Artificial intelligence, capital and the decisions that shape our world. Research and conversations from Sami Miettinen, Head of AI at Translink Corporate Finance.
Etla's chief executive Aki Kangasharju landed in a media storm when his view that Finland's debt brake is too strict was read as support for the line that nothing needs to be cut. In this episode he explains what he actually meant: Finland has drifted into an economic-policy impossibility triangle in which growth reforms cannot be made if they cost the public purse a single cent in the short run. EVA's tax specialist Emmiliina Kujanpää and Sami Miettinen ask what taxpayers get for their money, and Kangasharju concedes that the efficiency of Finland's public sector has barely been researched for some twenty years. The conversation covers the corporate tax cut, wellbeing services counties buying in-house from their own companies, falling learning outcomes against rising spending, the weight of pensions in public expenditure, inheritance tax and the Tax Administration's appeals unit. It closes on reduced VAT rates and the goal of Etla's anniversary book: to bring Finland back to Swedish living standards with more than 50 reforms.
Aslak de Silva is a change leader from Jyväskylä who gets called in when an organisation is in crisis or in flux: CEO of Nordic Business Forum through the pandemic years, CEO of Selfly Store, and now at Maventa. Behind him are more than 15 years of the Korean martial art Han Moo Do, about a hundred full-contact fights and a world title in 2000, from which he has built a leadership philosophy called The Black Belt in Leadership. Sami Miettinen and de Silva discuss how resilience, fitness, rest and staged goals carry over into leadership, why nobody reaches black belt alone, the cross-pressure of owners and boards, Paul Polman's long-term value creation and the loyalty dilemma of the Merita–Nordbanken merger. The episode closes on AI as a leadership coach, the power of stupid questions in the age of agents, and calibrating confidence at the decisive moment.
Ville Tolvanen argues that the Finnish board is a relic of the analogue age and that AI finally makes it possible to reinvent it as a process for leading the future: the owner creates the opportunities, the board creates tomorrow, and management delivers the results. Aura Boards combines an ideology, a thousand-day strategy and annual plan, and an AI situation room built from three sets of financial statements, the strategy and the budget, on the principle that the machine calculates and the human decides. Juho Jokinen and Sami Miettinen test it on three questions: what changes in the owner-board-management chain, how it works day to day, and what evidence exists. The target is 80 percent of board time on the future.
Elias Erämaja, chief economist of Suomen Ekonomit, and Sami Miettinen on why Finland's economy is in surprisingly good shape despite a halved Q2 flash estimate, the cut in the top marginal tax rate from 59 to 52 percent and the cost of a further step to 50, and the Finland–Sweden wealth gap read by mean and by median. The AI section covers business graduates as natural AI users, whether productivity is zero-sum, public-sector productivity and triage, and why an AI tax has no workable definition. Erämaja speaks for an interest group that campaigns for lower marginal taxes.
Anssi Nurminen and Lasse Mikkonen compare their own AI workers with the host's and ask what agentic work actually requires. Nurminen's agent Anneli turns WhatsApp photos, video or plain text into conceptualised videos and updates its skills from feedback; Mikkonen's Ossi and Dude reconcile receipts, keep a case memory and read email. The episode defines MCP, skills and the harness, sets out two memory architectures, and keeps three disagreements open: own hardware or cloud, whether security matters for hobby use, and whether the learning jump can be made in one go later.
Petri Lehmuskoski, founder of Gorilla Capital, explains why product market fit is an unusable term: it has never been defined, so nobody can measure where they are. His book More Scars Than Trophies replaces it with four measurable proof stacks. Buyer proof requires somebody to commit money, resources or significant time, and it is deliberately buyer rather than user proof, because a user makes no economic decision. Value proof is the original move: value counts only after the customer's own cost has been subtracted — training, switching systems, or at worst the reputational cost to whoever signs. That is what explains large system failures, and it is what tells you whether you can raise the price. Running through the conversation is hull speed: a displacement hull has a mathematical top speed and more engines only burn fuel, so pouring funding into a company whose proofs are not yet open is not merely wasteful but dangerous. A company's hull speed is the product of the entrepreneurs in it. Lehmuskoski argues revenue is a poor measure of value because a good salesperson generates revenue without repeatability, distinguishes growth (inputs and outputs rising together) from scaling (inputs falling while outputs rise), and insists the exit be built from the day the company is founded, because every funding round closes doors. On AI he makes three points: evidence decays faster, bad evidence can be dressed up beautifully, and the ability to build software is no longer a gate. What must not be outsourced is the search for buyer and value proof, because it starts from the questions you did not know to ask — and a salesperson only asks what they know to ask. The episode closes on the camel and the unicorn: the camel is a way of travelling, the unicorn an outcome, and Gorilla's third fund spreads over 70 investment decisions with realistic valuation as the key measure.
