--- title: "Is AI in a Bubble? | Ville Tolvanen, Timo Savolainen" titleOriginal: "Tekoäly kuplassa | Tolvanen Savolainen | #neuvottelija 364" episodeNumber: "364" guest: "Tolvanen, Savolainen" datePublished: 2025-12-08 duration: "01:04:17" youtube: "https://www.youtube.com/watch?v=IrWui2k4j8I" captionsVtt: https://www.neuvottelija.com/podcast/episodes/650-tekoaly-kuplassa-tolvanen-savolainen-neuvottelija-364/captions.en.vtt captionsSrt: https://www.neuvottelija.com/podcast/episodes/650-tekoaly-kuplassa-tolvanen-savolainen-neuvottelija-364/captions.en.srt originalLanguage: "fi" topics: ["ai_enterprise_tech","investing_markets"] subtitleMethod: "publisher-english-cues-imported-and-qa" provenance: "neuvottelija.fi Supabase episode-markdown API (platform captions)" fiCanonical: "https://www.neuvottelija.fi/fi/episodes/650-tekoaly-kuplassa-tolvanen-savolainen-neuvottelija-364" canonical: https://www.neuvottelija.com/podcast/episodes/650-tekoaly-kuplassa-tolvanen-savolainen-neuvottelija-364/ --- # Is AI in a Bubble? | Ville Tolvanen, Timo Savolainen ## Chapters - [00:00](https://www.youtube.com/watch?v=IrWui2k4j8I&t=0s) Ville Tolvanen and Timo Savolainen in the studio - [01:30](https://www.youtube.com/watch?v=IrWui2k4j8I&t=90s) The birth of the Digitalist movement, the dot-com bubble, and parallels to today's AI euphoria - [02:53](https://www.youtube.com/watch?v=IrWui2k4j8I&t=173s) The Radiolinja era, mobile services, and lessons from the dot-com crash - [04:16](https://www.youtube.com/watch?v=IrWui2k4j8I&t=256s) Finland's "internet winter" in the early 2000s and parallels to 2025 - [05:39](https://www.youtube.com/watch?v=IrWui2k4j8I&t=339s) The Digitalist movement, Apple's counterculture, and the slow adoption of digitalization - [07:02](https://www.youtube.com/watch?v=IrWui2k4j8I&t=422s) Finnish skepticism toward e-commerce and the decade-long shift to online sales - [08:25](https://www.youtube.com/watch?v=IrWui2k4j8I&t=505s) Savolainen's childhood: landline phones, modems, and everyday technological change - [09:48](https://www.youtube.com/watch?v=IrWui2k4j8I&t=588s) Creative destruction in Säynätsalo: three shops and two banks down to one surviving service - [11:11](https://www.youtube.com/watch?v=IrWui2k4j8I&t=671s) Agentics networks, the Digitalist legacy, and ecosystem thinking - [12:34](https://www.youtube.com/watch?v=IrWui2k4j8I&t=754s) How AI is used for coding, investing, learning, and agentic browsing - [13:57](https://www.youtube.com/watch?v=IrWui2k4j8I&t=837s) Timo Savolainen doesn't ask if you use AI — he asks why you don't - [15:20](https://www.youtube.com/watch?v=IrWui2k4j8I&t=920s) Ville Tolvanen's use cases: health data, self-help, investing, and weekly reports - [16:43](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1003s) Gemini vs. ChatGPT experiences and accelerating technological progress - [18:06](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1086s) The AI bubble debate: Nvidia and Palantir valuations and depreciation practices - [19:29](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1169s) Three areas of rapid breakthrough: investing, coding, and a new era for consulting - [20:52](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1252s) The changing role of IT consultants, the difficulty of buying AI, and companies declaring themselves "AI companies" - [22:15](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1335s) Y2K vs. "2030K": technical upgrade vs. a shift in thinking, and AI's fourth wave - [23:38](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1418s) Technology waves compressing: from IT to ICT to digitalization to the AI era - [25:01](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1501s) Gemini game demos, on-the-fly rendering, and next-generation interfaces - [26:24](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1584s) Palantir analysis: deep data, consulting, security, and business models - [27:47](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1667s) Microsoft's stake in OpenAI, valuations, and strategic position in the ecosystem - [29:10](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1750s) Monetizing technology, slow adoption, and resistance across industries - [30:33](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1833s) AI in healthcare: the potential for an AI assistant to support doctors - [31:56](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1916s) Electric cars as an example of slow disruption, and the parallel to AI adoption - [33:19](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1999s) Turning the "Delete, Delegate, Do" model into logic for directing AI agents - [34:42](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2082s) Why individual AI skills matter: the "Super Me" idea and the birth of new jobs - [36:05](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2165s) The "agentic state": Ukraine, Estonia, and examples of the public sector's next wave - [37:28](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2248s) The public sector's reluctance to change vs. the pressure to innovate that crises bring - [38:51](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2331s) Nest AI, Tesi, Peter Sarlin, and the role of Finnish capital in lifting the AI wave - [40:14](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2414s) The Finnish mindset — a lack of hunger and the need for a new kind of tech optimism - [41:37](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2497s) Mapping individual AI roles: the AI banker, AI fisherman, AI teacher models - [43:00](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2580s) The need to tell stories about AI professionals to inspire and accelerate learning - [44:23](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2663s) Sami's multi-location lifestyle and building his own agentic task-manager app - [45:46](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2746s) What AI means for productivity amid the disruption of project and knowledge work - [47:09](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2829s) From passive tools to active agents: the "Pera" example in the management team - [48:32](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2912s) The myth of the "digital leap" — the linear reality of technology adoption - [49:55](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2995s) What's different about AI transformation: from tools to processes to culture change - [51:18](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3078s) Knowledge work disappearing as agents take over: a new division of labor and skills - [52:41](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3161s) Individual learning, exposure, and the psychological shift in attitudes toward AI - [54:04](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3244s) Tolvanen's critique of misplaced priorities in the AI ethics debate - [55:27](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3327s) Youth mental health, TikTok algorithms, and technology's everyday impact - [56:50](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3410s) The Neuvottelija empire: banker, media producer, coder — hybrid careers ## Summary AI in a bubble | Tolvanen & Savolainen | #neuvottelija 364. This conversation rides the crest of the AI bubble wave. Talk about AI — and investing in it — echoes the tech and dot-com bubble of 2000. A decade after that bubble burst, Ville Tolvanen founded the Digitalist movement, and Timo Savolainen has worked alongside Sami on the rise of Agentics Finland. Who does the AI wave sink, and who does it lift? ## Transcript *English transcript derived from validated English subtitles ([WebVTT](https://www.neuvottelija.com/podcast/episodes/650-tekoaly-kuplassa-tolvanen-savolainen-neuvottelija-364/captions.en.vtt) · [SRT](https://www.neuvottelija.com/podcast/episodes/650-tekoaly-kuplassa-tolvanen-savolainen-neuvottelija-364/captions.en.srt)). Timestamps link to the original video.* [**00:00**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=0s) Alright, welcome to the Negotiator channel, and our guest is Ville Tolvanen Via Piedmont to Slush, here at the negotiator studio, and I must say a little detail, that when I was just arriving at the Apollo nightclub's like, Singa party, the guy was already leaving, so what happened to your endurance. That's how it went, a long week. Thanks for the invitation. Great to be here again. Yeah, and your original merit, meaning this kind of building of the Digitalist movement, where once again everything changed on the IT side especially, and we can think about whether AI is then a bit like this digitalist movement, and then I also have Timo Savolainen and a multi-published author. Here is the book on AI's competitive advantage and then we met at Mikko Alasaarela's Agents Finland event, where you were a speaker in that second seminar. I missed the third seminar. [**01:00**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=60s) We'll get to that soon. I was at Google's DeepMind Lord Boat in exclusive company drinking champagne there, so number three will be skipped, but if you could give an update on what was being discussed in Agent Cosmos 3 regarding the AI field. But which one should we start with? Maybe since he's the senior gentleman, Ville gets to reminisce about this Digitalist? It's terrible when you've reached this 'old man' category. I am, after all, the oldest guy here. The young challenger has reached the end of the road. Actually, I'll start from even further back. So, 25 years since the dot-com boom, the future is calling. the future calling. From another direction, yeah, so the first well, the first tech booms have been happening since the mid-nineties, but the first tech boom of our generation was at the end of the 20th century, with those 3G licenses and dot-com companies that had no revenue at all but huge valuations. And '99 was such a hot year, when stock prices were rising and everyone—all [**02:11**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=131s) companies that were hiring people received a huge valuation, and back then I was at Radiolinja in charge of WAP services. as a manager, as a young man, and we created IOBox, IOBX, and various companies. At that time, Radiolinja's most popular SMS service was Noppa. So when you sent the word 'noppa' via text message, it returned a number from one to six, and it was used quite a lot, and then when the dot-com bubble burst and Nasdaq fell something like 89%, so we thought, luckily we have this mobile internet, meaning these mobile phone value-added services won't collapse. Well, by the time we reached 2002-03, there was quite an 'internet winter' in Finland, and Finns pretty much wanted to give up on the whole internet. It's interesting, and the reason why I'm telling this is that now, at the end of 2025—it is November—there are many of the same features as in '99. It's explained differently. Now they say, 'hey, these companies have massive revenues, this is profitable business, and there is a future,' but there is the exact same euphoria, over-excitement, and [**03:28**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=208s) concentration on certain stocks on the exchanges and then these shady and strange acquisitions. I recall from the year '99 when Time Warner AOL operator, a landline operator, bought for cash Time Warner, which was over 100 years old, a big Hollywood company and publisher. So, in technology there are these phases, where we tend to, like, overreact and over-imagine the change in the world. We reinvent the steam engine and then sort of under-react and in the long run, it can now be said that all the dot-com boom's prophecies have actually come true. Everything that could be done then. Well, fast-forward from that to 2012, in Finland there was a huge, depressing, sort of current-like mood, apathy. Finland hadn't recovered from the financial crisis and I read then Fast Company and Wired, where they talked about e-commerce and the death of department stores and this kind of structural change. Uber and others were starting and, well, then I looked for from an Oxford professor's presentation material or like from a presentation from lecture material I found online somehow, I think it was Slide Share service, which had slides, and I dug up a word like digitalist. And well, that a digitalist in those [**04:55**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=295s) materials is a person who has a positive attitude towards technology's possibilities and promotes its utilization. And from that was born this kind of Digitalist network, the Digitalists, where they organized various events, things and more. And we had such a counterforce, which was of course familiar from the Apple story, because Apple's developers were underdogs and there's this old Apple vs PC confrontation. So we had like Digitalists versus Analogists, where Timo too as a young rowdy man was involved and well, and back then digitalization was promoted through these kinds of cases and examples and through international speakers and well, there wasn't this kind of prestige bubble, but in a way Finland is quite slow to adopt And I remember when I started Digitalist, in Finland they said that no, you can't sell shoes online because you can't try them on, and the retail chains really said no, we won't do e-commerce because then people will return them and so on. Come here to Pukimo, and quite a lot has happened in ten years, but now it may be that AI is a totally awesome thing, especially for the user, but it could be that we have this kind of bubble here that will burst one day. [**06:12**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=372s) I've always said, or I've said in a few places, that January 2026 will start, like, with beach parties on a volcano island. Hey, but come on, yeah, but a little reflection before I throw to Timo the hopefully optimistic torch, so... In '99, my subordinate and friend Henry Eilert went to iobox and then it succeeded—no, no, he really does live in Finland still, a Swedish guy, a really, really great angel investor and so on, but well, Henri, so then iobox was sold to Telefonica, meaning this to a similar one like Warner, well, to that Spanish phone giant like, for a massive sum and then Henri served his time there, in Madrid for a few years and well, hardly suffered too much, but then with money in his pocket, he was able to return as a Swedish, well, London-based investment banker to improve the Finnish tech scene and has made quite significant investments here, so these criticized tech deals, so maybe [**07:21**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=441s) Telefonica made a bit of a loss here and an error in judgment, but for us Finns it was some kind of blessing anyway, that from there Sonera's Zed and Plaza world conquest, massive investments, billion-dollar deals, well, Tuomioja could indeed have not blocked the sale of Sonera to Telefonica, and then we would have paid off all the national debt with it, but these leftist politicians don't really know how. Well yeah, a side note like this, but now we get to Timo. So well, you've heard that the flood is coming and then a new rise. So what did you think about this? Well, I brought that surfboard then. Right there so that like I don't know where the towels and swimwear are. It doesn't seem to be in the specs, but but I'm bringing it as a surprise, like hey surprise, that well the deluge is coming. So this is a bit like a matter of faith, that that what Ville set in motion back then and thanks for that. Back then we did a guerrilla marketing campaign I brought that that first book there to Sami's desk. By the way, it's a great color, I mean [**08:26**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=506s) digital transformation. You were ahead of your time and indeed and and this is sort of like that now comes the big bad and it wasn't dan but specifically digitalization came back then back in the day. But then if you still look at that one from my childhood, that that wine-red landline phone and yeah I had one like it or I mean sorry my family had the same color, so well the future is calling yeah and and well this is like you just said well Sami, how these young people do they even know how to use this rotary dial and and well, it's not even a joke, it's actually true that if there are now listeners and viewers here, that's how it goes, you have to take the number from there and well, it was the time when I was a child and then when we said hello and the phone rang, you had no idea who was calling, so a Savonian answered, and then there's this question about what happened when modems arrived. I'm from that generation, that I remember when we had a landline, the first PC arrived and then that then from there and then you knew that okay, now some line has to be closed since there was only one landline and so on. But let's fast-forward this image, so to speak, forward and I—I always say here that I'm from the island of Säynätsalo originally, and that's why this board is a really important element overall, so many people know Säykki because it has Alvar Aalto's town hall and and people come from all over Asia to see it, because it's like a huge, huge, fine structure overall. But then how this [**09:52**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=592s) it relates to this whole digitalization, so in my childhood this is a community of 5,000 people, 16 km from the center of Jyväskylä, so there on the island there were three shops, two banks, a post office, a health clinic, and a factory. Logs were brought to the factory along Lake Päijänne, they were floated. Then logs also arrived along the rails as well. Young boys were jumping there on top of the log piles. And now fast forward, 40 years ago I went there with my own children during the autumn break. What was left there? They had just shut the door of the health center and put up a note. That services are now found at the Central Finland Hospital 16 km away and, well, only one shop was left and this town hall. Mm. So this creative destruction here, this is, this won the Nobel Prize this year. So if we look at economic history, creative destruction happens. It's one of these technology waves. Are we going to surf it or do we pick up that landline phone and hope for the best? Yes. So now this Agentix network, well, it's quite massive wave. I mean, I think it's very similar to this digital movement by Ville Tolvanen, which, by the way, as a side note, Ville's idea was the Negotiators community, in the Stockmann cafe, that legendary meeting where Ville basically said roughly, 'steal from this', meaning 'copy this owners community', and I did, I did a good job, and I think everyone should create some kind of platform or [**11:16**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=676s) ecosystem. Actually, I just came across ecosystem thinking in many contexts; it was talked about a lot 10 years ago. It's a bit like it's been forgotten that the web originally connected computers, linking computers between universities, but it's a network of networks and in the sense that you kind of spread out like you do multi-channel via audio, YouTube, Facebook groups, communities, your guests spread the word, so it's like a kind of digital era way of sharing information, of spreading information. Yeah. And Agentics Finland, they have this massive WhatsApp, well, the discussion group Saari, where I'm quite active, and since I am also a coder, I am also active in that Coding Agents group, so well, I've found this vibe coder in myself, who, who and your previous episode is worth listening to, but I listened to it like that. It was a good episode about vibe coding, but at this point I could ask you, because you also talk about these communities that are important [**12:16**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=736s) in this era, and quite big ones too, thousands of people involved, so they have their own role, sort of like we said sharing is caring, meaning sharing is caring as a knowledge sharer, and in that way promoting things within communities, but I'm interested, for the basis of this AI discussion, how you use AI: what does it mean to you? Yeah, well, AI LAN parties, meaning uncles or daddies sitting side by side and sharing via laptop, like, 'I do it this way. What do you do?' This is like my primary way of learning. So you use it for learning and you've now chosen coding. Do you use it for anything else, like investing or Yeah, yeah, well, I have this agentic perspective, which then from there has become Agentix Finland. So, for example, browsers are agentic control tools for me, more than just these Google browsers used to reach some homepage. So they perform this active user interface access to various data sources and then run poor user interfaces. It's like really common. So for me, this, uh, like before with these modems and phones you would call somewhere like Napster or web browsers, so in a way this web browser experience has changed to be agentic for me. So I use it in a way that I give work [**13:45**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=825s) instructions to that browser's AI. As a dialogue, the browser internet was still this kind of mouse-clicking for browsers and it was actually, the mouse is a completely artificial invention, meaning humans don't move a ball-like object on a table anywhere in the jungle or nature and that's why adapting to it was harder, but now you can talk. or it's like interaction, it's dialogue, Timo Suni so I thought that since we don't always agree on everything, then these are just completely wrong questions, that in a way, are we asking like, well, 'do you use AI?' when we should be asking 'why don't you use it?' and well, I know that we are pioneers in this, in a way Ville has already been sure like, like there was already this whole Digitalist movement and what Sami is referring to, this Agentiksi community, basically those who have seen that, hey, the tsunami is coming now, that wave is there on the horizon, it's actually already here, and in a way, like whatever comes to these, like, agents, this is just so true. Our organizations are built for that and our whole, like, knowledge work, that people sit at the office. I don't know what's being discussed in Italy about this, like, office thing, but we have a big hullabaloo here about whether you have to come to the office and how many days you come, when that's not the point at all, but rather what value we produce and how we produce that value. This [**15:01**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=901s) the question is, okay, what do you use AI for? Value creation, competitive advantage, and these are those kinds of terms, whereas we're just talking about getting cool cat videos, and well, we'll get into what Google released later, the new Geminis and such, you can do all sorts of things with them, and anyone can, like this 'vibe coding', and the previous episode is a good example that anyone can make, with a single prompt, an online store for themselves, vibe code it, and like Eleven Labs' valuation increase, I mean it's a one-year-old company, and if it has like 200 million in recurring revenue as a service, it's pretty incredible. They don't have very many people working there. And this flips things completely, like those let's break this down. Read only memory. So ROM memory is limited. So much information is coming in. Let's try to break this down, I think you started off well and we agree on many things, Ville. RAM is the one, ROM is permanent, and EPROM, you can burn an EPROM like ROM, Kalle is now specializing in alcoholic beverages. Better not not anymore. Well, I was probably in London when you're talking about the prom. Well no well not so, er, technology has a tendency to sort of sometimes develop extremely fast and sometimes there are these slow periods, where not much really happens, which then makes for example investing really difficult, because it's hard for you to time the time-to-market. [**16:24**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=984s) someone wise once said that if you invest, the technology risk is worth taking, because the markets, those you can't influence, but that technology risk could supposedly be analyzed, but regardless, I think it's clear that Timo is right, that now we have moved from the ADP or ICT digitalization to the fourth era, which is the AI era. And I said, the reason it's the AI era is because AI dominates all development from here on. You're right about that. And then I'll just ask everyone this, what do you use AI for. Because I want to understand how others perceive it and learn myself. I specifically use it as a learning environment too, meaning for improving my own health, for navigating. I've input there health information, and we're talking all the time about eating and weight management and all sorts of middle-aged man topics. Then I use it a lot for investing. I have a lot of weekly tasks there. For example, even though I follow Novo Nordisk, it makes a weekly report for me for Monday. Yesterday was a bit—yesterday was a bit of a gloomy day. Let's keep that in mind and, well, and then this kind of self-help quite a lot. M it's really strong in that, because all those help tips can be found in the language models and I've asked, for example, that 'hey, I have a 30-year career now, so what should I do? Read all of Ville Tolvanen's public writings, [**17:45**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1065s) appearances, and tell me—give me five options for what I should do next and it's capable of—I check the language model. M and by the way, I did that same test with Gemini and it's—it's better than ChatGPT. It's considerably better. It's like great news for this whole ecosystem, that the Alphabet product is—it's very good, Google. But then if we go into what value the user gets and how we play with it and we are looking for significant changes in productivity or companies that would have done these 'leapfrog' cases, there are very few of them. And it is related to this technology, that there is always that 'teen sex' phase where there is more talk and hype in those communities, but the real changes are, well, modest. Well, and maybe we are not able to, we are not able to publish about these or we can't throw out big, big cases yet. In this, like American baseball, so I was there at the Berkshire Hathaway annual meeting and there [**18:49**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1129s) in the last significant Warren Buffett one, of course the young buck continues there, the 62-year-old CEO there, for this 95-year-old Buffett's career, but Berkshire Hathaway invested in Google and I am also a happy Google or Alphabet shareholder, and well, I was then invited there at Slush to this kind of exclusive Google I signed a five-year NDA, so I have to be a bit careful with what I say, what I can say about this, but well, the five-day session Yeah, you could get a five-year Google lawyer after you, but basically it was just that an hour and a half before the global Gemini 3.0 press release, they demoed it to us. So I got this slightly self-destructive feeling of taking some photos and posting them to social media, so both the police and Google lawyers would have probably hit me over the head, like you broke the embargo, did you see what you signed on this paper, so but you were among the chosen few and it was very exclusive stuff. The tables at BioRex were reserved and I had a Swedish venture capital investor sitting next to me. We had our own champagnes while they demoed Gemini 3.0, apologies for this elitist description, but it was just so cool. And then they demoed it, [**20:13**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1213s) the first one was this 'lovable killer,' an agentic community where it coded a sort of SaaS project right there, and then for us, it was doing this kind of journey and rule of law analysis, then like for the company's figures there and then, well, that sort of thing came up and there were all the graphical interfaces and then it made a kind of 3D game there as well, and we started with this and then we discussed, there were about ten of Google's, you know, those Deep- Mind unit's top researchers and then they explained how this agent coding environment and this Gemini 3.0 LLM model works and beats everyone and came first in the Humanities Last Exam, which is this humanity's last intelligence test, and in other tests too it was on top, and of course while filming this, Anthropic's Claude released Claude 4.5, 5.0, so this is like a constant rise, which is also prone to breaking those moats and justifying that no company is as valuable as Nvidia right now, or Palantir. [**21:31**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1291s) And I think there is a structural flaw here. I agree that luckily we have a surfboard. The wave is coming. Can all three of us fit on it? I'm really itching to go to Sweden or or Tallinn. Well, we are at a very early stage and now if we I I have found three areas where I believe there will be quick breakthroughs. The first is investing, where AI-generated analyses are just massively better and faster than this kind of analog analyst banking process. Because you can ask daily or weekly, you can actually set media monitoring on for your stock. Tell me everything published about it. And the second one is coding, but as discussed in your great episode about vibe coding, it doesn't go into production yet. So it's in the experimental phase, demoing, and it saves time. Well, then here is the third breakthrough: reinventing consulting and evaluation, because now we've been in this browser internet [**22:32**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1352s) digitalization phase in these large teams, with a 100 percent utilization rate and a massive cost base. And then when you add something like Yle Areena or an online store or ...change something that's been working for 20 years, well, that's legacy. So it's sort of like a quarry in a certain way, as things are being updated there and it's extremely complex, and well, in those matters I see that there is a need for changes, but we should recognize if we believe that it's coming, that ICT consultants will face really big layoffs, that the buying behavior actually starts to change. Now ICT consultants have declared themselves as AI companies, meaning they are basically using a steam engine on behalf of the customer. So these are those kinds of problems, those difficult things where you can't help but remember that '99 vibe in a way. Hey, I have to jump in here. Do you guys still remember Y2K? Yes, you do, and the kids won't remember, but let's refresh our memory. This was specifically the time around '98, '99, and then there was the fear that, okay, nuclear power plants will stop, you won't get cash from banks, and this was a technical problem. Sami can explain it to the youngsters. [**23:46**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1426s) Yeah. That's right, so on 1.1.2000, these bits no longer fit into that day counter, so people thought it would roll over to 1100 and then bad things happen when, well, time loops go back 100 years like, in the wrong direction. Exactly. And a massive global project was launched, Y2K, for which everyone prepared; coders of that time and everyone fixed, fixed that field, for which this was a purely technical update. The way of expressing time was changed. But now we are actually in a situation like this, and it relates to these waves and this, like, pace, and when we look at steam engines and all this that has, like, happened, so Y3K—and now we are not looking toward the year 3000, but rather 2030. So, so now we are, like—at the end of '25, it is exactly three years since GPT was released to the public. I mean, previous versions had been developed, but probably none of us used them; it was only when it surfaced on the market, and that was about three years ago, and in a way, when looking at this, like, long perspective, specifically these, like— remember, you are quite right about what, like, has happened and where the breakthroughs are, so in a way, this has come really, really... [**24:57**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1497s) Like, cautiously, but if we look at these previous waves, we're talking about how the first IT wave was from when the landline phone came and the internet broke through, so it's been about 30 years since we got, like, smartphones and the internet and so on. I'm a Master of Science in Engineering by training, sorry for that, but the point is I approach it from the angle that in this 'Competitive Advantage with AI' book, I describe these waves, that the first wave was terribly long and it was actually quite shallow, but now we are at the point where this fourth wave comes in a shorter time, but its impact is much higher, and where this '3 Ys' comes from: individuals, companies, and society. In a way, it being a technical change, the previous one was enough if we just, like, turned the dials a bit. This is a major change in mindset right now. And about this speed, I'll say briefly about Gemini still, it was absolutely amazing what happened there, you remember computer games from back then where you could either drive a rally or fly an airplane and you'd press those arrow keys and, well, then you went there and you could fly. So this was demoed at that Google side event where I was. I was on a shoestring budget. I wasn't at any exclusive one. Retrobit. [**26:05**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1565s) This really was a retro episode. So, the latest Gemini there can do it on the fly, when you press a button to fly down into that cave, it renders it on the fly based on what you do. Let's dig in again. Let's dig in. But maybe before Ville, I want to pick up on those two previous big trends. There's this kind of Nvidia reinvestment loop, which is a bit of an investor smokescreen for investors. And then this other one, the consultants, so maybe quick reflections on both, so since I mentioned Palantir as well, Palantir is in a way this kind of deep data consulting company that will disrupt companies, but in a way that they have all the world's data, including military intelligence. well, deep secrets and and then sort of like PE3000 yeah, and then they come with, well, an overwhelming mass of data, well to embrace your business model for the better and hopefully maybe then maybe then they leave your data alone and don't, like, in Borg style, hijack it from you, and there's a bit of doubt whether it's then this kind of malevolent amoeba that then grabs the information from there after it has consulted companies. Are you analyzing a stock now or is this some [**27:19**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1639s) Marvel thing? Well, a bit of the business model too, but I mean, it's reflected in the price. But fundamentally it's quite a lot like this very deep data, well, a very data-heavy, well, like a really competent consulting firm. At least now the business model has gone in that direction from that sort of hardcore, like, intelligence data and other stuff they have had. Well, then also to monitoring citizens and this Chinese-style... Well, will Gofor ever reach this level? Well no, but it can get in on this, like, sort of data embrace. So in a way you So you're basically that group that understands, like, data and AI, layers on top of it and all these abstractions on top of API calls, like Model Context Protocols and, well, the interfaces created through them, and being able to renew them in this AI digitalist wave to the, customer's added benefit. Well, maybe that. Is there more on this, or shall we move to this Nvidia situation? Well, maybe from there one could say that Palantir is still missing that industry in a way, because I don't believe Palantir serves [**28:31**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1711s) banks like Osuuspankki or European companies, so whether they are governments or whatever they are, but in any case, whatever the future is, that 3000-year valuation multiple is absurd. It's a Marvel movie, really. That's why I think it's good that we put this on record and see. In all my skepticism, I hope I'm wrong, but I'm afraid this is like November '99. Yeah. I don't own Palantir myself. Hey, I could add a little to that and Exactly, like where the value comes from and how it's formed and, and In a way, I'm looking at this positively to get a bit of a gap here. I mean, I strongly agree with you, but the question of whether this is a bubble is the wrong question again, and well, it's true that in the valuations there's definitely this inflated, inflated, inflated and cross-invested stuff and so on. I also have Microsoft socks, as they just turned 50, they were handing these out here and we know they own 95... by the way, you are both quite biased. I am also a Microsoft shareholder. I didn't say that. I just had the socks, as there's this beautiful, happy smiling thing or... [**29:45**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1785s) Anyway, Sami asked to talk a bit about Microsoft. So, the point is that if they own, say, half of OpenAI, then it's completely valid to ask what happens to them. What happens to the process? What happens if the valuations stall? That is one question, but I think the real question is, what to do with this technology. Even if all companies went under just like that and we believe that our data and algorithm sovereignty is met in Finland. We still get those services from across the big pond here to us. We get those algorithms, we get them all on our own servers the same. But wait, wait Ville, wait. We agree, but the point is what we do, how we produce value in Finnish society with this technology that we currently have. And, and to this I want to bring a kind of angle and maybe a swing, that let's take for example, the occupational pension field. We have good private pension companies and we have national ones, and about Finnish pensions one could say a thing or two, but if someone today decided to establish a new pension company, how would it work? We talk about [**30:49**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1849s) these things called AI-native companies and we are here now, I, I, I, as I said to Ville just now, that we have now founded here a new AI-native company. We want to disrupt this consulting business, because it's in a massive transformation and this goes through all industries. But if we use this pension company as an example, if they took all the technology, everything we have available right now, how would they make investments? Would a human do them? Not necessarily. There would be quite a lot of automation. AI exists and these kinds of investment services. Ville said, Ville already uses these, they're all available, money is there. Would there be people processing those pension applications or would it be based on having a group of agents who could already practically do this? And this is where the value creation comes from, like okay, how many people would work in a pension company like this, so it could disrupt the pension sector. Pause. Great. Now we're at the core. We are right at the core. You just said something important. Earlier, when we were talking, you spoke about this rapid adoption. So Users have now adopted, er, the use of language models very quickly. It is true. So the adoption of technology has been accelerating all the time. If this is now the fourth wave, for instance, in this discussion. I agree with that, but one of the slowest things also during these 20 years has been the technol- [**32:12**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=1932s) monetization of technology. I mean, that has been one of the slowest things. We haven't really seen, like, in digitalization the process of giving things up. So you're talking about a native pension company. There probably isn't really anyone working there. Maybe some CEO or their secretary, but we haven't seen any sign that different sectors would quickly and proactively utilize technology. I'll give a few examples now about AI. We talked a bit about consultants, we talked a bit about banking investments, but if we talked about how customers should buy ICT differently than consulting, but then if we could now be sitting in a Mehiläinen clinic during a middle-aged man's check-up and AI could listen to everything we say and after that talk with a doctor about what is bothering Ville and what might bother him next. And we don't, we don't see this. I haven't at least heard that doctors at Mehiläinen or Terveystalo would use, in a certain way, an AI assistant there and change the treatment process in a certain way, because then, of course, the consumer could also continue with the AI with. Doctors don't trust it, they don't want it. [**33:30**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2010s) Education, are our teachers currently responding positively and encouragingly to using AI in learning? I bet we believe that the old ways of learning are still the right ones. And this has actually been, even though I am a digitalist and thanks, Timo, for saying so beautifully that I've been a pioneer, this has been the hardest thing for me for 25 years, how extremely slowly technology eventually arrives in everyday life. We held electric car seminars 10 years ago, where we looked at Oslo's curves and thought this would happen very quickly, and the share of electric cars in new registrations is quite high, it is it's been a hugely slow process, 10 years. Yes. And this is where our problems lie in a way, when we think about technology and maturity and especially those valuation levels. I agree that if investors lose a bit of money, then it isn't necessarily a societal problem as such. Of course, there may be other jobs lost along with it, but we should understand that the fact that something is possible in technology doesn't mean at all that it will be adopted quickly. This is exactly how it is. [**34:44**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2084s) I have a hundred things to say about AI service design, but let's take this loop about Nvidia. So in a way, it is now the kind of primary source here. It's the hardware, those Nvidia graphics chips that power this AI computing capacity almost everywhere, and Nvidia's market value is rising to 4,000 billion. It's not enough, and in a way, well, and then they have discovered about themselves that they are also the cash register, and then when that money comes in, well, then it has been invested in ecosystem partners, and good old Nokia—I also own just a tiny bit of it—got a small bounce from it, when Nvidia magnanimously invested a billion, completely selflessly, in Nokia. And was it so that a small announcement came later that Nvidia invested a billion in Nokia, and was it so that Nokia is investing four billion in Nvidia, and well, these have all been these kinds of strange trade-offs. [**35:54**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2154s) Unfortunately, I don't own Nvidia. In investing, perhaps the most important basic thing is that you shouldn't invest in something you don't understand. Mm. And if you don't understand it, then maybe you should keep your money out. But now we are in a situation where this 'Magnificent Seven' or these most valuable American companies make up about 30-35% of the American index, and the American index is 65% of the whole world. So we are also close in that sense to such a bubble, questioning if these can be this valuable. Well, now it's being explained over there on Twitter, where the bank chief analysts are. They're all shouting at the top of their lungs that there is no bubble, because, well, these are generating massive cash flow, as you described. Except for Palantir—and well, that was started from a bit of a weird angle—then these hyperscalers, meaning Meta, Alphabet, and others who are forced to buy these chips, depreciate them quite slowly. So now there's this phenomenon where these companies have started... it's even been said that these circuits are relevant for about two years. Development is so fast, the wave is almost [**37:10**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2230s) like a storm wave, yet they depreciate them over five years. Mm. And then it's claimed that this is a kind of sleight of hand, which has made these hyperscalers appear extremely profitable and kind of hidden the scale of the massive AI investment, because it was said somewhere that this total amount of depreciation is greater than their annual profit. Mm. Which would mean that these, these well, the next most expensive companies after Nvidia would make zero profit. Mm. And there is one problem here related to the monetization of this big technological arc, which I see as quite clear, and I'm really worried about the fact that actually the users are benefiting now. Blockchain was very similar. The user benefits immensely. In Finland, real estate transactions are done on the blockchain. A great system. It works quite fantastically, but that wave didn't result in productiv- profitability, productivity, or really new business models. A lot of better IT came out of it, and it might be that AI is in the first phase like a revolution for users. Yeah. And, and I'm refining this idea, I mean, I like this, and in a way, we're talking about those valuations and [**38:26**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2306s) hopefully none of us are giving any kind of investment tips or, or hints here. Isn't this all—isn't this like, let's update this saying, that if a Savonian or AI speaks, then... it's the listener's responsibility, so if someone now takes a tip from this to heart and starts investing, well, hopefully you won't call that landline to ask what you should invest in. But one thing, I would invest in, one thing I would invest in, would be mastering this technology yourself. So I'll grab onto what you, Ville, said, and what you, Sami, use this for, and what the listeners and viewers use it for, because you can't understand this; I could come with that book and hit Ville and Sami and everyone's temples, and then you hear 'ouch' and things just continue in the exact same way as before. Oh, the headache is already such that nothing changes. That was, in my opinion, 'unless someone gets angry.' Ville, this was thanks to your slogans from there. Tere Sammallahti has stolen your slogans. I don't know who... I think these are memes going around for you middle-aged men. But the point is that if you don't actually start using those agents yourself, there's no way you can really get this change, and then the discussion about whether this is now valuable, does this change the value, so basically where one could invest, because we all, whether we are Warren Buffetts or just ordinary Savonians, we only have 24 hours a day, and [**39:54**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2394s) if I could give one investment tip, invest in using it every day, whether for fifteen minutes or an hour, depending on how much you believe in it, and to get back to this, I use it solely and exclusively, AI first, and only then if my own skills aren't enough, and this is the kind of mindset, I must also give credit for this to Risto Siilasmaa, props to him, he was speaking at the entrepreneur days and was asked for a tip: use AI, ask it anything, and this is like what's also in the Agentiksi community, where in a way they openly share what is already possible, and then of course there is probably—I don't know if there's a channel like investing and such—there's a lot, we should set it up. as that sort of father-sun figure, he can also do it for the negotiator channel. There is technology, there is this education you mentioned. I wrote this AI 2030 white paper for it in the spring because we have like absolutely everything, we have social and health data. We have large wellbeing services counties, which is incredibly cool. [**40:57**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2457s) That's like really good. I like that, and we're like very much in agreement that now, if AI should be used for something, then create that Super Me version of yourself, a learning version. So, no matter if you're into cats or hunting or whatever, you start using AI to spar with your own hobby. It's—it's like a really, really good idea, and actually, I'm wondering if we should actually create this kind of Super Me community that would just sort of share, you know, like those playing cards, the Happy Families cards, so, what I do with AI, because we should spread it to teachers and lawyers and engineers and, well, doctors, and get people to learn, because then they could something is starting to really happen. I like that a lot. Really much. Yeah. And then just straight up go for it. I have Oliver Palmuni here on an internship, just as you arrived, you interrupted my brainstorming session, I was just saying, let's code an app for me, where I can do this task management. So in a way I have this multi-location challenge, where I have, well, investment stuff at Translate Corporate Finance, board work at Frederman Capital, no sorry, Korona Invest's Ferman Group, then [**42:24**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2544s) I have this Negotiator Empire and then still my own investments and family stuff and other things to manage, so this multi-location thing doesn't exist yet as an app from any of these organizations, so I thought that I'll just make it myself, yeah and in this, well, it's like one of those things where you can make your own app for these very unique use cases and leverage the data you have and put it in there. But I think that kind of doing is a bit of pseudo-doing, but it is the right direction, because you're learning what works for you and what works well with AI and and then you learn about AI, so you understand the possibilities. But then when we say that now you can code or now you can like create images or shows or presentations and such, I've been like thinking about productivity and the nature of work, since most of the work, if we forget for now, like ditch digging and physical labor and health work, which is all like human work and important, but quite a lot of knowledge work has become this like project management, or participating in projects, so companies don't really produce massive amounts of [**43:40**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2620s) shows or animations or images or presentations. I know many large firms that make one presentation a year and that's like enough, that it's the annual plan. So, like, the capabilities or what we seem to be possible now, it doesn't solve this productivity question and doesn't really create anything new. But I'll say this at the end, as I'm so passionate. This is good when it gets intense. Sami has like the right approach and I have a bit of a similar one, like four horizons, where I utilize AI, from which I learn, and from that I got this Super Me idea, in a way, I actually create like a smarter version of myself or an interface to the world, maybe better. And an interface to the world. Yeah, you can use many terms, but the goal is then when AI evolves and the tools platform evolves, to maybe compare those environments to each other and through that create something new. Maybe you can code at night and and this is like the creation of new value. Now I maybe also there tossed out at the start and earlier, whether we compare Finland and Italy a bit like the Fatherland and well Ville has already lived there longer and commuted here, maybe I don't know what kind of. [**44:52**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2692s) I'll say this quickly. I did 600 keynotes about digitalization. I always I said that if this thing ends up on the pier, I just had to leave. Great. But it's sort of like, what are we worried about in Finland now? Where do we get that renewal and that drive and sort of putting on those rose-colored glasses, like, hey, we can bring value here. So now the challenge has been, and I totally agree, this AI transformation must be led. This cannot be outsourced, as we operate as an AI-native consulting firm and we intend to disrupt the consulting market, and the point is that there are five of us and well, I'm aiming for 3x. Meaning a threefold impact equivalent to 15 people, and well, how do we do it? We take that AI utilization as far as we can. We have Pera, for example, and instead of him just being this passive listener who transcribes our meetings. We have our operations so well-organized that we don't discuss whether we have an office, so one of our members is in Central Finland, one in Eastern Finland and three people in the capital region. So we don't really need an office. This is also the era we're living in. But hey, but these are also good starting points. I have to praise this, because now if we want to change the establishment, we need these kind of revolutionaries and we need companies that say that we are like anti-Ovio or anti-this and we do this differently, because it also develops those companies. They have to think about things and above all [**46:20**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2780s) customers listen. Yeah. Yeah, and we need people for this. We need to hold their hands, like thank you Ville and Sami, for carrying this message, so really from the management team level, holding hands and taking it from here, like, can Sami and I send our CVs to you? Yeah, yeah, yeah, surfboards, we'll take all of you and if nothing else, let's go to Tripla to try that indoor wave. I don't know if you've tried it, but the point is that for example at our place, we've been doing this for three years, transcribing meetings. I don't know if some AI or something is listening to us here right now. Everyone transcribes meetings and that is, after all, what, does it provide value? No. right, it is passive use of a tool, when this mindset should be flipped around, so that we were introduced to Pera in our management team a couple of weeks ago. Yeah. It is like active. at some point when this became a kind of IT history lecture series, then well, at some point there was talk about a digital leap. You remember, everyone remembers it, it was a bit after the information superhighway came the digital leap, so it annoyed me immensely, what I heard in this story, that it doesn't unfortunately development doesn't leap; instead, development is like very linear. [**47:28**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2848s) But in that sense you are absolutely right and you are talking total sense, I will even read that book too, actually you have sold it well, so that digital or now AI development cannot happen if the individuals do not develop. And I still want to get to To echo Sami, the problem here is that this AI transformation is a different kind of beast than the previous digital transformations and all the big D's. So, we've approached it through individual-level tools first. Copilot, ChatGPT, cat pictures and videos and so on, when we should specifically buy into it on a process and organizational level, regarding how we generate value. And this flip is actually very slow. There's a lot here if we talk about the agent community that Sami has mentioned many times, where they are considering how these processes can be handled; they talk about agent farms and such, as technology evolves constantly, we need to create boundary conditions and so on, but in a way, regarding what you said about knowledge work, Yle did a good story just a moment ago on how Finnish society has evolved: [**48:41**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2921s) first we were in the fields, then in industry, then we did services and the majority of working hours are currently spent specifically on knowledge work, on us typing on that keyboard with two or eight or, some even reportedly with ten fingers, and then moving that mouse, like, eight hours a day, and this is supposed to create value, while the point is that when agents can eventually do this same knowledge work that we do, and now we should be able to imagine the future we're heading towards. The CEO of Box, a large American company. With quite high regard currently, the CEO of Box said a couple of weeks ago - it's on X if someone wants to check it. If we look ahead about five years, 95% of the work agents do is something that humans have never done before. And this is, I think, essential: what humans have never done before. So if we're there, returning to the history lecture, in '99 we imagined what SEO is. Now we talk about what GEO is. Back then we talked about what an internet browser is. Now we're talking about agents doing stuff there shopping. Um, OpenAI just released that inside ChatGPT there are now its own stores, and all these credit cards from Visa and Mastercard are available for agents. [**49:55**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=2995s) Well, we couldn't have imagined back then where this was going. And this is the biggest obstacle right now. People can't imagine it. We keep turning that crank because it's fun to do. But those who want to go to the very end of the pier, they've really decided that, hey, this is a bit like a matter of faith. Do I believe in the coming wave or do I just go where everyone else does? Yeah. And then in a way, we 'AI-ize' these digitization-era models. I've also been preaching for a long time since I came up with this '3D' model as a young man: delete, delegate, do. So in a way, as a young man, you are the one on the receiving end of management's right of direction, so your ability to say no and not accept delegation and to work like crazy, well, the power balance is a bit skewed, but as the 'uncle-leader' of your own life, that pyramid flips upside down, so you have a tremendous opportunity delete things that aren't significant. Just say no, start with no also good negotiation wisdom. Then, like, a lot of power to delegate and and then, well, top-tier stuff gets refined there, and you do it with your own hands, so in a way, how this sort of already digitized model as a supervisor and between people is translated into the AI era, because all of those can be agentized, so a sort of delete filter, [**51:17**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3077s) delegate filter, and do filter, so in a way there are incredible efficiencies in all of them, and I'm thinking quite a lot about those kind of digital leaps in that sense. A digital leap, meaning the way digitization happened, when we moved from analog activity to, say, direction rights between people to where bits are being moved now, then in a way how you move that interaction and direction over to AI agent farms and such, and I'm putting a lot of effort into this myself, so I can develop these further. I have to say that both of you are talking very smart and well. A one-man investment bank would be quite modern. I mean, a coding-capable banker with a Rolodex, then everything an investment bank does agents would do for you, wouldn't they, in practice. So that too becomes a bit of a this kind of 'super-me', meaning user empowerment. Great stuff. Well then, what Timo said just now and you mentioned what we should start doing and how to behave, in my opinion, to move this evolution forward in a certain way, because we have such strong counterforces in the familiar world and way of being. And for many people, like digital devices are already too burdensome. So we don't have, we don't have that hunger. We don't have the hunger, that drive to [**52:42**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3162s) version the world better, and especially not Finland right now. Maybe it's this community we're involved in, or some other community. That this kind of time really needs some kind of escapist simulated reality, where some nerds go to a garage or some lab or digital event and start to believe that, that The world is different, and then, little by little, it will emerge from there. Yeah. And there are these kinds of things, and so the Digitalist agentic community is, or... the negotiation community is, and I want to highlight one example and... in a way, if we look at the Finnish public sector, for instance—I don't know... what kind of debates are being had in Italy about how good the social and health services are, how good the education is, or what kinds of services municipalities and cities provide, but the thing is, we have a massive public sector in Finland right now, and good services have been created there, and society has been built upon that. But if we start from the mindset—for example, we were there with Virpi Hotti and the State Treasury, just to give a shout-out—we were there thinking about this together, what it [**53:42**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3222s) means to have this kind of agentic state, and specifically, where is that hunger and this sense of purpose. If we take a Finnish hockey player who finds that, hey, the SM-liiga is a big deal, but those who want to go to the NHL are the ones who succeed. And now we should take like this mindset and and if we think about this, then from there you can find those partners who are already planning how we can like play in these. And in a way, our vision is that Finland is really raised to the crest of this AI wave and we take the most out of this and it really isn't about the money. The biggest obstacle right now is between our ears. And and now then let's take this public sector and and I want to bring an example from out there in the world and and here we find something called the Agentic State Report and and this vision paper is worth looking into. Luukas Ilves has, among others, been involved in it and he told about what Ukraine has done in the middle of a war. They have to invest, they have to innovate. They don't have... as we know, we are now at the point whether there will be like a forced peace, but unfortunately war has always been a time for innovation, so in a way we in Finland have been really good at rising from crises; Ukraine published the world's first agentic service this autumn in the middle of this space and, in a way, now this is like, I want to bring this up as an example, we have examples in Estonia, this autumn, uh, students and certain groups of teachers were given [**55:08**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3308s) AI tools, while in Finland we did a mobile phone ban. As a father, it's an incredibly good thing, but it's a completely different mindset, whether you implement a phone ban or something like this, and there are these kinds of ecosystems and thinkers who want to build this Tesi, and well, regarding AI investments. I'd like to bring this up, meaning it was said earlier that this money isn't necessarily needed, but this Nesta, meaning Nokia and Tesi, Petter Sarlin's Nesta took 100 million and Petter Sarlin made a good profit as a founder of Silo AI, and that's good, and he has also been in this 'lavaboli' investing, so he is in a way, I think, genuinely close to the Finnish tech scene and in a good way—I mean, the word 'Tech' is often reserved for types like Bezos and Elon Musk, but who have then there might be like tens or hundreds of billions. Okay, like Peter Sharlin operates in a world of tens of millions of euros, but in Finland, a wallet like that has an insane amount of value if it is focused on the crest of this AI surfing wave. And I would hope that they specifically find this kind of new digital wave, meaning that the AI enthusiasm would catch on specifically with people who are, say, even single-digit millionaires. In a way, you [**56:30**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3390s) get incredible power right now and start leading this field and raising them, and capital really does matter, because because if you play around with tens of thousands, setting up companies, it's a completely different thing if you start with a hundred million. Yes. Hey, I'll say this in between, I'm listening to both of you excitedly and and we often, I myself have also spoken in the same way, we should start and we must find something like this and we must, we like push the collective. Well, now 25 years later, when I've been in these previous technology sectors and occasionally taken on a bit of seawater and, well, then waited for the time to be right, so I would say that now, if I could make a wish, or what should be done now is that we should do this in a thoroughly individual-centric way. So I would actually start highlighting people who are excited about doing this. We have an AI banker, we have an AI consultant, we have an AI board member, and we would start sharing, a bit like the HS Black Box, the stories of these people, what an AI fisherman is like, what an AI teacher is like and so on, because [**57:42**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3462s) now AI is at its most mature in the user's daily life, while everywhere else it's a bit of that 'teen sex' or difficult pilots, or not yet in production use, but it can already change the consumer world and that is why I'll repeat for the fourth time this 'super me' idea in a certain way. So everyone should create that kind of AI-seasoned and developed learning individual. And through that, I think that when we get plumbers and architects and pastors and various professions to show their own glow, their AI-driven creative environment. Sami speaks of empires, so I think that through that, other people would learn, learn, like, learn faster and better, and through that, the resistance perhaps in society as a whole or among professional groups would decrease. Yes, and I'll definitely grab onto this. Hey, Ville, could you invite us all to visit you, because here in Finland it's quite gray right now, meaning the weather and this kind of cloud of thought is like the color of gray long drink at the moment. I believe that in Piedmont it's a bit more colorful right now, even though it's November, but the IT there is so advanced that you don't need to travel or ruin nature, but we can meet in a meeting service. Well, like in Meet or whatever it is. But then I'd also like [**59:04**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3544s) the next generation to be included. We have a work experience student with us, and then as a father, thinking about what world we, what kind of world we are building for them, and in a way when we talk about, like, that mindset shift and where to, like, invest, we need these bankers who put that, that like, money on the table. But then again, if you look at these, like, latest AI tools specifically, this is maybe the biggest shift compared to before, that GPT costs 20 euros a month, so if you have that much to, like, invest in it and start learning it, if you are a student, we know that these big, like, firms offer, for example, annual licenses for free. So if there's, like, a young person listening, I'll say this much at the end: start building, start building, just start making that next thing and well, then we need to be supported a bit there it costs the same, that Chat GPT—not the absolute most expensive, but the Plus version—as Netflix. I think it's almost down to the cent. [**60:04**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3604s) Which one, but think about which one is more valuable, but which one is more desirable. Maybe that's this podcast's Christmas gift tip, to buy your child that annual AI license or a premium license for double the price of Lavab, and then straight to coding. But that... okay, coding is there, but I mean it's kind of like a WordPress killer and of course I understand that, and also cloud and others, I know I have a friend who is a restaurateur. He does all the ads automatically now. So small or those unfortunately job-providing layers, like website design, AD work, photography, and everything have already disappeared or are disappearing in an instant, because of course pizzeria owners tell each other how to do it and the learning happens, but they are not surfing with them yet. No, no, it's like an income transfer in a certain way. What about ethics? Do we want to go there at all or is it settled? I was AI-style, and then we have challenges like the youth's mental health. I can say a few words about the ethics. [**61:09**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3669s) It [ __ ] me... sorry, am I allowed to swear? It already slipped out. You can delete it. It... [ __ ] it, people talk about things like 'human AI' and all that kind of nonsense, because it's nothing but people using it. It is always human. It's just total nonsense. And okay, I understand that it stems from the fact that for five years people talked about whether AI would take over and make wrong decisions and so on, but I mean, AI does what we want. It's in human hands; it's always human. And yeah, this point where we're at—these language models are coming out constantly, but we are still at the point where they don't learn on their own yet. Sorry, but we are on the threshold of seeing if we'll soon find a self- developing one—like if you remember the movie The Matrix, where they put a cable into the back of the head and then martial arts and everything went to the human user and they were able to develop. But no I mean, these language models are still, as you said, these are human- developed, and this can be challenged and disagreed with, whether there is any [**62:09**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3729s) thinking and consciousness and whatever else. But the idea of whether we are making ethically correct choices now; there are these videos on TikTok circulating for youth, and maybe I don't know if you are there too, but but especially like, how much water something uses. Well, should we not use mobile phones then, or should we travel; what are those, what are our ethical choices, if we have youth who are doing poorly in Finland, and this Sekaisin-chat for example, as a good example, I said a bit harshly about that, which is that if we start with an ethical question, we will never even try it, because we'll get stuck we'll get stuck dwelling on something that takes the attention away from it actually, because to everyone, my advice in a way for listeners of this episode, it's important for everyone to expose yourself. You have to expose yourself and, in a way, demand that renewal from yourself in such a way that you try things out, learn things that interest you in that environment. That's when interaction begins, as well as development and learning, and Sami is of course a brilliant example now as a podcast host, of how you bring wype-coding here and do it yourself and mix in a certain way the careers of bankers, media moguls and, well, wype-coders or experts, and that's what's needed. Alright, thank you Timo Savalainen and Ville Tolvanen. And now that we have this retro [**63:34**](https://www.youtube.com/watch?v=IrWui2k4j8I&t=3814s) gear, I feel like making funny modem sounds and talking about, well, Commodore 64 and other stuff where it all started, but maybe instead we'll go over to the inner circle side and start some completely irresponsible speculation. As in, where is this AI ultimately going? So let's toss that surfboard thinking and go straight with a moon rocket into the future, or maybe, maybe use some warp drive and go wherever this is heading. Maybe we are in Elon Musk's simulation. Who knows. Uh, in the inner circle, we see and speculate very heavily, and since you have watched or listened this far, please subscribe to the channel and you can also give a thumbs up. Thank you and gentlemen. Thank you. Thank you. M. --- Cite as: Sami Miettinen, Neuvottelija — Is AI in a Bubble? | Ville Tolvanen, Timo Savolainen, https://www.neuvottelija.com/podcast/episodes/650-tekoaly-kuplassa-tolvanen-savolainen-neuvottelija-364/, 2025-12-08. For quotes include episode 364 and timestamp.