Sami Heinonen

The through-line

Sami Heinonen's consistent argument is that the metaverse is best understood as the internet of the future, and that this framing does more useful work than any of the tighter definitions on offer. He concedes the phrase "sounds a little bit lame" [1], but the point is deliberate: nobody can say definitively what the internet is either, so nobody should expect one fixed answer for the metaverse, and the same fragmentation the internet already shows, isolated regions shaped by culture or restrictions on speech, will reappear as organisations try to control the new layer [1]. Around that core sits a working method rather than a manifesto. He has lived through a full hype cycle already, and describes visiting customers roughly five years ago who told him they had tried VR and it was now sitting in the meeting room, which was the sum total of their "reality" [1]. His conclusion is not disappointment but relief: "now I can say that the once the hype is more or less gone and people are more or less talking about jet GPD and stuff like that now the real world actually begins" [1].

What follows from that is a bias toward the unglamorous end of the field. He steers attention away from consumer spectacle and toward business problems, away from token economics and toward whatever payment method the customer already finds easy, and away from standalone XR toward AI as the thing that makes any of it work. He continues to talk in this register about augmented reality and the state of the metaverse, and about concrete builds such as a digital twin project in which a virtual copy of the city of Gothenburg was constructed [5][6].

On meaningful use cases versus gimmicks

The recurring test he applies is whether something still creates value after the novelty passes. Every time a new technology arrives, he says, everybody wants to show off, and the result is marketing gimmicks that let you do this and click like that, and then "after a week or two or maybe half a year it's not actually creating value" [1]. The job of anyone designing in this space is therefore to find the genuinely meaningful challenges and areas worth enhancing, alongside customers and their end users [1]. For organisations that do not know where to start, he suggests a facilitated format, a kind of metaverse boot camp, to surface opportunities and build understanding at the same time [1].

He also positions the conventional service design toolkit as necessary but not sufficient. The Double Diamond gets you through the honest first question of whether you are solving a real problem, then divergence, then focus on one or two ideas, then a product to test [1]. His reservation is about what comes out the other end: the usual result is a more or less static solution, a fixed answer of the "this is how we solved it" kind, whether it is a digital service or a retail floor plan with the cashier here and the products there [1]. AI changes that because it sits underneath the whole thing continuously rather than being bolted on as an additional layer [1].

On why B2B is where the value is

For consumers, he sees the metaverse showing up mainly as entertainment, as a marketing surface, and as a virtual escape from reality [1]. The commercially interesting territory is business to business. His examples are all about labour and risk: if you cannot recruit people, you can bring them into a virtualised version of your manufacturing line so they work as part of the physical line; you can raise work satisfaction and employee experience; and you can improve safety through training for environments such as construction sites [1]. Co-designing forthcoming products with consumers belongs in the same category [1].

His favourite example is the value chain. When mineral, aluminium, battery and component prices went through the roof, the pressure travelled down the whole chain, from manufacturer to importing agency to retailer to after-sales organisation, each taking its own margin, with the consumer at the end [1]. Once production costs rise far enough there is "not enough a cake to be shared with any everyone" [1]. That is what pushes manufacturers toward direct-to-consumer models, and it raises real questions about how to set up services and maintenance, but one clear need is a virtual showroom in which customers can inspect the product and be looked after [1]. He calls this a million dollar opportunity for the manufacturing industry, a route to better product quality, a better whole stack, better customer service and potentially lower prices [1]. He traces the pattern back to the pandemic, when there were no physical venues to see or test products and virtual showrooms and car configurators arrived hard [1].

On AI as the substrate, and the human interface

He is emphatic about how his own company should be read: Younite AI is not an XR or metaverse company, it is an AI company that builds tools for the age of the metaverse, in the same way that nobody calls themselves an internet company and equates that with the internet itself [1]. The company was formed by merging three businesses spanning AI, XR and fintech, starting effectively from scratch in January and already numbering more than 80 people, with headquarters in the States and a base in Helsinki [1]. His description of how they work is candid: they get into genuinely new and innovative things, tinker with the latest technology, sometimes over-promise, and then find themselves working out how to deliver before the premiere, with success flowing out of roughly 90 percent of those projects [1].

On interfaces, he starts from information overload. He tried at one point to follow everything tagged as metaverse or AI and found it was killing him, which is his argument for AI-driven filtering and recommendation that tells you what is actually relevant to you [1]. From there he makes his most speculative claim: the most natural counterpart for a human being to interact with data is another human being [1]. So when he is using a mixed reality device he wants his assistant standing next to him, invisible to everyone else, context aware about what is happening in the room and proactive enough to suggest that he should go somewhere else and do something different [1]. He calls that humanlike agent the ultimate human data API, or viewport [1].

On web3, ownership and platform gatekeepers

He draws a hard line here: "met is not web 3" [1]. Web3 is a stack of technologies and enablers, and he likes them, particularly the browser-level prospect of owning your own data rather than handing it to Meta or anyone else, along with smart contracts [1]. He does not claim expertise in it, and he notes that as the ChatGPT bandwagon has filled up, fewer people are talking about web3 at all, which is precisely when the field needs "more doers than talkers" tinkering their way to real use cases [1]. His practical advice is to keep it out of the critical path: when building a metaverse strategy, an MVP or a proof of concept, you do not need any of it yet, and you can introduce pieces when a genuine need appears. If you need a payment method it does not have to be crypto or anything fancy; PayPal is fine if it meets the consumer's need, is intuitive and easy, and leaves both you and the customer happy [1].

The decentralisation problem he expects to persist is the device layer. Whatever apparent freedom of navigation you have, whether through mixed reality glasses, a tablet or an iPhone, you enter through the manufacturer's yard and their ecosystem [1]. Friends, payment options and digital ownership all sit inside it, and those companies have no intention of letting you go free; they still want their 30 percent of every transaction [1]. Free alternatives will exist, but people tend to gravitate to the mega brands and buy there [1].

On artificial scarcity and NFTs worth having

Artificial scarcity is a term he picked up relatively recently, and he treats it as a value-creation mechanism that runs on fear of missing out [1]. Applied to Bored Ape Yacht Club, where a digital monkey carried a value of around 3.4 million dollars, his verdict is that it makes sense if you are speculating on the value of such things, and that otherwise, if you are not jumping on the bandwagon, "you're actually missing nothing" [1]. What he does value is a token tied to a physical product as a digital twin. His example is a vehicle: before the car comes off the manufacturing line, before it is born, the customer already receives its digital identity and can start learning about it in advance of a delivery twelve months out, and when the car is sold on, the token travels with it [1]. That, for him, is a meaningful use case rather than a speculative one.

On access, ageing and the case for virtual worlds

He uses Ready Player One as a familiar reference, where the world has gone to hell and the only safe haven is a virtual one, and then deliberately inverts the pessimism [1]. If you are disabled in some way, if you want to protect the environment by limiting your travel, or if you simply do not have the time, a virtual world becomes a way to experience life and beyond [1]. He points back to VR work already done with older people, bringing residents of care settings into shared worlds where they can move freely and play multiplayer games with friends, with the effect that "my back is good but I'm all all world all" [1]. He extends the argument to himself without irony, saying he will be doing it in bed at eighty or ninety, and acknowledging that it sounds a little depressing while insisting he will most definitely be there [1].

Takeaways

• Treat the metaverse as the internet of the future rather than chasing a single definition, and expect the same fragmentation and attempts at control that already exist online [1].
• The end of the hype cycle is when the useful work starts; the VR era left headsets sitting unused in customers' meeting rooms, which is what happens when adoption is driven by novelty [1].
• Consumer applications skew to entertainment, marketing and escape; the durable value sits in B2B, including virtualised manufacturing lines to offset recruitment shortfalls, employee experience, safety training and co-design [1].
• Rising component and raw material costs squeeze every margin in the value chain, which is why virtual showrooms and direct-to-consumer models are a million dollar opportunity for manufacturers [1].
• The metaverse is not web3; keep blockchain and crypto out of your MVP and use PayPal or whatever else is intuitive until a real need appears [1].
• Speculative NFTs like a 3.4 million dollar Bored Ape are safe to ignore; tokens tied to a physical product as a digital twin, transferring with the car when it is sold, are the version worth building [1].
• The most natural interface to data is a humanlike, context-aware agent standing beside you in mixed reality, proactively telling you what matters [1].
• Younite AI is an AI company building tools for the age of the metaverse, not an XR company, and its work includes 3D worlds and a digital twin of the city of Gothenburg [1][5][6].