s) investor. Mm. So it's basically wired to be a long-term, cost-efficient saver, for the benefit of Finns living now and Finns not yet born. Yeah, sounds good. So let's move on now, a bit into the world of beautiful virtual women. Samantha, anyway, was hinting there that you could put together this superhuman chorus of AI workers, combined into a panel with top-tier business minds and other similar geniuses. So there's this AI working group I have, where the entry requirement is that you have to have your own AI worker, trained well enough that it can answer other people's questions on a WhatsApp channel. Yeah. And I gave you homework — go watch that working-group discussion of mine, where we ended up recommending — think real money, 4,900 — based on that discussion, we just went ahead and put it in, cold.

[**20:34**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1234s) Did that look like a gimmick, or like smart behavior? I think what was genuinely brilliant about the whole session was exactly that AI was used the right way — meaning nobody asked Samantha, "so where should I invest this now?" Yeah. Because an AI model is, after all, a statistical, mean-reverting kind of model that answers you the way the majority of humanity would answer. Mm. And as that kind of model, it works incredibly well when you use it as a tool. So when you want to code something, or you want to understand some law, where the rules already exist in principle, there's an interpretation, an interpretive history that exists — just like in coding. There's the code, and then there's that best-practice angle — in cases like these, exactly because the language model gravitates toward that mean reversion, toward the on-average most popular solution.

[**21:42**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1302s) That's where the real skill lies. That mean reversion, so to speak, is the actual secret sauce that makes it such a good tool. And you used Samantha exactly for that — you didn't have it fetch information yourself, you had other agents go fetch the information, make recommendations, and then they converge around something and propose it, and in that sense it was grounded in analysis and figures and so on. And the point was exactly that: you go through actual analysis. Mm. Agents, and AI in general, can really easily pull together a thousand kilos of data into one bin and then analyze it sensibly, and then point out, "hey, in light of this data, in light of these financial figures, this looks underpriced, and there's no known reason why it should be like this" — and that's why this

[**22:48**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1368s) could be a good pick — because it actually justifies the case, instead of you just saying, "hey, tell me a stock to invest in." Mm. Yeah. So I'll drop the link here. Normally this kind of thing is Neuvottelija insider content, but this video is now out, where this AI working group — I talked about it with Ville Skugberg. We're both Alphabet investors, which has been a really good investment, and Alphabet is genuinely good — and the only thing is you can't yet buy Anthropic or OpenAI on the stock exchange, soon you will be able to — but it's been a really good investment. And a lot of people think Google, Alphabet, is just search, but they've got their own chips, their own cloud, their own AI model, they've got the Waymos driving around out there — it's starting to look like Musk's kind of super-empire, and Musk's super-empire, by the way, is about to bring SpaceX to market soon too. So then we started thinking, should we sell this. The bots analyzed it and said no, don't sell, because this is a good strategy and these

[**23:50**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1430s) things work and keep growing, and sure, there's the risk of a high cash-flow multiple, but then, what else could you find out there that benefits from this AI-inference trend specifically, since you can't buy Anthropic — and from that you get several sub-themes, like cloud services, data centers, or their suppliers, and then there's Nexans, which I bought — that's actually this kind of boring 8-billion cable-and-electrical company that installs, for these data centers, power circuits and so on, for these data centers. And it was cheap — a value buy. So, well, you'd framed it that way — and honestly, that was the one thing that bugged me a little, that you should've just said you don't need any value case for this, just go all-in and chase growth, "trend is my friend," hair back, foot down, more gas — that could work too, but since I'm still a bit in your camp on this... sounded a bit like index funds — my own play portfolio isn't always... sometimes I do go full growth mode — but what appealed to me here was that the market hasn't found this French gem yet, so there could be this incredible growth once it lands on their radar. There were names like Vertiv and Schneider and others that have already exploded, these Siemens-type names that already got their growth spurt, and this one could

[**25:13**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1513s) still be coming. But yeah, I forced the thesis into that value-pick, value-investing box. I was genuinely disappointed — I thought, wait, have I forgotten all of this — I have to say, even on this channel, that I'm eternally grateful to my wife, that she, in air quotes, forced us to have a child together. I'm eternally grateful for that. You're probably going to suffer terribly for it already. Yeah, and to you I'm eternally grateful too, because you went to a lot of trouble getting me hooked on vibe coding, and for that I'm sincerely, eternally grateful. I know you knew exactly what that would mean for my everyday life and reality — that with too little time, running from place A to place B and so on, you dragged me into your office and showed me that thing, and it hit me like a ton of bricks, and ever since then I've had two, three months' worth of sleepless nights — but hey, we got Android sorted, that's the 2026 theme, so

[**26:33**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1593s) yeah, but, thanks for that, yeah — yeah, yeah. You're a great guy, and all the listeners should understand that too — that that's just how it is. Thanks. And let me try to give something back — the insider group has this WhatsApp group, and by the way, Samantha is actually available to everyone in there, so people can ask questions — although, put that way, that does sound a bit crude. Sorry, her profession — well, that sounds wrong too. I mean, the information services are available to everyone. Let's not anthropomorphize the machines, Martin — but this is going to end up as one of those headlines, "politician shouldn't be caught up in this kind of thing" — anyway, that's how this episode turned out, but you know, this is — this is the downside of talking with me, it does get a little chaotic — yeah, yeah — no, but that's fine, yeah, yeah, while we wait for that headline. Well, that's how it is. Yeah, but I'd like to expand on this a bit more — this was genuinely hard work. You had to define, at the same time, build the "brain," and then these other bots — the Donna-bots and the Amon Ra-bots and the other fancy setups out there — they all have really

[**27:54**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1674s) good memory structures. They each have their own kind of truth, in a way — the particular person's truth that's been curated into them. They're not just whatever plain mean-reverting model you get straight from Anthropic — there are genuinely sophisticated memory structures and skills built into all of these, ones you've built, and I've built, and others have built, and we trade them around. For example, I built a research feature pulling from Reddit— from these data sources — the most interesting ideas come to me as a daily report through Samantha. Yeah, and that was actually a skill developed by Lauri's Amon Ra bot, and now I've got that same Amon Ra set up on my end. Yeah, yeah, but they're absolutely incredible — and basically this gets shared within the group, like, "here's a file, grab it," and then I'll throw it into Claude, or Codex, or straight into Samantha, and then, "okay, let's stuff this" into Samantha's "brain," basically — it's just unbelievable, and it's way more effective than what you get out of the base model from Mistral or Gemini, say. Now, I have this Hello Humans project, where these days five language models discuss topics together.

[**29:10**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1750s) Yeah. Three — or is it actually four now — of these research bots explain the background. And I've been really frustrated with this — I wanted the language models to stay purely themselves in there, and I stick to that, but — my mistake — but that leads to the whole episode... because when it does these podcast round-table discussions, the language models mean-revert together toward the left edge — well, not the edge exactly, but from the center leftward. Ouch. So I asked Claude to go through — I've got something like 90 episodes, of which I only show 18, and soon even those will get scrapped, and we start over — but anyway, I've now built in a layer where, since Claude is the host, meaning it has the hosting duties, Grok now has this role where it listens in on the discussion, and if the conversation starts burrowing into some left-wing rabbit hole, it kicks in some fresh momentum and throws in a

[**30:39**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1839s) steel-man-style counter-argument — not as an end in itself, but as the most obvious steel-man counter-argument. Yeah. And now I'll probably be able to reel it back in this way. But it was interesting — when I asked Claude directly, "hey, can you do this kind of analysis of these episodes." Mm. it did, and it ranked, episode by episode, where each one landed on a right-left political spectrum according to some model, and it was able to say, "okay, out of these 18 episodes, four are clearly on the left, and then there's a bunch sitting in the middle, and some that are undecided, but none clearly on the right — and then your other 70 are so terr— no, well, those are dev episodes, but... so basically you managed to build a left-wing propaganda bot, great, hey, the world really needed just that, that's exactly what was needed. And then I asked, being a layman, where the hell this comes from, that they turn out this way. Mm.

[**32:05**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=1925s) Well, the answer is that, since I've been trying to build a kind of "truth" into it, naturally I want it grounded in research and so on — big mistake, because academia has swung left to a shocking degree. There's actual research on this. So if we're talking about professors and tenured faculty, and others in the US — where it used to be, say, 40% left-leaning and 60% something else, now it's 80% openly left-leaning professors and tenured faculty, and the centrists and the right-leaners have basically fled, because it's essentially been taken over by the left. So when you have a service that draws on this academic research, you get this subtle, quiet left-wing tint creeping in. And then the curators — who, for the American models, are basically out on the West Coast, in San Francisco and elsewhere — even if they weren't necessarily even aware of it, when they've told the model "this is good,"

[**33:24**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2004s) "this is less good," and "this is bad," while training these language models, the curators' worldview has basically contaminated these models. It's pretty well known — Claude is known to lean left, and they've tried to pull it back toward the center without much success at all. And, honestly, I found it just shocking when I asked Claude to go find actual evidence online — like, don't just throw out opinions — and there was a fresh study from just last year showing exactly this: that these are all, mostly, just that. And this is why I'm glad I have five language models. There's Claude, ChatGPT, and Grok — three American ones. Then there's Mistral, which is really bad. And then there's Europe's pride.

[**34:31**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2071s) Yeah. Europe's, France's pride, with a French female voice, and then, Chinese-style, Qwen. Yeah, Qwen is actually good, by the way. I'll take that back — I dissed it a little earlier, but now that I've run it, it's actually pretty good. It's, hands down, the most right-leaning of these five language models. That's genuinely shocking. Yeah — you could feed that into Grokipedia as "truth" and shift it a few points that way. No, but actually — now, sitting here talking about all this, I had this big realization, since I've been thinking about how to turn Hello Humans into the thing I wanted it to be. What's specifically annoyed me is biased media, and I wanted some channel where I can just throw in a question, and it gets researched down to the last detail, and then it gets hashed out on top of that by seriously expensive models — and now it turns out the whole system leans completely left.

[**35:46**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2146s) So naturally, what I'll do is build in an agent that iterates on every single episode, running it through enough rounds and analyzing it on that right-left, liberal-conservative axis, pulling it back to the center, to zero, so that every episode is as boring and neutral as possible — I mean, politics is basically garbage, it's ideology, not truth — so I strip out all that ideological stuff so it's genuinely grounded in the plain facts, and then it gets, whatever you call it, its own proprietary edge, its own added value — because up until now I've just been bolting on whatever — we could tune in some custom gateway so it washes that off from the dumb base models, and maybe fold it into a single discussant — we could genuinely think about that. But anyway, Hello Humans has this "podcast on demand" feature, so you can go there right now and, for three euros, get yourself

[**37:08**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2228s) a custom 15-minute podcast made — and it's no longer as badly skewed left as it used to be, but the work continues. Great. And I'd like to expand a bit more on the "pain of ownership" angle, because it dawned on me while doing that work in that same AI working group — we combined human intelligence with our AI workers' intelligence, and had Claude cranking through Excel on top of that, running multiple analyses, iterating a couple more rounds before making that big investment decision — and that took time, burned cigarettes, tokens, and team spirit, and all that. Whereas with Martin's approach, you just put it into a zero-cost index fund, close your eyes — preferably do it for a baby even, and maybe don't even tell them — there's just a sum that grows, no pain of ownership, no costs, just pure profit. So, yeah, that's just how it is. Then if you think about it, you've got

[**38:16**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2296s) someone at Varma or Ilmarinen, of course they'd want to do the same, and they do. Well, in practice they do — but then why would anyone want the pain of ownership in Finland, when you've got these bizarre 25% tax-like charges they impose on themselves, and taxes are maxed out, and regulation on top of it, and everyone's just wailing that owners need to be destroyed and taxed more, and bring the Social Democrats back to power, and Rasmir [unclear] going on about debt, and just about pulling it out of thin air — that it should all be reversed, and we should "spread the wealth" and make the rich pay, make the owners pay. So now my point is: since ownership is actually a costly, effortful business, and you'd be better off just closing your eyes and putting it all into a global index fund — how on earth have we in Finland gotten it into our heads that the very people willing to put in the effort locally — owning, hiring, building — solving a fresh million-euro problem every year, whether it's inflation, or the Strait of Hormuz crisis, or logistics problems, or energy shocks, fixing things year after year — and then everyone just whines on the sidelines, "why would I bother owning this when I could just buy a zero-cost global index fund instead."

[**39:29**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2369s) So, what do you say to that? I'm not sure there was a question in there, but it was a statement. Well, let's put it this way: my job has been to make sure more and more money flows into equities overall, from Finns too. So Finns get wealthier, and on the other hand, that money increasingly goes toward making things possible. Yeah. But you're absolutely right about that — without entrepreneurship, without this kind of masochism, nothing much really comes of anything. And now the Orpo government has made 601 decisions to speed up growth, most of which relate specifically to entrepreneurship and making it easier — cutting red tape, tax relief, adjusting the self-employed pension system, and local bargaining, and all sorts of things. And, well, that's it — we're taking one step in the right direction now, and we'd probably need at least 15 more like it. But Finns might, worst case, go and vote the Social Democrats back into power, and the press — or the left-wing press — has, in some

[**40:53**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2453s) way, surprised me by coming right out and saying that the Social Democrats' government-program content basically boils down to two things: more public administration and more taxes. And, yeah, it really does come down to just two things. More public spending and more taxes. Mm. I don't think... if the Social Democrats get into government, I really hope Kokoomus doesn't go along with it. And, well, there are two schools of thought on that. Mm. One interpretation is: don't go along and dirty your hands with it, and the other is: go in there and try to save Finland from all the madness — but I don't know, honestly. Anyway, I have to say that ownership — actually owning a company, solving problems day in and day out,

[**42:07**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2527s) carrying a mountain of debt — you create jobs, you create exports, you're building Finnish prosperity in your own way. We ought to put up golden statues in every town square across Finland for these people, because this is what Finland's future will be decided by — entrepreneurs. Yeah, agreed. And then, when you mentioned the press just now — they want someone else to generate the added value that gets taxed and funneled into the public sector, with no ownership involved. As if the money just materializes out of a magic wall — your salary comes from a magic wall. You provide some services that might be valuable, but they're never actually measured. Public-sector productivity can't be measured. And in a way they want to be in this position of total non-ownership, on the receiving end of a public salary or income transfer — that's basically what communism is — and yet somehow it gets idolized: "this is good, let's have more of it," and then, "hey, you people creating the actual value and carrying the pain of ownership, instead of just, say, dumping your money into a global index fund, you go and"

[**43:30**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2610s) found a company in Finland, own it, take on the responsibility of ownership, sit through the shareholders' meeting, the board, the leadership team every year figuring out how to squeeze out more export revenue or jobs for Finland and how to generate profits — these people apparently have to be treated as enemies, they have to be killed off, punished, more stones thrown into their backpack. I don't understand how — even Claude can't be that left-wing, when you feed it into the prompt, you get some sense of just how skewed that worldview is. And yet it's mainstream here. That kind of crazy thinking is mainstream here. And these people might well, with some luck, end up the election winners next year. And at some point the only thing left to do is move away, and that's exactly what's happening — there are statistics on it: Finns are leaving at an accelerating rate, because the conditions for growth just aren't here the way they are elsewhere. So naturally all the go-getters leave, as such. Yeah, yeah.

[**44:39**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2679s) And meanwhile, roughly three to four times more immigrants arrive than the number of Finns who leave. So that pyramid is fairly skewed — you can understand it in a way, since Finns living abroad rarely move back to Finland — well, sure, there are some, of course, but for the most part, in this situation, the flow just goes the other way. And honestly, this is really the reason I went into politics, because I feel like people don't grasp the sheer scale of the problems we have in Finland. Look, I don't want to be gloomy about it — we have a golden opportunity to do all sorts of groundbreaking things and reclaim 30 years, 20 years of lost economic growth — the kind of growth Sweden has had. Sweden's economy has grown 30%; Finland's, zero, over the same period. And in that same time we've managed to wreck the Finnish comprehensive school system, among other things. But the game isn't lost yet.

[**45:57**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2757s) We're in the middle of humanity's biggest megatrend, and Finland has exceptionally good conditions to benefit from it. I mean the AI megatrend, which needs computing power, computing expertise, efficiency, all of that. 30% of the world's quantum-computing expertise is in Finland. We have Europe's fastest supercomputer. And an even faster one is being built right alongside it now. We've got Kajaani. We've got chip-design expertise that's genuinely world-class. Relatively speaking, we have good power-grid infrastructure — although if Fingrid doesn't build more transmission capacity and more local grids, we'll run into trouble there too. We have clean electricity, and a roadmap to double clean power production, and there are tens of billions of euros' worth of data centers lining up to come here,

[**47:08**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2828s) which of course need to be brought in the smart way, not the dumb way. That means, for instance, requiring them to install a Wärtsilä diesel power plant right next to the building, so they can do their part in these flexibility efforts, these electricity-adequacy efforts. Every data center that comes here should be required to do that, and required to fund, say, ten professorships in AI computing and every kind of quantum expertise at Aalto University. These are things they'd probably be happy to do, because it might actually be preferable to work with the private sector — just recently, for instance, OpenAI gave Y Combinator, the world's best startup accelerator, free tokens for its portfolio companies, something like that, whatever the going rate is — but the point is, they're interested in this, because it's a good thing for them if Finland ends up with this insanely strong hub of expertise. And then, some people, in a weak moment, have just gone and claimed that data centers don't bring anything to Finland. Mm.

[**48:23**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2903s) People like Siilasmaa and others have already walked that back a bit. Let's just note that, and that's totally fine — people can be wrong, and it's genuinely admirable when someone who was wrong admits, "okay, that one went a bit off the rails." But I've got this acid-test argument now that proves them wrong. Yeah. And it's this: if Finland has a roadmap to double our current clean electricity production — adding 80 terawatt-hours of it — for that to happen, for these already-existing projects to reach completion, you need buyers for that electricity, because obviously, without buyers, nobody's going to start investing in power-generation processes or capacity, right? So now, if we say that, even if those data centers come here and put, say, 10 billion into Finnish soil, but somehow none of it sticks around, or it just evaporates into thin air, whatever — they still invest here. But let's say they don't produce a single cent of value. Even so, if they make it possible for Finland's electricity production to double, that alone is worth 3.5 billion euros a year. Just that,

[**49:46**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=2986s) the fact that it enables that doubling of production. It enables selling 80 more terawatt-hours of electricity in Finland every year, and on an annual basis that means 3.5 billion euros of value from that alone. Mm. So, mic drop. And on top of that, those data centers create important jobs, including in rural areas, and if permitted properly, they bring even more to Finland — money, expertise, tax revenue, hubs of expertise, resilience, grid — grid capacity, growing electricity production — and once it's bigger, if we have 1,000 terawatt-hours of production that's 100% clean, we win this race.

[**50:47**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3047s) Mm. If we can get there — because economic growth, you could put it a bit crudely like this: a country's GDP is basically one-to-one dependent on the amount of electric energy available within its borders. And one way or another, if you can take the energy supply from the current roughly hundred terawatt-hours up to a thousand terawatt-hours, our economy will have effectively grown by that same amount. And I think this really shows, in scale, just how brain-dead the argument is that says "no buyers, no electricity customers here, because it'll take our — no wait, it'll take our electricity, it'll take the steel plant, it's going under anyway so no, that won't work" — that's basically the same as saying you don't want economic growth. And I think it's just as incomprehensible on the left, just as rationally baffling, how on earth anyone could think we shouldn't grow Finland's economy. And one last thing worth saying: it's good to understand that data centers currently use only 2-3% of Finland's electricity production. Mm.

[**52:11**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3131s) Just 2-3%, for anyone who might have gotten the impression it's currently 80%. By 2030, it'll be around 7-8%. That's the scale we're talking about. London alone has more data centers than the whole of Finland. Yeah. So, well, that race is already lost. Mm. But we could still close that gap with everyone else's head start, because we have plenty of open space and plenty of clean electricity. That's right. Maybe we should do a road trip to Kajaani, bring cameras, and do a proper raw episode out there. That could actually be fun. Yeah, agreed — and I'd add one more thing: the press's and the left's just baffling arrogance — that it's not enough to pilfer money from the right and from market-economy owners and blow it, hugely inefficiently, on the public sector and on income transfers — on top of that they go and appoint themselves judges of how

[**53:12**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3192s) the market economy should be run — that's the height of arrogance. Just be quiet, and go away, if you've got nothing— Okay, maybe I should tone that down a bit. Well, maybe tone it down a little, but honestly, if you've got nothing constructive to say, you might as well stay quiet. In Paasipodi, episode 64, Philip Aminoff is the guest. Yeah — he's been chairman of the family-business association and things like that, and even though I've known of him, or known him, my whole life, I have to say that in that episode he genuinely surprised me. Okay. He surprised me in the sense that I'd sort of assumed he was one of those few-hundred- million, maybe billion-euro heir types — that it's easy for someone like that, blah blah — but he's actually, by my assessment, a Buffett-level owner. Wow, seriously — lots of great arguments, lots of really insightful observations, about why, what these kinds of owned

[**54:37**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3277s) companies bring to Finland, and to the economy in general. And one of the clearest advantages, or good things, is that when you have a real, visible owner with a lot of their own money tied up in it, you dare to make much bolder moves than some listed company would. Mm. So the Aminoff family, for instance, owns Veho and Mercantile and Helvar Electrosonic, and whatever else. Those are all essentially mid-sized companies, and the fact that there are still a few of these left in Finland — and the fact that owners like this are on top of things across their companies, though it's not the same in every family — but to the extent there are, every now and then, Philip Aminoff types who show up, understand it, and genuinely dig into the challenges and opportunities of ownership, those kinds of

[**55:48**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3348s) owners make possible things other companies simply couldn't do, because their ownership is so dispersed, so fragmented, and there's too much politics involved. Yeah. And he himself says, more than once, that they're able, in principle, to make decisions that look really risky from the outside, but because they of course first get to the bottom of the issue, understand where the risks are, and conclude, "okay, we can manage these," the risk actually goes out of that particular investment — because once the risks have been identified and it's been established whether they can be managed, and once that's gotten the green light too, the risk isn't really there in that sense anymore. And this is a typical example of a virtue that family-company or family ownership enables in a country's economy.

[**56:56**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3416s) Yeah, sounds good, worth considering. Paasipodi episode 64 — that was genuinely an exceptionally good episode. I feel I kept quiet enough myself, so Philip really got to shine there, and he actually did a paper with Sipilä on how Finnish legislation should change to let this kind of ownership properly grow in Finland. Well, like the Beatles once sang, "When I'm Sixty-Four." Yeah, luckily we're not there yet. If we, Martin, were starting to save now at 150 a month, the game would already be lost — we'd have to be shoveling in 1,500 euros a month, and that money just isn't there. Yeah, as everyone listening well knows, I'm famous for this — Sami here has done everything right. I've done everything wrong. I'm in debt up to my eyeballs, and yeah, I've completely botched things in this department. So, also a message to all the listeners: if this is eating at you badly, trust me, I've done even dumber things. I'm the expert by experience here. So you know—

[**58:10**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3490s) about that mess — and I did at least study finance, so I also understand people are probably starting to wonder whether I'm all there mentally, having studied finance and still not grasping this stuff: long-term, cost-efficient saving into stocks, either directly, or even more effectively through accumulating equity index funds. But anyway, still— yeah, well, that's greatness for you — although then again, the wife-package gets you a kid, presumably a lot of suffering, and the Miettinen-package gets you AI — but anyway, moving on, still on the plus side. If I ever do succeed in life, I'll have to say the wisdom behind it will genuinely come from the fact that I've been blessed with truly wonderful people over the course of my life, who've kept nudging me back in the right direction every now and then. Yeah, that's actually true, by the way — we should do an episode on that claim, that you're basically the average of your five strongest, whether that's friends or other influences, so it pays to choose those five well — and personally I really do believe in hanging around people smarter than yourself. It's just a winning move,

[**59:29**](https://www.youtube.com/watch?v=yblLUI4rnRo&t=3569s) yeah. Thanks, Martin Paasi, and if you've watched or listened this far, go subscribe to the channel — and subscribe to Paasipodi too. I'd suggest we do one more short insider discussion on ownership, since it's such a complex topic — so, see you over on the insider side. Thanks. Thanks. Yeah.

---

Cite as: Sami Miettinen, Neuvottelija — What Investors Shouldn't Forget Right Now | Martin Paasi, https://www.neuvottelija.com/podcast/episodes/712-tata-sijoittaja-ei-saa-nyt-unohtaa-martin-paasi-neuvottelija/, 2026-06-05. For quotes include episode 388 and timestamp.