Imagination 
Manifesto of the Future

The 3rd millennium or Common Era is the current millennium spanning the years 2001 to 3000. Ongoing futures studies seek to understand what will likely continue and what could plausibly change in this period and beyond. The rise of a global economy and Third World consumerism marked the beginning of the 21st century that will end on 31 December 2100. 

We are leaving behind the 20th century, an era characterised by massification, ideological totalism, unprecedented speed that historiography should define as The Age of Velocity with particular reference to the dizzying rate at which life transformed within a single human lifespan (going from horse-drawn carriages to moon landings in under 70 years). Living at the dawn of the  21st century is a rare and powerful opportunity to shape our future.  

Over the past two decades, global markets have experienced unparalleled upheavals. The early 2000s were marked by the dot-com bubble burst and subsequent recovery. The 2008 financial crisis exposed systemic fragilities causing widespread losses and regulatory overhaul. Quantitative easing has become the norm in developed economies, fuelling asset bubbles and low interest rates. The rise of China and emerging markets shifted economic power centers. Technology stocks soared driven by the internet boom, followed by a sharp correction in 2000 and again in 2021. 

Cryptocurrencies and blockchain technology gained prominence for financial decentralization and autonomy, propelled by the idea of individual sovereignty and clear dissatisfaction with monetary policies, leading to a speculative frenzy and regulatory scrutiny. 

Real estate and infrastructure have converged into a unified opportunity, bringing governments and businesses together under increasing popularity of Public-Private Partnerships (PPPs). 

Capital markets remain unremarkably stable with a few dominant players controlling a significant share that make the whole ensemble look fantastic. 

Private equity and venture capital grew exponentially, with record levels of dry powder, while increasing barriers to entry. 

Complex fund structures, regulatory cost and asymmetric GP-LP interests created adverse business conditions, causing asset underperformance and difficulties in exit: the proliferation of “zombie funds”.

SPACs and ETFs emerged as alternative listing vehicles, gaining popularity until regulatory crackdowns in 2022. 

The investment landscape is no longer a neat collection of distinct silos. Hedge funds, Venture Capital (VC) and Private Equity (VC) are increasingly engaging in strategic arbitrage, bleeding into each other’s traditional territories. 

Driven by a relentless search for alpha and constrained capital deployment in their core mandates, these giants are chasing growth in adjacent spaces. 

VCs are extending their reach into later-stage growth equity, PE firms are dabbling in venture-style operational turnarounds and hedge funds are building significant stakes in private companies. 

The overlap signifies a desperate scramble for opportunities, a testament to the narrowing of traditional investment avenues and the recognition that superior, sustainable returns lie in embracing broader operational and corporate horizons, not just traditional blind pools.

The “evergreen” or “holdings company” model, exemplified by titans like Berkshire Hathaways, Jim Ratcliffe, Lord Ashcroft, Access Industries, Bollorè, Arnault, Tencent and other savvy media conglomerates, is not just a nostalgic relic but a potent, enduring strategy. These industrial groups thrive on long-term value creation, eschewing the short-term obsession, typical of a fast paced world. 

They possess the capital, flexible infrastructure and patient vision to nurture and grow businesses, not merely trade them. 

In a volatile economy and stark contrast to the often-fragile nature of many contemporary funds, their diversified portfolios are less susceptible to single-sector downturns, their ability to allocate capital internally cross-pollinate best practices and weather market storms with unwavering conviction. Smart ownership (control) is a powerful antidote to demise.  

The startup ecosystem boomed but faced challenges in sustainable growth, and profitability. Access to start up capital counterintuitively tied up to existing wealth or traction, after government intervention led the markets to risk aversion, and the average venture capitalist to prefer “ready meals”. 

The ascent of search funds and entrepreneurship through acquisition (ETA) is not a fleeting trend but a logical shift in the entrepreneurial playbook of a fast-paced, risk conservative ecosystem. 

For aspiring founders, the arduous, capital-intensive and high-risk journey of bootstrapping a business from zero (middle class privilege) or raising start up capital (upper class privilege) is being supplanted by a structured, and de-risked entry strategy. 

Harking back to the "buy and build" model, ETA allows to acquire established, cash-flowing businesses, leveraging leadership and operational acumen, instead of pursuing organic growth. 

This shortcut to size offer immediate revenue, proven traction and institutional qualification, drastically reducing the Valley of Death, the time to market and the probability of venture idea failure (Blitzscaling). 

In lieu of crossing the pre-seed capital desert, ETA is a pragmatic and alternative route to the mythological "garage start up" narrative, flagging a permanent evolution in how new enterprises are built. 

GP seeding/Sponsor financing is the actual funding gap risk-averse VCs are solicited to close to serve their market function, and avoid extinction (read “The Ineffective Venture Capital Interference”, Vincent Bozzino).   

Preventive red tape did not reduce the excessive risk-taking with financial stability, rather weakened markets with artificial inefficiencies, inequality and an overall systemic fragility: margin or capital requirements for lending, venture capital and sponsor finance are routinely transferred to the borrowers, what I call regulatory cost pass through.

Systemic risks nonetheless increased with rising debt levels, corporate leverage and geopolitical tensions. 

Market volatility returned with frequency, driven by inflation shocks, contractionary policy shifts and technological disruptions. 

Financial engineering accelerated under the veneer of innovation and still outpaced regulation, introducing new risks (e.g. the bankruptcy of FTX, the Wirecard scandal, Bill Hwang’s Archegos Capital and Credit Suisse collapse). 

Governments can reduce economic crisis and manage stability by either (i) reducing the population, (ii) re-distributing the fortune or reducing the wealth gap between the bottom, and the top of the society or (iii) expand markets. Reducing the population is not a practical or acceptable approach, whereas redistributive policies (largely fiscal, tax income-dependent such as social welfare or civil society-funded) failed to address inequality or actively stimulate economic growth, for the last 30 years. 

Insufficiently balanced policy mix have made the financial system indolent and markets dependent on money stock. Previously, start up access to high-yield bonds ignited entrepreneurship, economic growth and social mobility; banks and private funds were incentivized by collaborative debt monetization to reinvest the return on leveraging productive enterprises back into the economy, for maximum liquidity creation. 

Today, market activities are fundamentally driven by passive gearing (margin trading, warehousing, third party capital base) ever refraining from pre-money investing or returning profits via taxes, in absence of government backing. 

Savings from average workers (pension funds) and banks finance government deficits, junk bonds were repackaged into ETFs and others to tilt flat portfolio returns of the same pension fund contributors; ultimately, lower and middle classes have real liquidity and bear the first loss position, at any market level (read more in “The Institutional Construction of Equity”, Vincent Bozzino).

Selfish governments retreated to hands-off authoritarian control and  interventions, afraid of bailout capitalism or losing national debt buyers; however, neutral economic stimuli such as bail-ins, write-offs, CARES act equally incentivise accountability and risk discipline, without destroying existing capitalism structures or breaching the fiduciary duty of a State toward its legitimators, and financiers. 

Financial actors inherited myopia, following market inertia and a sort of modern real bills doctrine. The current state of the world is a paralysed, illiquid closed loop with apparent risk parity and economic dynamism. 

There is no shortage of capital, the problem is lack of expansionary credit, transparent administrations and fresh ideas on how to create sustainable value, and commonwealth. 

The war on leverage is political, hypocritical as the war on terrorism or immigration: a false front by the State for control and economic advantages, caused by their own actions, while demonizing the very tools they wield.  

Technology has caused a global shift that enabled unlimited consumer choice, boundless scale and uncapped opportunities. Experience economy and immersive tech are dominating the markets. Cross-sector, we are living in the era of personalisation and egocentric brands. 

Various sectors have been dismantled by the internet, IT and AI boom, whereas the creative industries smoothly integrated innovation, becoming a playground for disruptive tech applications. 

Media & entertainment turned less sensitive to market cycles, less cyclical than ever, thanks to the direct, “digital” delivery method.  

Mediatech is now everywhere, 24/7: it’s in the music you listen to, the movies you watch, and the books you read. It’s in the advertising you click, the video games you play and the text conversations you have.  

Over the past two decades, institutional allocators, family offices and alternative asset managers understood financing media is not a speculative gamble, rather a structured, high-yield asset class. 

Spreading capital across a diversified portfolio of 5 to 15 titles transforms movie financing, from binary "hit-or-miss" risk into a predictable, non-correlated financial vehicle. 

Historical data and private equity benchmarks for structured film slates show target Internal Rates of Return (IRR) between 14% and 26% with Multiple on Invested Capital (MOIC) of 2.0x to 2.5x over 3-5 year investment cycles. 

Digital streaming infrastructure turned creative IP into a globally distributed asset that generates revenue around the clock, across every timezone and region, without incurring marginal logistics costs.

Even if the box office of a $15m 4-title slate with $7m risk capital yields modest results ($8m worldwide), subsequent streaming rights (Netflix/Amazon licenses), TV syndication and FAST channel ad revenue generate $16m long-tail monetization over a 5-year window, yielding a net LP return of ~$24m (20%+ IRR and 1.6x+ net payout). 

Authors can sell niche or specialized books perpetually through platforms, generating ongoing royalties. 

By investing $100,000 upfront mainly into music marketing, a record label launches a rising DJ producer with $100,000+ annual recoupment from touring (50-80 shows/year @ $3,000 booking fee), digital sales (streams & downloads) and licensing income ($20,000 for a dance track used in commercials, movies or video games), generating revenue indefinitely. 

Media cash flows continue perpetually at minimal ongoing maintenance cost, outpacing typical tech product cycles. It’s an evergreen catalog value. Terminal IP value makes money forever. 

Film and TV slate returns exhibited near-zero correlation with public equity or fixed-income markets, offering genuine orthogonal diversification during economic downturns.

With charitable activities and culture tech at their peak, academic institutions and research ecosystems endure as the world's largest pools of capital and intellectual property catalysts. 

Endowments cater to $1.6+ trillion academic R&D and philanthropic giving ecosystems, around the world, with a distinct, structural advantage. 

The "perpetual reinvestment" flywheel leverages non-dilutive, zero-cost capital to fund early-stage development, drastically de-risking venture investments, while preserving full equity upside in high-barrier intellectual property.

Similar to how university endowments (e.g. Harvard, Yale, Stanford) compound capital over decades, combining equity exits with recurring top-line IP royalties activated a self-sustaining engine for discovery, and enabled business-disciplined sociocultural impact.

Media portfolios, academic R&D/IP and donor markets display zero marginal cost of replication with strong downside protection from 30-40% government match funding and instant, global geographic reach (infinite scalability), compared to high CapEx of traditional tangible assets, risk mitigation limited to physical asset collateral and geo location limit. 

Intangible asset portfolios operate on a power-law return structure with an asymmetric risk-reward profile. 

While downside risk is reduced by upfront soft money incentives, grant co-funding or debt, the upside on a runaway hit horror movie, major tour production or a blockbuster scientific patent is mathematically uncapped, providing massive portfolio tilt and abnormal alpha.

Intellectual property, backed by long copyright (70–95+ years) and patent protection periods is an income annuity for the next century. 

As global distribution platforms expand (e.g. new AVOD services, international broadcast networks, digital health platforms), old IP library assets experience periodic "re-monetization waves”, without requiring new capital expenditure.

The expansion of immersive digital environments (metaverse, mixed reality) generated new markets for entertainment, social interaction and commerce. Geographically, North America is the biggest media, tech and entertainment market followed by UK - countries where mid-market Asian companies operate too - and selectively, Europe. 

Middle markets present larger opportunities than the majors: (i) still fragmented and undercapitalised, (ii) product invariably easily transferred. 

Despite the rise of the creative class, AI is where everyone is looking now, motivating unprecedented enterprise and capital spending to deep tech, AI-startups or AI-first, expecting Artificial General Intelligence (AGI) or Superintelligence to overtake human intelligence, by the end of the century. 

The world is on the brink of a transformational point in human history as transhumanists cross the Rubicon.  

There are two possible futures for humanity. The first is the apocalyptic  scenario currently being sold to us: artificial general intelligence (AGI) arrives, cognition is substrate-independent, machines acquire superintelligence: an intellect that greatly exceeds human capacity across virtually every domain, as Nick Bostrom put it. The machine ceases to be an instrument and becomes an agent. 

Should we gift full agency to a powerful digital mind, technological singularity and instrumental convergence would be serious, large-scale existential risks for human beings, becoming endangered species or extinct. Cross-partisan groups advocate for AI limited to serve humankind rather than replace it or compete against it.

The second future is considerably less cinematic, consequential and much more beneficial. AGI does not make it. Instead, we exploit increasingly extraordinary systems of sensing, prediction, generation, optimization and action that we built, without ever crossing the ontological boundary between computation and cognition. Nouvelle AI, narrow AI, synthetic intelligence, intelligence amplification and machine agency grow spectacularly useful for specific tasks but remain fundamentally different from human-like intelligence. Machines do not wake up: they get extraordinarily good at acting, while remaining incapable of originating the reasons for.

An increase in the material means at the disposal of individuals may still present dangers, unless it is accompanied by a corresponding spiritual effort.

To achieve the vision of AGI, we need an unknown number of major, historic breakthroughs crossing humanities and sciences, namely: deciphering, computing and replicating (i) human consciousness, (ii) intentionality, (iii) representation, (iv) free will and (v) imagination; in other words, full human intelligence or our infinite mind. 

Not merely our genome, our neural architecture, our behavioural patterns or the statistical regularities of language. Defining, computing and replicating our infinite mind is the lacuna. 

Not just an engineering flaw waiting for smarter fixes, in the decade ahead. It's a long-standing existential conundrum keeping both natural and social scientists up for centuries. Not because they are less clever, in fact, computer scientists exhibit a bad case of Dunning-Kruger effect.

The strong AI hypothesis is the acid test. Marvin Minsky’s paradox earmarked the difficulty in achieving super intelligence: some of the hardest human abilities to reverse engineer are precisely those operating beneath conscious awareness. The rational sphere is computationally easy, because you can formalize it. The irrational sphere is hard to, entangled with embodiment, biography, emotion, sociality, memory, desire and the peculiar phenomenology of being a person. 

The unconscious mind has unlimited processing power. That is the territory AI has not conquered and likely, cannot conquer until we do.

The paradox should have made AI researchers humble. On the other hand, the industry got accustomed to dismiss the concern, flaunting a laser focus on behaviour only. 

Many in the field consider it out of scope, because machines need not be an imitation of human beings, rather useful, controllable systems that exhibit intelligent or superintelligent behaviours, without implying autonomous intentions or consciousness (synthetic intelligence). 

Others call the challenges undermining, because AI achievements are constantly reclassified as non-intelligent, as soon as machines accomplish them (AI effect). 

Transhumanists believe that the reproduction of an “actual” mind (strong AI) is unnecessary, because irrationality or unconscious aspects are inefficiencies or vulnerabilities to overcome through technological means: they reject the idea of intentionally replicating or preserving reducible or undesirable features of human minds in machines; it’s counterproductive, especially if such limitations lead to flawed judgments or destructive behaviours. 

The future human or posthuman condition is supposed to be one of more rational, more capable or enlightened beings, driven by optimization. 

Excellent engineering and insufficient metaphysics does it, just like construction projects or behavioural economics. 

Cognitive performance may not exhaust what it means to rival humanity in intelligence, though; a machine can become better than us at calculation, memory, search, translation, diagnosis, prediction, coding and strategic optimization. 

To ultimately match or beat our brain power as a superhuman, something much stranger is required: the capacity for motivation, judgment, imagination, self-awareness and the endogenous generation of purpose. 

Agency is not just behaviour produced in response to an organism's environment (extrinsic motivation) but mostly the product of intrinsic motivation. What we proudly call free will. Here the mainstream AI argument gets circular. 

De facto and de jure, consciousness is not just sentience, it’s also self-awareness. Intelligence is not merely a force (Alexander Wissner-Gross), an agent's ability to achieve goals in a wide range of environments (Shane Legg & Marcus Hutter) or goal-directed adaptive behaviour (Robert Sternberg). Intelligence is still poorly understood: it’s not just the capacity for logic, understanding, learning, reasoning, planning and problem-solving. These capabilities represent the rational half of cognition but human beings are not merely reasoning machines. Luckily. 

Reductive definitions are useful only to abstract away the core dilemma, failing to address the will to power that brought human race, far and wide, at the summit of the animal kingdom across 21 centuries. 

Intelligence is a human psychosocial competence that also entail the potential for finding or creating problems, as Howard Gardner put it. 

Optimization presupposes an objective, intelligence can invent the objective. The distinction matters enormously in creating superintelligence.

Contemporary AI enthusiasts infer agency from responsiveness, understanding from linguistic fluency, intention from coherence and interiority from the appearance of personality. The interface became evidence of the thing behind the interface, as I elaborated in my latest book "The Aesthetic Shift. Normative Conflict and Value Corruption". 

Behaviourist descriptions tell us what an organism does in response to an environment but don’t necessarily explain why an organism originates behaviour, in the first place. That’s the demarcation between extrinsic and intrinsic motivation, machine agency and human agency. To wit, a machine responds to incentives, a person can manufacture the incentives. Agency is not merely the ability to execute an objective, it’s the capacity to generate, revise, reject or replace objectives. The contrast is subtle enough to disappear inside a benchmark but enormous enough to determine whether we are building tools or new forms of life. 

Current AI remains overwhelmingly extrinsic, as a result. Prompt it, reward it, optimize it, evaluate it, give it a context window, assign a goal, feedback, then observe the output. The system is magnificent at responding but where is the original input? Where is the desire that was not supplied? Where is the problem it invented? Where is the question nobody asked to ask? Where is the aesthetic impulse that precedes the instruction? Where is the refusal to optimize, because the objective itself has become meaningless? 

Hence, much of what mainstream AI call agency is, at least in its strongest marketing form, a category error. We are too focused on optimisation, at the expense of creative exploration. AI slop is the tell: there is no original input.

The enormous volume of synthetic content flooding the internet is not evidence that machines have become creative; creative AI is proof that humanity is the ceiling and machines are extraordinarily efficient at producing only artefacts that resemble the statistical surface of existing works of human beings. 

A human artist can decide to make something; present machines are asked to make something. The fine line is not semantic, it’s causal. 

Until machines possess some form of intrinsic motivation, creative AI remains an impressive, licensed or unauthorised simulation of a copyrighted human output from the past, rather than an explanation of the impulse that generates the output. 

A machine can write essays but writing is not just language generation. A machine can generate images but creativity is not just pattern recombination. 

We are suffering the ELIZA effect, our tendency to project mind onto systems that merely manipulate symbols in operationally convincing ways and attribute human-like qualities to machines that are really just executing scripted or pattern-based responses, with no emotional capacity. 

Based on interdisciplinary research, we now know that consumers mostly purchase for emotional reasons and then rationalize their decisions afterward. Ultimately, the only way to create real profit is to bank on the emotional rather than rationality of both internal and external stakeholders, by appealing to feelings, personality and imagination.  

Major companies who build stoic AI tools simultaneously announced that their own technology may destroy civilization, at length, soliciting governments to take greater control over it, as soon as possible. 

Andrew Ng has correctly pointed out how the joint warning by Google DeepMind, OpenAI and Anthropic about our excitation risk from AI is just regulatory capture. 

Fear is a market instrument and their choice of staking on irrationality to promote hyperrational products should suffice alone to validate the whole post-AI thesis. 

Regulation is not inevitable, AI is not an existential threat, the unwise creators are. No conspiracy intended, no incentive architecture wanted either, though.  

There is a strange tendency in Silicon Valley to regard anything short of the maximal outcome as failure. Ambition, like fear, is proof of how the emotional is the tip of the iceberg. 

A technology can be transformative, without fulfilling the prophecy attached to it. Humankind should spend considerably more time understanding its own ontology, before attempting to instantiate it elsewhere. 

Before we can embed the unconscious into a machine, we must discover what free will actually is and coherently address it across social sciences.

We should not manufacture an artificial version of a phenomenon that we have not yet philosophically, scientifically and humanistically reconciled with ourselves. The human species has been spending centuries - not product cycles - to reach unanimous confidence about consciousness, intention, irrationality and will to power. We cannot reproduce what we cannot define yet. We cannot distribute to machines what we cannot first locate in our genes. Scaling statistical systems does not necessarily produce agency and human-like intelligence. That is not pessimism, it’s epistemic hygiene.

Another AI winter is due, before the next AI alignment phase, another cycle where the frontier is going to simultaneously expand and reveal its own limitations.

The history of technology is not a straight line toward an inevitable destination. It’s a succession of intellectual fashions, engineering breakthroughs, broken promises, opaque funding cycles and conceptual revisions. It will force us to distinguish intelligence from performance, agency from optimization, creativity from generation, consciousness from simulation and autonomy from execution. It would also force investors to distinguish a technological revolution from a technology narrative. Anything can happen, if bankrolled; even revolutions we don't need or actually exist.

We don’t need AGI for AI to change civilization, we don’t need conscious machines, we do not need digital gods. 

We need systems that amplify human intelligence enough to improve the economics of work, production, healthcare, education, happiness and fun.

That is already a monumental transformation and more analytical than the intellectual laziness of treating AGI as inevitable. 

This isn’t an AI-driven economy, it’s an economy driven by AI infrastructure spending, which is very different. The entire sector is running on commitments, not returns. Revenue is booming but profitability is not. Valuation without operational truth is theater. 

The mistake is not believing that AI will become powerful, it’s assuming that power necessarily culminates in intelligence. AI is poised to increase both demand (by creating new opportunities) and supply (by making production more efficient) but irrationality is the intermediate neglected variable in the broad AI impact.

That’s why the hard AI mirage deserves total skepticism with an educated guess. Treat it as a weak scenario, not as an approaching train. 

I am confident that AI is going to take the three-sector economy and post-industrial society by storm, shortly, for our emancipation and offer brief glories but will never, ever fully bleed into the creative industries - the Ars Longa - across the Indigo Era, where our infinite mind, imagination and originality are the motherboard. The best future is smaller than AGI.