Marius Schober

  • The First Principles of a Post-AGI Business

    OpenAI released its new o3 models and numerous people argue that this is in fact Artificial General Intelligence (AGI) – in other words, an AI system that is on par with human intelligence. Even if o3 is not yet AGI, the emphasis now lies on “yet,” and – considering the exponential progression – we can expect AGI to arrive within months or maximum one to two years.

    According to OpenAI, it only took 3 months to go from the o1 model to the o3 model. This is a 4x+ acceleration relative to previous progress. If this speed of AI advancement is maintained, it means that by the end of 2025 we will be as much ahead of o3 as o3 is ahead of GPT-3 (released in May 2020). And, after achieving AGI, the self-reinforcing feedback loop will only further accelerate exponential improvements of these AI systems.

    But, most anti-intuitively, even after we have achieved AGI, it will for quite some time look as if nothing has happened. You won’t feel any change and your job and business will feel safe and untouchable. Big fallacy. We can expect that after AGI it will take many months of not 1-2 years for the real transformations to happen. Why? Because AGI in and of itself does not release value into the economy. It will be much more important to apply it. But as AGI becomes cheaper, agentic, and embedded into the world, we will see a transformation-explosion – replacing those businesses and jobs that are unprepared.

    I thought a lot about the impact the announced – and soon to be released – o3 model, and the first AGI model are going to have.

    To make it short: I am extremely confident that any skill or process that can be digitized will be. As a result, the majority of white-collar and skilled jobs are on track for massive disruption or elimination.

    Furthermore, I think many experts and think tanks are fooling themselves by believing that humans will maintain “some edge” and work peacefully side-by-side with an AI system. I don’t think AGI will augment knowledge workers – i.e. anyone working with language, code, numbers, or any kind of specialized software – it will replace them!

    So, if your job or business relies purely on standardized cognitive tasks, you are racing toward the cliff’s edge, and it is time to pivot now!

    Let’s start with the worst. Businesses and jobs in which you should pivot immediately – or at least not enter as of today – include but are not limited to anything that involves sitting at a computer:

    • anything with data entry or data processing (run as fast as you can!)
    • anything that involves writing (copywriting, technical writing, editing, proofreading, translation)
    • most coding and web development
    • SAAS (won’t exist in a couple of years)
    • banking (disrupted squared: AGI + Blockchain)
    • accounting and auditing (won’t exist as a job in 5-10 years)
    • insurance (will be disrupted)
    • law (excluding high-stake litigation, negotiation, courtroom advocacy)
    • any generic design, music, and video creation (graphic design, stock photography, stock videos)
    • market and investment research and analysis (AI will take over 100%)
    • trading, both quantitative and qualitative (don’t exit but profit now, but expect to be disrupted within 5 years)
    • any middle-layer-management (project and product management)
    • medical diagnostics (will be 100% AI within 5 years)
    • most standardized professional / consulting services

    However, I believe that in high-stakes domains (health, finance, governance), regulators and the public will demand a “human sign-off”. So if you are in accounting, auditing, law, or finance I’d recommend pivoting to a business model where the ability to anchor trust becomes a revenue source.

    The question is, where should you pivot to or what business to start in 2025?

    My First Principles of a Post-AGI Business Model

    First, even as AI becomes infallible, human beings will still crave real, raw, direct trust relationships. People form bonds around shared experiences, especially offline ones. I believe a truly future-proof venture leverages these primal instincts that machines can never replicate at a deeply visceral level. Nevertheless, I believe it is a big mistake to assume that humans will “naturally” stick together just because we are the same species. AGI might quickly appear more reliable, less selfish than most human beings, and have emotional intelligence. So a business build upon the thesis of the “human advantage” must expertly harness and establish emotional ties, tribal belonging, and shared experiences – all intangible values that are far more delicate and complex than logic.

    First Principle: Operate in the Physical World

    • If your product or service can be fully digitalized and delivered via the cloud, AGI can replicate it with near-zero marginal cost
    • Infuse strategic real-world constraints (logistics, location-specific interactions, physical limitations, direct relationships) that create friction and scarcity – where AI alone will struggle

    Second Principle: Create Hyper Niche Human Experiences

    • The broader audience, the easier it is for AI to dominate. Instead, cultivate specialized groups and subcultures with strong in-person and highly personalized experiences.
    • Offer creative or spiritual elements that defy pure rational patterns and thus remain less formulaic

    Third Principle: Emphasize Adaptive, Micro-Scale Partnerships

    • Align with small, local, or specialized stakeholders. Use alliances with artisan suppliers, local talents, subject-matter experts, and so on.
    • Avoid single points of failure; build a decentralized network that is hard for a single AI to replicate or disrupt

    Fourth Principle: Embed Extreme Flexibility

    • Structured, hierarchical organizations are easily out-iterated by AI that can reorganize and optimize instantly
    • Cultivate fluid teams with quickly reconfigurable structures, use agile, project based collaboration that can pivot as soon AGI-based competition arises

    Opportunity Vectors

    With all of that in mind, there are niches that before looked unattractive, because less scalable, that today offer massive opportunities – let’s call them opportunity vectors.

    The first opportunity vector I have already touched upon:

    • Trust and Validation Services: Humans verifying or certifying that a certain AI outcome is ethically or legally sound – while irrational, it is exactly what humans will insist on, particularly where liability is high (medicine, finance, law, infrastructure)
    • Frontier Sectors with Regulatory and Ethical Friction: Think of markets where AI will accelerate R&D but human oversight, relationship management, and accountability remain essential: genetic engineering, biotech, advanced materials, quantum computing, etc.

    The second opportunity vector focuses on the human edge:

    • Experience & Community: Live festivals, immersive events, niche retreats, or spiritual explorations – basically any scenario in which emotional energy and a human experience is the core product
    • Rare Craftsmanship & Creative Quirks: Think of hyper-personalized items, physical artwork, artisanal or hands-on creations. Items that carry an inherent uniqueness or intangible meaning that an AI might replicate in design, but can’t replicate in “heritage” or provenance.

    Risk Tactics

    Overall, the best insurance is fostering a dynamic brand and a loyal community that invests personally and emotionally in you. People will buy from those whose values they trust. If you stand for something real, you create an emotional bond that AI can’t break. I’m not talking about superficial corporate social responsibility (nobody cares) but about authenticity that resonates on a near-spiritual level.

    As you build your business, erect an ethical moat by providing “failsafe” services where your human personal liability and your brand acts as a shield for AI decisions. This creates trust and differentiation among anonymous pure-AGI play businesses.

    Seek and create small, specialized, local, or digital micro-monopolies – areas too tiny or fractal for the “big AI players” to devote immediate resources to. Over time, multiply these micro-monopolies by rolling them up under one trusted brand.

    Furthermore, don’t avoid AI. You cannot out-AI the AI. So as you build a business on the human edge moat, you should still harness AI to do 90% of the repetitive and analytic tasks – this frees your human capital to build human relationships, solve ambiguous problem, or invent new offerings.

    Bet on What Makes Us Human

    To summarize, AI is logical, combinatorial intelligence. The advancements in AI will commoditize logic and disrupt any job and business that is mainly build upon logic as capital. Human – on the other hand – is authenticity. What makes human human and your brand authentic are elements of chaos, empathy, spontaneity. In this context, human is fostering embodied, emotional, culturally contextual, physically immersive experiences. Anything that requires raw creativity, emotional intelligence, local presence, or unique personal relationships will be more AI resilient.

    Therefore, a Post-AGI business must involve:

    1. Tangibility: Physical goods, spaces, unique craftsmanship
    2. Human Connection: Emotional, face-to-face, improvisational experiences
    3. Comprehensive Problem Solving: Complex negotiations, messy real-world situations, diverse stakeholder management

    The inverse list of AGI proof industries involve some or multiple aspects of that:

    • Physical, In-Person, Human-Intensive Services
      • Healthcare: Nursing, Physical therapy, Hands-on caregiving
      • Skilled trades & craftsmanship
    • High-Level Strategy & Complex Leadership
      • Diplomacy, Negotiation, Trust building
      • Visionary entrepreneurship
    • Deep Emotional / Experiential Offerings
      • Group experiences, retreats, spiritual or therapeutic gatherings
      • Artistic expression that thrives on “imperfection”, physical presence, or spontaneous creativity
    • Infrastructure for AGI
      • Human-based auditing/verification
      • Physical data center operations & advanced hardware
      • Application and embedment of AI in the forms of AGI agents, algorithmic improvements, etc. to make it suitable for everyday tasks and workflow

    The real differentiator is whether a business is anchored in the physical world’s complexity, emotional trust, or intangible brand relationships. Everything pure data-driven or standardized is on the chopping block – imminently.

  • This is AGI

    I say that current AI is AGI. It is not obvious yet, because we haven’t yet connected very complex and fragmented software and data environments – and for R&D to turn into real-world change is a multi-year process anyway.

    Even if we stopped and freeze AI development here and now, we’d only realize that we indeed have AGI 2 or 3 years down the road. In some niches it will be faster (software or law) in others slower (complex logistics).

    However, AI development is not stopping here and now. It continues to improve – I say exponentially. Even if you are more conservative, then the linear growth still has undoubtedly a large rate of change.

    Today (!), we have AI models that evolved from barely completing sentences to writing code that ships to production, we have AI doing PhD-level research, and achieved gold medal-level performance on the International Math Olympiad. AI is solving medical problems that baffle experts.

    Again – what is currently mostly manually prompted work in long chat conversations will soon develop into agents that can do almost all knowledge work fully autonomously.

    I’m not talking about AI as an assistant, as a co-pilot. It will just straight up finish the work while you are napping on the beach.

    The difference between the GPT-3 model and today’s models – whether Grok 4, Gemini 2.5 Pro, or ChatGPT o4 – is like comparing a Nokia 1011 to an iPhone 16 Pro. We went from purely text based chats to multimodal understanding – models that can see, hear, and reason across domains simultaneously. AI is starting to genuinely understand context and nuance in ways that feels human.

    The next phase is not purely larger AI models, but models that learn continuously. They can remember you, plan and execute multistep tasks over days, weeks, or months.

    An AI system that perfectly remembers, understands context, who never sleeps, and gets smarter every day. This is being built today in AI labs around the globe.

    We have AGI today, and it is only a matter of time for us to arrive at superintelligent AI systems. Is it 2 years? 3 years? 4 years? 5 years? Irrelevant. Whether it is 1 year or 10 years, the implications are the same: everything is going to change forever.

  • Invisible Genius

    Every time I wish the minds I admire would open up and show the full depth of their thoughts, ideas, and experiences, I realize – I’m doing the same thing by staying too silent online.

    The idols I look up to – for example Ido Portal, Bruce Poon Tip, Dr. Nun Amen-Ra, my Sifus, countless peers – are all masters of mind and body in private but leave little or no footprints online. The wisdom is real, yet it’s undocumented and the world barely sees it.

  • Why a Marathon?

    Marathon running is socially encouraged orthopedic self-harm disguised as virtue. 42 km of joint erosion, cardiac overreach, and dopamine-chasing disguised as discipline.

    Before you commit to run a marathon, take a moment and pause. Ask yourself: Why is running a marathon my goal?

    Odds are, you’re either unconsciously copying someone else’s coping mechanism or seeking social reward and Instagram likes through performative suffering – and calling it “fitness”. That’s not discipline. That’s mimesis. Read Girard. You’re not chasing health but someone else’s pain prescription, hoping it will fix your own.

    And if you really need to suffer publicly to feel alive, maybe it’s not your body that needs training, but your fear of sitting still without applause.

    If the idea of running a marathon truly comes from your heart, then you’d run 42 km on a Tuesday at dawn with no witness. If you need a start line, a medal, a round of applause, and pictures for social media, then you are suffering through a socially accepted identity crisis.

    Let’s get physiological. Marathons attack your hypothalamic-pituitary-adrenal axis, suppress your immune system, and can induce irreversible joint damage – knees, hips, ankles, worn one kilometer at a time. Cardiac scarring is a real risk. Cortisol floods. Lymphocyte counts crash. This really isn’t health but a ritualized system failure.

    Discipline is not a spectacle. Movement is not masochism. Read Body by Science (Doug McGuff & John Little). Explore movement systems like Ido Portal’s (or try Kung Fu). Train for capacity, not applause.

    Stop running from the void. It keeps up.

  • How We Use AI

    Whether current AI systems qualify as AGI is beside the point. Five years ago, if you had asked me to define AGI, my answer would’ve closely described what GPT o3 or Gemini 2.5 Pro are now. So if this is AGI, then where are the breakthroughs?

    Valid question. The answer: we are the bottleneck.

    The limitation is no longer the model. The real limitation is that we haven’t really figured out how to use LLMs properly. Even if AI development froze today, and all we have available are o3 and Gemini 2.5 Pro level LLMs, then we would still see a decade of profound disruptions and innovations across entire industries.

    Most users treat AI like Google, a friend, a mentor, or a novelty. Few understand prompting. Those who do don’t even scratch the surface of what is possible when you give AI the right prompt, the relevant context, and access to specific or perhaps proprietary data.

    Worse, we are not augmenting human intelligence, we are outsourcing it. TikTokified workflows, mindless automation, and prompt-template copy-paste culture are commoditizing subpar outcomes. Instead of expanding our minds, we’re paralyzing them.

    The real potential lies hidden in tandem cognition. Reimagining how we work with AI systems in a way that ensures our uniquely human traits (intuition, creativity, vision, …) aren’t ignored, but amplified. Without this shift, outputs will commoditize (across humans and organizations).

    We urgently need two things: first a methodology for extracting maximum value from LLMs and second a philosophy for not replacing our human genius, but empowering it.

    The future is not AI versus human. It is human with AI, at full capacity. Currently, the focus is on maximum capacity for AI compute. Now it’s time we focus on maximum capacity for human genius.

  • NOW

    The future is an illusion. Just as is the past.

    There is only now.

    Everything that is truly meaningful is happening now.

    You shape the future by your actions now, by your thoughts now, by your love now.

    Any moment is the future.

    There is only now.

  • Creativity and Commoditization

    Most products and technologies became commoditized. I believe this is mostly due to a lack of creativity.

    When was the last time you saw a product that totally caught you off-guard in absolute amazement?

    Nobody dares to create something uniquely NEW.
    Commoditization happens because everyone is just focused on incremental improvements, calling it proudly innovation. It is a delusion.

    Creativity is not listening to customers. It is creating something completely new that your future customers don’t even know can even exist. It is possible in your imagination but impossible in their imagination.

    Vision, creativity, intuition.

    I 100% believe that listening to customers is a trap that leads to mediocrity. True innovation comes from ignoring the noise of customer feedback and daring to invent what they can’t yet imagine – yes, it’s risky, but the only way to avoid competition.

  • Meritocracy Without Balance

    Meritocracy without equality is basically a ladder with missing steps, which only the fortunate few can climb. The rest is basically left at the ground to gaze up.

    Furthermore, a pure focus on meritocracy can become a weakness if it’s a zero-sum game. While talents thrive in systems that value it, without fairness, you ultimately get exploitation, resentment, or fragility.

    Ergo: meritocracy needs to be balanced with equality, and equality needs to be balanced with meritocracy.

    In a meritocratic system, you basically need that the fortunate and the able are compassionate towards those less fortunate and able.

    If you are more on the libertarian side, you need the fortunate individuals to compete as capitalists and then be socialists within their communities and families.

    If you prefer a state, then the state must encourage meritocratic contribution of everyone that is able, to the best of their ability, while balancing it with a fair welfare system that nurtures and supports those less able and fortunate.

    I think this is something Germany did historically quite well, but at one point we lost the balance: we lost the culture of merit by putting too much emphasis on equality – even to a point that Germany now cares for millions of non-citizens that never contributed, when it should instead demand these individuals to contribute to the best of their ability.

    If we want to have a state, then we don’t want a welfare state, and – I think – we should also not want a pure capitalistic state. We need balance and thus a social meritocracy.

  • Trump Tariffs

    I don’t believe that Trump uses the current tariffs as a negotiation leverage – at least not across the board.

    The U.S. will not go back to 0% or < 10% tariffs, because Republicans (JD Vance) will with absolute certainty lose re-election.

    Trump promised to revive the Rust Belt, which are the swing states. He can only “save” them through tariffs against China and all current and future China alternatives.

    Some regions, and I’d include the EU, UK, Argentina, Australia, New Zealand here, now have the unique chance to negotiate towards free trade or at least a friendly ≤ 10% tariff.

    Yet, speaking of the EU, I think they are mostly too arrogant or stupid to do so. 10% seems to be the new base anyhow.

    The relevant question is how the EU will respond.

    Currently, it looks that the EU is looking for sovereignty. This could mean new entry barriers for US technology products. Perhaps even a straight out ban of key technologies like Palantir to replicate local key players.

    This in turn is the ultimate and biggest threat to the USA. A world that is shifting away from US technology will erode the vital pillar that is currently keeping the US dollar alive.

  • Corporate AI vs. Visionary People

    BCG says that 83% of firms prioritize innovation, but only 3% feel able to execute.

    I ask: what’s the bigger flaw: overestimating corporate AI or underinvesting in people?

    Probably both.

    I see a trend where AI is seen as the cure-all.

    Instead of investing in visionary and creative leaders plus the engineers to execute, companies invest in AI gimmicks whose ROI is (for now) mostly in efficiency – not real innovation.

    Everyone – at this point – basically pretends it will lead to proprietary innovation.

    At the same time, the innovation companies end up running after, are the innovations that AI tells them to – and because AI is commoditizing logic, it simultaneously commoditizes innovation.

    Congratulations: You end up in a red ocean instead of blue ocean.

    Therefore, innovation becomes competition, and is not innovation anymore.

    The biggest mistake companies can make today is investing all in AI and not investing in the human genius of their workforce: visionary leadership, their collective intuitive intelligence, Human-AI-Symbiosis

  • Decentralization versus Authoritarianism

    Erdoğan’s calculated elimination of Imamoglu through academic technicalities and alleged ties to PKK is not really an isolated Turkish case but an example of democracy’s global collapse.

    Yesterday, Germany rushed constitutional changes without proper scrutiny and with a majority that was already voted out of office, Romania disqualifies candidates on procedural grounds. The list goes on: Hungary under Viktor Orbán, Serbia under Aleksandar Vučić, Israel under Netanyahu, Poland under the Law and Justice party – the democratic backsliding transcends regions and political systems. 72% of humanity now lives under authoritarian control.

    There is a new playbook: weaponize legal institutions against opponents, manufacture legitimacy through procedural theater, and dismantle democratic safeguards while maintaining the illusion of constitutionality.

    We’re witnessing not democracy’s dramatic assassination but its methodical strangulation through bureaucratic manipulation. This erosion isn’t coincidental but the inevitable outcome of centralized power structures that invariably corrupt even well-designed systems.

    I believe that our only viable path forward lies in radical decentralization: distributing governance to local communities, financial sovereignty through crypto networks, and communication via censorship-resistant platforms that no single entity controls.

    Decentralized systems restore human dignity by establishing unbreakable cryptographic guarantees rather than depending on the hollow promises of centralized authorities, career politicians, and unelected bureaucrats .

    The future belongs to networked individuals collaborating voluntarily through systems designed with liberty as their foundation. Decentralized and globally networked societies are antifragile societies that unleash innovation by enabling thousands of concurrent experiments instead of single-point failures.

    Only decentralization can safeguard freedom in an increasingly authoritarian world.