Solertio — Writing https://solertio.pl/blog Long-form essays on AI, business decisions, stoicism and building a company of one. en Sun, 04 Oct 2026 12:00:00 GMT Growth will not split the gains from AI by itself https://solertio.pl/blog-anthropic-scenariusze-ai https://solertio.pl/blog-anthropic-scenariusze-ai Sun, 04 Oct 2026 12:00:00 GMT In the extreme scenario of an Anthropic Institute report, all the GDP gain goes to capital. Why growth will not split it, and one rule to write down. Stop measuring development by what the model does for less https://solertio.pl/blog-heutagogia-capability https://solertio.pl/blog-heutagogia-capability Sat, 03 Oct 2026 12:00:00 GMT Competency is the correct execution of a known task under predictable conditions. That is also what the model does for less. Four objections included. What the winning 28% do https://solertio.pl/blog-co-robi-28-procent-wygranych https://solertio.pl/blog-co-robi-28-procent-wygranych Fri, 18 Sep 2026 12:00:00 GMT Gartner: 28% of AI use cases in I&O meet ROI expectations. Winners credit embedding into an existing workflow. I test how much of that list survives. Governance without theatre https://solertio.pl/blog-governance-bez-teatru https://solertio.pl/blog-governance-bez-teatru Fri, 18 Sep 2026 12:00:00 GMT One question for every control in AI governance: what would it have shown if the incident had happened yesterday, and after how long? One operator, a fleet of agents https://solertio.pl/blog-jeden-operator-flota-agentow https://solertio.pl/blog-jeden-operator-flota-agentow Fri, 18 Sep 2026 12:00:00 GMT My claim: a one-person operator with agents skips the internal layer of coordination between employees. Brooks, Coase, a falsification test. This is not an IT project https://solertio.pl/blog-to-nie-jest-projekt-it https://solertio.pl/blog-to-nie-jest-projekt-it Fri, 18 Sep 2026 12:00:00 GMT A project has an end date; AI value comes from continuous process adaptation. The day-400 test: who owns it, whose result it is, who pays for degradation. I taught my bot to confuse important with urgent https://solertio.pl/blog-important-without-a-deadline https://solertio.pl/blog-important-without-a-deadline Tue, 01 Sep 2026 12:00:00 GMT A triage rule welded importance to urgency and dropped 32 of 84 items as noise. Simon and Leroy on what a personal attention tool really does. The company brain and the knowledge nobody wrote down https://solertio.pl/blog-enterprise-brain https://solertio.pl/blog-enterprise-brain Sat, 22 Aug 2026 12:00:00 GMT A company brain holds only what the company recorded. Hayek, Walsh and Ungson, and a 2026 benchmark on internal data mark where that boundary sits. Antifragility: how to stand on the right side of volatility https://solertio.pl/blog-antifragility https://solertio.pl/blog-antifragility Sat, 08 Aug 2026 12:00:00 GMT Taleb's antifragility as an operating discipline: convexity instead of forecasts, the barbell, a worked example in hours, and three honest objections. Flow has a spec sheet — and hyperfocus does not meet it https://solertio.pl/blog-flow https://solertio.pl/blog-flow Sat, 08 Aug 2026 12:00:00 GMT Csikszentmihalyi measured flow with pagers: three entry conditions and a channel between anxiety and boredom. ADHD hyperfocus misses the control one. Mimetic desire: why you want what others want https://solertio.pl/blog-mimetic-desire https://solertio.pl/blog-mimetic-desire Sat, 08 Aug 2026 12:00:00 GMT Girard's triangle of subject, model and object explains borrowed wanting in careers, rivalry and feeds, with a scorecard audit of one career decision. Via negativa: the discipline of subtraction https://solertio.pl/blog-via-negativa https://solertio.pl/blog-via-negativa Sat, 08 Aug 2026 12:00:00 GMT Taleb's via negativa: negative knowledge outlasts positive claims, and removal beats addition. A calendar audit that gives back eighteen percent. The agent demotions will start after the incidents https://solertio.pl/blog-agent-demotions https://solertio.pl/blog-agent-demotions Tue, 04 Aug 2026 12:00:00 GMT Gartner forecasts 40% of enterprises will demote autonomous agents by 2027. The cause is binary governance: autonomy is not access scope. 95, 80, 28: the war of AI failure numbers https://solertio.pl/blog-ai-failure-number-wars https://solertio.pl/blog-ai-failure-number-wars Tue, 04 Aug 2026 12:00:00 GMT MIT says 95%, RAND is cited for over 80%, Gartner reports 28% success. Each counts something different, and methodology writes the headline first. AI-ready data is a power problem https://solertio.pl/blog-ai-ready-data-power https://solertio.pl/blog-ai-ready-data-power Tue, 04 Aug 2026 12:00:00 GMT Gartner forecasts 60% of AI projects without AI-ready data will be abandoned. The binding constraint is usually ownership and incentives, not pipelines. Mainframe and the boundary of GenAI competence https://solertio.pl/blog-mainframe-genai-boundary https://solertio.pl/blog-mainframe-genai-boundary Tue, 04 Aug 2026 12:00:00 GMT Gartner forecasts over 70% of 2026 mainframe-exit projects will miss their benefits. Forty-year-old COBOL is the worst case for GenAI, for named reasons. Too much, too fast: the mechanics of expectation inflation https://solertio.pl/blog-expectation-inflation https://solertio.pl/blog-expectation-inflation Mon, 27 Jul 2026 12:00:00 GMT 57% of I&O leaders who reported an AI failure blamed expecting too much, too fast. Expectations, not the model, and a layer you can contract. What Gartner's 28% actually says https://solertio.pl/blog-gartner-28-percent https://solertio.pl/blog-gartner-28-percent Mon, 27 Jul 2026 12:00:00 GMT Gartner's 28% is not an AI failure rate. It is I&O leaders self-rating their own use cases, one survey with two denominators. What to ask first. Pilot purgatory: where AI projects go to die https://solertio.pl/blog-pilot-purgatory https://solertio.pl/blog-pilot-purgatory Mon, 27 Jul 2026 12:00:00 GMT S&P puts it at 46% of AI proofs-of-concept scrapped before production. The purgatory is not incompetence; it is how the pilot was designed. ROI doesn't live in the model https://solertio.pl/blog-roi-not-in-the-model https://solertio.pl/blog-roi-not-in-the-model Mon, 27 Jul 2026 12:00:00 GMT Gartner says AI ROI tracks integration, governance and fit, not model sophistication. A Theory-of-Constraints read on stalled projects. Your company is a tribe — anthropology says design the rituals on purpose https://solertio.pl/blog-anthropology https://solertio.pl/blog-anthropology Sat, 11 Jul 2026 12:00:00 GMT Geertz's thick description aimed at your own company: culture as a designed system, onboarding as a rite of passage, and where 'we are a family' lies. The cat is an accidental stoic, and it has something to teach the founder https://solertio.pl/blog-cats https://solertio.pl/blog-cats Sat, 11 Jul 2026 12:00:00 GMT A cat embodies Epictetus and Laozi without reading them: indifference to approval, presence, no chasing. Worth borrowing, and impossible to copy. Cognitive sovereignty: choosing whose model of the world runs in your head https://solertio.pl/blog-cognitive-sovereignty https://solertio.pl/blog-cognitive-sovereignty Sat, 11 Jul 2026 12:00:00 GMT Feeds optimise for engagement and LLMs supply default frames. The second brain as infrastructure, AI without outsourcing judgment, the bubble risk. Deep work for an ADHD brain: blocks, or sprints? https://solertio.pl/blog-deep-work https://solertio.pl/blog-deep-work Sat, 11 Jul 2026 12:00:00 GMT Newport's deep-work insight holds; his monastic four-hour block assumes a neurotypical brain. Hyperfocus sprints may fit ADHD better, with caveats. Mismatch: why a Stone Age brain keeps sabotaging a modern life https://solertio.pl/blog-evolutionary-psychology https://solertio.pl/blog-evolutionary-psychology Sat, 11 Jul 2026 12:00:00 GMT Evolutionary mismatch: ancient drives firing at modern supernormal stimuli. A design tool from Tinbergen, and where the field earns its bad name. First principles: a real lever with a real bill https://solertio.pl/blog-first-principles https://solertio.pl/blog-first-principles Sat, 11 Jul 2026 12:00:00 GMT Reasoning from first principles occasionally buys a decade of advantage, and misused it reinvents the wheel. When to decompose, and Chesterton's fence. Houellebecq's warning: you can win the game and still have nothing https://solertio.pl/blog-houellebecq https://solertio.pl/blog-houellebecq Sat, 11 Jul 2026 12:00:00 GMT Houellebecq as the dark mirror of the optimisation canon: Durkheim's anomie, and a life maximised for freedom that is stripped of meaning. Engineer the identity, not the willpower https://solertio.pl/blog-identity-engineering https://solertio.pl/blog-identity-engineering Sat, 11 Jul 2026 12:00:00 GMT Behaviour follows identity, not willpower. Bem's self-perception theory and James Clear's votes, plus the point where an engineered identity cages you. Invert, always invert: the founder's cheapest edge https://solertio.pl/blog-inwersja https://solertio.pl/blog-inwersja Sat, 11 Jul 2026 12:00:00 GMT Munger borrowed inversion from Jacobi. It is the founder's highest-leverage, lowest-cost thinking tool, and it fails in three specific ways. Writing is thinking: the case for long formats in a short world https://solertio.pl/blog-long-form-thinking https://solertio.pl/blog-long-form-thinking Sat, 11 Jul 2026 12:00:00 GMT You do not know what you think until you write it at length. The essay as a forcing function, compounding, and where depth becomes procrastination. Not everything that counts can be counted — but "meaning over metrics" is a dangerous excuse too https://solertio.pl/blog-meaning-over-metrics https://solertio.pl/blog-meaning-over-metrics Sat, 11 Jul 2026 12:00:00 GMT Goodhart's law kills dashboard companies, and 'meaning over metrics' is the phrase founders use to dodge accountability. The McNamara fallacy, both ways. Naval's formula for wealth — and the base rate it hides https://solertio.pl/blog-naval https://solertio.pl/blog-naval Sat, 11 Jul 2026 12:00:00 GMT Specific knowledge plus permissionless leverage, code and media, is the solo builder's operating system. The survivorship bias behind it is real too. Stop fighting symptoms: the structure is the problem https://solertio.pl/blog-patterns-over-symptoms https://solertio.pl/blog-patterns-over-symptoms Sat, 11 Jul 2026 12:00:00 GMT Recurring problems are your system working as built. Meadows and Senge on leverage points and shifting the burden, and when to grab the bucket first. Your brain is a prediction machine — and that reframes ADHD https://solertio.pl/blog-predictive-coding https://solertio.pl/blog-predictive-coding Sat, 11 Jul 2026 12:00:00 GMT Friston's predictive coding says the brain guesses the world and checks with the senses. Attention becomes precision; ADHD, a dysregulated loop. When machines do the execution, taste is the job https://solertio.pl/blog-rick-rubin https://solertio.pl/blog-rick-rubin Sat, 11 Jul 2026 12:00:00 GMT Rick Rubin cannot play or engineer, yet defines records. As AI collapses the cost of execution, judgment becomes the scarce thing, with limits. "And then what?" — the two words that separate good decisions from clever ones https://solertio.pl/blog-second-order-thinking https://solertio.pl/blog-second-order-thinking Sat, 11 Jul 2026 12:00:00 GMT First-order effects are obvious; second-order effects arrive late and reverse them. Howard Marks, the pricing tax, and where the question overreaches. Solitude is a skill, and you have to build the room for it https://solertio.pl/blog-solitude https://solertio.pl/blog-solitude Sat, 11 Jul 2026 12:00:00 GMT Solitude is not loneliness but the chosen state where original thought happens. Storr and Newport on solitude deprivation, and the isolation risk. You are playing a status game right now — the only question is which one https://solertio.pl/blog-status-games https://solertio.pl/blog-status-games Sat, 11 Jul 2026 12:00:00 GMT Most behaviour is status-seeking: dominance, virtue, success. Storr's three games, how to swap yours, and the limits of status reductionism. Stoicism is an operating system for uncertainty — not a mug quote https://solertio.pl/blog-stoicism https://solertio.pl/blog-stoicism Sat, 11 Jul 2026 12:00:00 GMT Dichotomy of control, premeditatio malorum, amor fati: Stoicism as a working system for judgment. And the counterfeit, suppression sold as strength. The second life: what changes when you actually believe you get only one https://solertio.pl/blog-the-second-life https://solertio.pl/blog-the-second-life Sat, 11 Jul 2026 12:00:00 GMT A second life begins when we stop treating time as something guaranteed. On memento mori, deliberate choices, and the traps of productivity.