THE BLOG

Antony Slumbers 4/14/25 Antony Slumbers 4/14/25

Exponential Industries & Advanced Infrastructure: A Strategic CRE Investment Thesis

Executive Summary:
The next decade will be shaped by exponential industries – high-growth, high-dynamism sectors ranging from AI and robotics to new energy and biotechnology – that are poised to reshape the global economy.

In their October 2024 ‘The next big arenas of competition’ report, McKinsey Global Institute identified 18 “arenas of tomorrow,” and posited that they could generate $29–$48 trillion in annual revenues by 2040.

Critically, many of these emerging arenas require highly specialised real estate and infrastructure to reach their potential, presenting a compelling opportunity for a global CRE fund focused on “Exponential Industries & Advanced Infrastructure.”

This thematic strategy prioritises flexible investment across cutting-edge sectors with unique property needs – from semiconductor fabs and life science labs to AI data centres and spaceports – while avoiding areas that rely only on generic, oversupplied real estate (e.g. traditional offices or standard warehouses).

By aligning with fast-evolving industries and their infrastructure requirements, the fund can capitalise on premium yields, resilient demand, and long-term defensibility.

The approach emphasises adaptability: rather than anchoring to a fixed sector taxonomy, it focuses on the underlying drivers (tech innovation, R&D intensity, policy support) that cut across sectors, ensuring the portfolio can pivot as technologies converge and new industries emerge.

Below, we detail the investment thesis in five parts:

Filtering McKinsey’s arenas to prioritise those with specialised CRE needs.
Defining “exponential industries” and adjacent sectors beyond McKinsey’s list.
Justifying a thematic flexible strategy over a static sector approach.
Outlining the specialised real estate requirements of priority industries.
Quantifying the global market opportunity for such specialised CRE over the next 10–15 years, including expected demand growth and superior investment characteristics.

1. Prioritising High-Impact Arenas Requiring Specialised Real Estate

Not all of McKinsey’s 18 arenas of the future are equal from a real estate perspective. We focus on those arenas (or logical clusters thereof) where highly specialised properties are a prerequisite for industry growth. These are the arenas in which cutting-edge facilities – with features like cleanrooms, high power density, advanced logistics, or proximity to talent hubs – form a critical bottleneck or enabler for the sector.

Conversely, arenas that primarily use generic offices or warehouses (or those already over-served by existing real estate) are de-emphasised, except as part of broader mixed-use ecosystems.

Below we categorise the arenas of tomorrow:

Category
CRE Intensity
Investment Priority

Semiconductors – e.g. chip fabrication
Extremely high: Requires semiconductor fabs with cleanrooms, ultra-clean power, water and chemical handling, plus proximity to supply clustersHighest – core focus (advanced manufacturing hubs)

AI Software & Services – incl. AI infrastructure
High: Requires large-scale data centres and edge computing sites with massive power and cooling for AI training.
Highest – core focus (digital infrastructure)

Cloud Services – hyperscale & cloud computing
High: Data centres (often overlapping with AI needs); network fibre connectivity hubs.
High – (digital infrastructure, established but growing)

Cybersecurity (services)
Low: Largely office-based or cloud-based; no unique property beyond secure ops centres (can piggyback on data centre/office).
Low – deprioritised (generic office/IT)

Robotics (industrial robotics, automation)
High: R&D and assembly facilities with high ceilings, heavy floor loads, robotics testing labs. Often co-located with manufacturing or warehouse space designed for automation.
High – core focus (advanced manufacturing)

Autonomous Vehicles (shared self-driving fleets)
Medium: Testing tracks, sensor-laden infrastructure, and fleet depots with charging; many needs overlap with robotics and EV infrastructure.
Medium – selective (as part of mobility hubs)

Electric Vehicles (EVs) – incl. batteries
High: Gigafactories for batteries and EV assembly plants, requiring large sites, high power, and skilled workforce proximity; also charging network infrastructure.
High – core focus (transport manufacturing)

Batteries (Advanced energy storage tech)
High: Battery cell manufacturing facilities (similar to EV needs), with chemical handling and clean environments; co-location near auto or grid hubs.
High – core focus (energy infrastructure)*

Future Air Mobility (e.g. eVTOL, drones)
Medium: Vertiports and charging hubs in urban areas; airspace integration. May use retrofitted rooftops or new helipad-like structures. Early-stage, but infrastructure will be needed in cities and airports.
Medium – emerging focus (city infrastructure)

Space (commercial space launch, satellites)
High: Spaceports and rocket test facilities (special siting – remote/coastal), satellite assembly cleanrooms, control centres. Few global sites exist – significant greenfield development potential.
High – selective (as part of “NewSpace” infrastructure)

Industrial & Consumer Biotechnology (non-medical)
High: Biotech production labs (fermentation plants, bio-manufacturing) requiring sterile environments, specialised HVAC, and often near research universities. Clusters in biotech hubs.
High – core focus (life sciences manufacturing)

Drugs for Obesity & Conditions (biopharma)
High: Life science R&D labs and pharma manufacturing facilities (GMP compliant) – e.g. lab-office campuses, cleanrooms for biologics. Benefit from established life science clusters.
High – core focus (life sciences R&D)

E-commerce (online retail)
Medium: Large distribution centres and automated warehouses. While e-commerce drove huge warehouse demand, much of this is standard big-box industrial space. Supply has expanded rapidly, creating oversupply in some regions. Generic warehousing yields are under pressure.
Low – deprioritised standalone (except as part of mixed logistics campuses)

Digital Advertising (ad-tech)
Low: Generic office space for tech workers; computation on cloud (no unique real estate). No specialised footprint – essentially part of the digital economy supported by data centres.
Low – deprioritised

Streaming Video (media streaming)
Low: Offices and existing data/network infrastructure (CDNs). Content production uses studios (a niche CRE segment) but not core to streaming platforms themselves.
Low – deprioritised

Video Games (gaming industry)
Low: Offices for developers and generic server space. No special CRE beyond perhaps creative studio space.
Low – deprioritised

Digital Services (Consumer Internet) – e.g. segments of “consumer internet” arena
Low: (Similar to above digital categories – primarily office or standard tech space; relies on generic cloud infrastructure.)
Low – deprioritised

Modular Construction (off-site prefab)
Medium: Factories for modular building components. While this is manufacturing, it uses standard industrial facilities (warehouses adapted for construction assembly). Some specialised needs (heavy cranes, large bays) but generally can use or retrofit existing industrial stock.
Medium – opportunistic (as part of broader industrial portfolio)

Nuclear Fission Power (next-gen fission reactors)
High: Sites for small modular reactors (SMRs) or advanced reactors. Requires significant land, regulatory compliance, and robust infrastructure (cooling water, grid connection). Real estate is highly specialised (often government-partnered).
Medium – selective (depends on policy and partner opportunities)

Key priorities – Based on the above, the fund would prioritise arenas in advanced manufacturing, life sciences, and digital infrastructure: sectors like semiconductors, biotech/pharma, robotics/automation, EVs/batteries, and AI/cloud infrastructure all exhibit high specialised CRE intensity and strong growth drivers.

These industries cannot scale without new state-of-the-art facilities – for example, semiconductor and battery plants have seen a dramatic surge in construction, with U.S. manufacturing construction spending doubling since 2021 due to chip fabs and gigafactories being built under government incentives . Such facilities often cannot be met by existing stock and must be built new, as they have “highly specialised requirements not easily met by existing buildings” .

This supply-demand gap creates an opening for investors to develop or repurpose real estate tailored to these needs.

By contrast, arenas like digital advertising, streaming, or gaming are booming industries but primarily digital in nature – their growth does not hinge on specialised buildings. A video game studio or ad-tech firm can reside in a standard office (of which there is ample supply in the market – indeed global office vacancy hit a record ~16% in 2023 due to remote work). Investing in generic offices for “hot” digital sectors would thus not yield differentiated value – in fact, it could expose the fund to oversupplied asset classes with secular headwinds (e.g. traditional offices). Similarly, e-commerce logistics, while a major growth driver for industrial real estate, has in many areas become a victim of its own success with vacancies rising due to a wave of supply. Unless targeting a unique niche (such as cold storage or robotics-automated warehouses), plain vanilla big-box warehouses for e-commerce are not a scarce resource.

Integrated ecosystem plays – However, the strategy doesn’t outright ignore the lower-priority arenas. Rather, we deemphasise direct investment in those unless they are part of an integrated development that supports the exponential industries.

For instance, a large innovation campus might primarily feature lab and manufacturing space for life science and robotics companies, but could also include some flexible office or co-working space for digital startups (digital advertising, gaming, etc.) to foster an innovation ecosystem.

Likewise, an advanced industry park might have a logistics centre supporting an e-commerce or AI-driven supply chain as a complement to a cluster of high-tech manufacturing tenants.

In such cases, generic CRE uses are included as secondary components to enhance the value of the overall cluster (providing amenities, services, or vertical integration), rather than standalone investment themes.

In summary, our filter produces a concentrated focus on “hard tech” arenas – those where the physical infrastructure is as cutting-edge as the technology it houses – and places on the back-burner the “asset-light” arenas that ride on existing real estate. This ensures the fund’s capital is deployed where it is most indispensable and where competition from generic landlords is minimal.

2. Defining Exponential Industries and Emerging Sectors Beyond the List

The term “exponential industries” in our theme refers to sectors characterised by rapid, compounding growth driven by technological innovation – essentially, industries on the steep part of the S-curve. Key hallmarks include:

High R&D Intensity and Innovation Rates: Exponential industries invest heavily in research and embrace frequent technology upgrade cycles. They often feature step-change innovations in business models or technology – what McKinsey calls “technological or business model step changes” that are one ingredient of an “arena-creation potion” . Examples include AI (breakthroughs in deep learning), advanced materials (nanotech), or biotech (gene editing) – each of these can upend previous limits and unlock new markets.

Escalatory Investment and Capital Formation: These sectors see surging investment as firms race to build capacity or capture leadership (another arena ingredient noted by McKinsey ). Think of the current “AI arms race” or the battery manufacturing boom – companies plough in capital despite short-term losses, creating self-reinforcing growth cycles . This often yields fast growth but also high dynamism (market shares shuffle quickly as new entrants can displace incumbents).

Advanced Infrastructure Requirements: Exponential industries typically push the limits of existing infrastructure, necessitating new types of facilities or networks. Their operations may be mission-critical 24/7 (for example, data centres for digital services must run continuously ) and they often have extreme demands (power, precision, etc.) that standard facilities can’t accommodate. The need for specialised real estate – whether it’s a climate-controlled lab, a high-performance computing centre, or a high-throughput manufacturing line – is a common thread.

Policy Support and Strategic Importance: Many exponential sectors align with government priorities (for economic competitiveness or societal needs), attracting subsidies, favourable regulation, or public-private partnerships. Recent examples include the CHIPS Act for semiconductors, the Inflation Reduction Act for clean energy, and various national strategies for AI or biotech. This support lowers risk and increases the scale of what these industries can achieve (and by extension, the scale of facilities they will build).

Large and Growing Addressable Markets: By definition, an exponential industry targets a substantial market opportunity that can support its growth ambitions . Often they create new markets or radically expand old ones – e.g. the rise of electric mobility expanding the market for batteries, or precision medicine expanding healthcare markets. This ensures that investments in capacity (factories, labs) aren’t in vain; demand is expected to catch up or exceed supply.

Using these criteria, we not only encompass the 18 McKinsey arenas but can also identify adjacent or emerging sectors that fit the profile yet weren’t explicitly listed by McKinsey. Some such sectors on our radar include:

Quantum Computing and Quantum Technologies: Quantum computing is on the cusp of moving from lab research to industry. It squarely meets the exponential criteria: R&D heavy (countries and companies investing billions into quantum research), frequent tech breakthroughs (e.g. higher qubit counts, new quantum error correction methods), and infrastructure intensity (quantum computers require extreme conditions – cryogenic refrigeration, vibration isolation, etc., meaning custom-built lab facilities). Policy support is evident (national quantum initiatives in the US, EU, China), and the addressable market (revolutionising drug discovery, cryptography, etc.) could be enormous.

CRE Opportunity: Quantum research hubs – specialised facilities with shielded, low-temp environments – and eventually quantum data centres. If quantum hardware scales, one can envision quantum computing centres akin to today’s supercomputing centres, potentially co-located with universities or major tech clusters. Few such facilities exist today, pointing to a future niche for development.

AgriTech and Controlled Environment Agriculture: With food demand rising and sustainability concerns, there’s an emerging sector around high-tech farming – e.g. vertical farming, hydroponics, and lab-grown foods. These ventures blend biotech, robotics, and AI – for example, vertical farms use automated systems in warehouse-like structures to grow produce year-round. They tick the boxes of exponential industries: innovative tech (from LED lighting to AI crop monitoring), heavy upfront investment, policy interest in food security, and specialised real estate (custom-built indoor farms with climate control). The addressable market (global food and agriculture) is huge, though this sector is still nascent in proving its economics.

CRE Opportunity: Converted or new-build “agri-factories.” There’s growing interest in repurposing urban warehouses or building greenfield facilities for vertical farms. These require high ceilings, extensive HVAC and water systems, and proximity to city markets (for fresh delivery). While some early ventures struggled, the sector continues to evolve, and a flexible strategy allows optionality to invest when technologies mature – for example, if a particular vertical farming model achieves profitability, the fund could finance its real estate rollout as part of the exponential industries portfolio.

Fusion Energy and Advanced Nuclear: In addition to fission (which was on McKinsey’s list), the pursuit of nuclear fusion power is an archetype exponential bet – high R&D (decades of research nearing potential breakthroughs), massive prize (virtually limitless clean energy), and requiring colossal specialised infrastructure (experimental reactors and, in success, commercial fusion power plants). Several private companies aim to demonstrate viable fusion in the 2030s, supported by public programs (e.g. US DoE fusion milestone grants, UK’s STEP program).

CRE Opportunity: Positioning for the long term, the fund could target partnerships for building sites for fusion pilot plants or components factories. This would be a higher-risk, longer-horizon play (fusion is uncertain), but it exemplifies the benefit of a flexible thematic approach – we can allocate a modest portion to “moonshot” infrastructure that, if it succeeds, would generate outsized returns and cement the fund’s reputation in truly advanced infrastructure.

Other Cross-Sector Convergence Plays: The convergence of technologies is itself spawning new sub-industries. For example, autonomous logistics (drones and last-mile delivery robots) is at the intersection of robotics, AI, and e-commerce. Smart cities infrastructure (sensors, IoT networks, 5G small cells) sits between telecom and urban development. While individually these might be smaller niches, they collectively reinforce the need for flexible investment mandate – the fund can allocate to emergent themes as they arise, provided they meet our criteria of high growth potential and specialised physical infrastructure needs.

By defining “exponential industries” in terms of these fundamental characteristics rather than a static list, we ensure our scope remains open-ended. Today’s list of 18 arenas is a starting point, but history shows that entirely new industries can emerge within a decade. (For instance, ten years ago, industries like commercial space launch or AI-as-a-service were barely on investors’ radar; today they are multi-billion dollar arenas.) Our strategy is to be forward-looking and opportunistic in capturing these adjacencies. If a sector demonstrates the exponential DNA – say, a sudden breakthrough in food-tech or metaverse hardware that creates new demand for specialised real estate (e.g. mega data rendering farms or VR experience centres) – the fund’s theme allows us to pivot and invest there, whereas a rigid sector-specific fund might miss the boat.

In essence, “Exponential Industries” is deliberately defined by attributes and outcomes (rapid growth, innovation, infrastructure intensity) rather than narrow industry definitions. This keeps the investment universe expansive and adaptable, encompassing both the known high-growth arenas and the “unknown unknowns” that may arise.

3. Strategic Rationale: Flexibility Over a Fixed Taxonomy
Anchoring the fund to a broad thematic of Exponential Industries & Advanced Infrastructure – rather than, say, picking a fixed subset of sectors (like “life sciences, data centres, and logistics only”) – provides crucial strategic flexibility. Given the pace of technological disruption, an agile approach is essential for long-term outperformance. Key justifications for this flexibility include:

Adaptation to Technological Disruption: We are living in an era where disruption is accelerating across multiple fronts, and yesterday’s niche can become tomorrow’s necessity (and vice versa). A static sector-based strategy could quickly become obsolete if a chosen sector falters or a new one rises. By contrast, a thematic approach can “stay agile and rapidly adapt” to structural changes. For example, if advancements in biotech automation lead to lab layouts changing, we can adjust our development specs; if a currently unforeseen industry (like quantum computing) suddenly takes off in 5 years, we can include it under our theme without needing a mandate change. This agility is echoed in investment commentary:

“Disruption is accelerating, making thematic investing an essential tool for investors to stay agile and rapidly adapt” .

Cross-Sector Convergence: Many exponential technologies are converging, blurring the lines of traditional sectors. AI is not a siloed sector; it’s permeating healthcare, transportation, manufacturing, and more. Similarly, advances in materials or energy storage impact multiple industries.

A flexible theme allows us to invest along these convergence points. For instance, consider autonomous electric vehicles – this sits at the intersection of automotive, AI, and clean energy (EV batteries). A rigid fund might struggle: Is that “transport” or “tech” or “energy”? For us, it squarely fits the exponential industries theme and we can invest in, say, an R&D campus or a test facility for autonomous EVs as easily as in a battery plant. The modern economy is defined by such “digital and physical worlds fusing, accelerating adoption curves and compelling companies to rethink investments”, as noted by Global X analysts . Our strategy inherently embraces this fusion. We ignore artificial boundaries (like NAICS codes or traditional sector silos) and instead focus on the underlying drivers and needs.

Risk Management Through Diversification of Innovation: While all our target sectors are high-growth, they will not all succeed or cycle at the same time. A flexible basket of themes provides diversification within the high-tech universe. For example, if one arena (say, streaming media) saturates or faces a downturn, another (say, biotech manufacturing) might be in an upswing.

The fund can rotate emphasis accordingly.

This is analogous to multi-theme investment strategies in equities, which use momentum or conviction to tilt towards currently outperforming themes.

Similarly, our portfolio construction can dynamically allocate capital to sub-themes that show the strongest indicators (demand, rent growth, government support) while pulling back on those that cool off. The overarching exponential theme ensures we remain in growth areas, but we are not married to any single one if fundamentals shift.

Capturing “Ecosystem” Opportunities: A strict taxonomy (e.g. an “AI Real Estate Fund” or a “Life Science Fund”) might limit investments to properties that neatly fit one label. In reality, innovation happens in ecosystems – think of innovation districts that host academia, startups, mature companies, and supporting infrastructure in one place.

Our broad mandate means we can create or invest in mixed-use innovation hubs that encompass multiple exponential industries.

For instance, a large campus could have a data centre (AI/cloud), a lab building (biotech), a prototype advanced manufacturing facility (robotics), and even an attached accelerator or office for software startups – all elements reinforcing each other.

A flexible fund can underwrite such a project holistically. The benefit is synergy: tenants from different exponential sectors co-locate to collaborate (AI scientists working with biologists, etc.), driving demand for our space and creating a vibrant, defensible asset. We’re not forced to choose one narrow use; we can provide the full stack of infrastructure that tomorrow’s industries need.

Long-Term Megatrend Alignment vs. Short-Term Fads: The thematic approach keeps us focused on long-term megatrends (e.g. automation, digitisation, decarbonisation, demographic shifts in health) rather than short-term fads. We use the theme as a north star but remain flexible on execution. For example, the megatrend is that healthcare and tech are converging – within that, today it might manifest as demand for biotech labs; tomorrow it could be facilities for personalised medicine production or AI-driven diagnostics centres. We don’t want to be stuck only investing in “wet lab space” if in a decade the key need shifts to a different type of life science facility. By articulating our strategy as “Exponential Industries (with Advanced Infrastructure)”, we send the message to investors that we’ll pursue whatever sub-sectors best deliver exposure to those megatrends at any given time. This mitigates the risk of being in the right church but the wrong pew, so to speak.

Greater Resilience to Disruption: Ironically, a fund targeting disruptive industries must itself be designed to not be disrupted. If we pigeonholed ourselves (e.g. a pure-play “data centre REIT”), we could be upended by an unforeseen technological shift (what if, hypothetically, future computing shifts to a completely new paradigm requiring different facilities than today’s data centres?). Instead, with breadth, we can evolve our asset mix. In practice, this could mean if decentralised computing (edge devices) reduces the need for giant central data centres, perhaps it increases the need for smaller edge server locations embedded in warehouses or cell towers – which we could pivot to, because our mandate is broad enough to treat it as still “AI infrastructure.” Essentially, optionality is built in.

In summary, the benefits of adaptability far outweigh any perceived focus dilution. We maintain coherence by the unifying idea of exponential growth industries, but we avoid the trap of false precision in defining what the future will look like. The world’s top companies today largely were born from agilely riding multiple waves of innovation – our fund mirrors that ethos, aiming to be “unconstrained in capturing innovation, ignoring traditional sector limitations”.

This flexibility is not lack of strategy; it is the strategy – to always be where technological growth and specialised real estate demand intersect, regardless of how the labels evolve.

4. Specialised CRE Requirements of Priority Industries

A cornerstone of the thesis is that many exponential industries have non-negotiable real estate and infrastructure needs that differ markedly from conventional property. These needs create high barriers to entry but also attractive supply-demand dynamics for those who can fulfil them. We outline the nature of specialised CRE required for the key industries we’ve prioritised:

Semiconductor Fabs and Advanced Electronics Manufacturing: The semiconductor industry’s fab facilities are among the most complex buildings ever constructed. They feature large cleanrooms (often tens of thousands of square meters) that must maintain extremely low levels of particles and vibration. These fabs require specialised MEP (Mechanical/Electrical/Plumbing) systems: for example, massive high-capacity power transformers and HVAC systems to support tools that consume tens of megawatts, as well as ultra-pure water and chemical delivery systems.

The cost to build a state-of-the-art fab can exceed $10 billion, and location is influenced by access to skilled labour (engineers), a stable power grid, and often government incentives.

Key features: Class 10 or better cleanroom standards, 100% backup power, air re-circulation systems, and often on-site process waste treatment.

ESG note: Fabs are energy and water intensive, so newer projects incorporate sustainability measures (recycled water loops, renewable energy sources) to reduce carbon and ensure environmental compliance.

Investment implication: These facilities tend to be built custom for a specific operator (e.g. TSMC or Intel). As a landlord, opportunities lie in partnerships or joint ventures (for example, providing capital and development expertise in exchange for partial ownership or long-term lease from the operator). Once occupied, tenants are extremely sticky due to the huge capital invested in equipment and the impracticality of relocation – a chip plant cannot move without halting production for months. This stickiness can translate to stable, long-term income.

Life Sciences R&D and Bio-manufacturing Space: The life sciences sector (biotech, pharma, medical research) relies on a spectrum of specialised properties:

Wet Lab R&D Facilities: These are laboratory-office hybrids, often multi-storey buildings in innovation clusters (e.g. Cambridge (UK), Boston, San Francisco). Labs require reinforced structures to carry heavy equipment, enhanced ventilation and air filtration (fume hoods, 100% outside air systems), and piped utilities (lab gases, deionised water) that typical offices lack. Ceiling heights and floor-to-floor distances are larger to accommodate ductwork. Additionally, as science evolves, labs need flexibility – modular layouts that can be reconfigured, and increasingly, the ability to house computational research suites (for AI-driven drug discovery). Indeed, the rise of AI in pharma is “expanding the sector’s real estate strategy” – facilities now must plan for “onsite AI labs” with power-hungry computational nodes and secure high-capacity data storage alongside traditional wet labs . This means higher electrical loads and cooling capacity in lab buildings, blurring lines with data centre specs. Cluster synergy is important: lab buildings thrive in clusters near universities or hospitals, and often we consider not just the building but the campus (shared amenities like conference centres, collaboration spaces, or an incubator wing for startups).

Biomanufacturing & Pharma Production: These facilities (for example, cell and gene therapy production suites, vaccine manufacturing plants, or pharma pill production lines) take specialised to another level: they must meet stringent regulatory standards like GMP or even cGMP (current Good Manufacturing Practice) . They often contain cleanroom production areas (though typically less strict than semiconductor cleanrooms, still with controlled environments), specialised process equipment (bioreactors, filtration systems), and backup systems to protect valuable biological product in process (e.g. emergency power and refrigeration). Many are single-story large floorplate buildings in suburban tech parks or industrial zones. Zoning can be an issue due to handling of biological materials – sites often need appropriate zoning for light manufacturing and sometimes isolation from sensitive neighbours. According to CBRE, “Biomanufacturing facilities require unique quality assurances, infrastructure, architecture and zoning”, and developers have used strategies like build-to-suits and conversions to create them. They also tend to cluster in specific regions with talent and funding, such as Boston, Philadelphia, Basel, or Singapore.

Investment implication: Demand for these facilities is rising as more biotech therapies reach clinical approval stages – CBRE notes “surging demand for specialised bio-manufacturing facilities” in the cell/gene therapy era . These tenants (often big pharma or well-capitalised biotech) invest heavily in customising the space, leading to long leases and sticky occupancy. As one industry expert noted, “life sciences tenants tend to be ‘sticky’ because they are well-capitalised companies that invest a lot of capital [in their labs]” . We expect above-average tenant retention and the ability to command premium rents, especially when supply is constrained by the complexity of developing new lab space.

Data Centers and AI Computing Infrastructure: Modern digital industries require physical backbone in the form of data centres – effectively the “factories” of the digital age. These are typically mission-critical facilities with high power density, robust cooling, and ultra-reliable uptime. There are a few sub-types:

Hyperscale Data Centers: Large (often 100,000+ sq ft) server farms, usually in lower-cost areas or campuses, serving cloud providers or large enterprises. They demand huge power capacity (often 20–50+ MW per facility) and often significant water for cooling (though air and liquid cooling tech is evolving).

Special requirements: dual or triple power feeds, multiple backup generators, battery UPS systems, and cooling plants (chillers, cooling towers or economizer systems). Many hyperscalers also now design for AI hardware, which is even more power-dense (racks filled with GPUs/TPUs). Goldman Sachs projects that by 2030, power demand from data centres will increase 165% due to AI workloads . This is causing occupancy in data centres to tighten – forecasted average utilisation rising to >95% in the next couple of years . In some markets, power availability and grid constraints are the limiting factor – e.g. Northern Virginia (the largest data centre market) is seeing project delays due to grid upgrades needed . ESG and location: Data centres are energy hogs, so access to renewable energy is increasingly a factor; also community pressure can arise over water use and noise (from cooling equipment), influencing design and sometimes favouring certain geographies.

Edge Data Centers: Smaller facilities located closer to end-users to reduce latency (important for applications like autonomous systems, AR/VR, etc.). These might be 5,000–20,000 sq ft modules placed in urban areas or at network hubs. They still need strong power and cooling, but scale is smaller – could be within base floors of buildings or modular units on telecom sites. As industries like robotics, autonomous vehicles and smart cities grow, edge computing sites near city centres, highways, or cell tower aggregations become more important (for quick data processing).

Investment implications: Data centres have become a distinct real estate asset class with institutional interest due to their strong cash flows. Tenants often sign long leases (5-15 years) with significant upfront fit-out costs, which makes them sticky. In particular, “hyperscale data centre tenants also tend to be stickier due to the significant investment they make to outfit the property, and relocating the critical infrastructure can be challenging” – leases average ~10 years with extension options and built-in rent escalators . This stickiness and built-in growth translate to reliable income. Furthermore, current market dynamics are landlord-favorable: in major markets, vacancy rates are at record lows (≈3–5%) and rents are rising sharply (e.g. US data centre rents +19% year-on-year amid scarce supply ). This is in stark contrast to traditional office or retail. As AI adoption soars, one can reasonably project robust demand for new data centres globally for years to come. JLL predicts 10 GW of new data centre capacity will break ground globally in 2025 alone. Each megawatt of capacity roughly equates to ~$10 million in development cost, underscoring the scale of investment. Our focus would be on both developing new facilities in high-demand markets (often in partnership with operators) and potentially acquiring existing data centres with expansion potential.

Special considerations: We must navigate power procurement, fibre connectivity, and in some cases local moratoria (e.g. some cities have paused data centre approvals due to power strain). We also would incorporate sustainability (modern designs aim for PUE – Power Usage Effectiveness – close to 1.2 or better, and some use renewable energy or innovative cooling like heat reuse).

Advanced Manufacturing & Robotics Facilities: Aside from semiconductors (covered) and bio-manufacturing (covered), there is a range of next-gen manufacturing that includes EV assembly plants, battery gigafactories, aerospace/space tech plants, and robotics assembly and testing centres. These typically share needs such as:

• Scale and Layout: Large single-storey structures (for assembly lines) or multi-storey with freight elevators (if footprint constrained). High bay ceilings (to accommodate equipment or cranes), extensive floor load capacity (for heavy machinery). Open spans (few columns) to allow flexible reconfiguration of production lines as technology evolves.

• Power and Utilities: High electrical loads to run automation, robotics, and in some cases high-heat processes (like battery cell production ovens). A battery cell factory, for instance, not only needs huge power but also special HVAC to maintain dry rooms (low humidity) for lithium handling. Similarly, aerospace factories may need built-in testing rigs or wind-tunnel facilities. Robotics R&D sites might include dedicated labs with their own power and data setups.

• Connectivity and Tech Integration: These “Industry 4.0” facilities are bristling with IoT sensors and often require robust on-site data infrastructure (edge computing) to handle real-time data from machines. So they might have a small data centre on-site or at least advanced IT rooms. Additionally, if autonomous vehicles or drones are being tested, the site might include dedicated tracks, airspace permissions, or RF communications infrastructure (e.g., private 5G networks on-site for robot communication).

• ESG and Worker Amenities: Modern advanced manufacturing sites differentiate themselves with sustainability and employee-friendly design – solar panels on vast roofs, energy-efficient lighting, water recycling (especially for batteries where water usage is heavy), and amenities like training centres or collaboration spaces to attract talent (engineers often work on-site alongside production). For example, many EV and battery plants being built in the US and Europe tout renewable energy integration and community benefits as part of getting local approval.

Co-location with talent & suppliers: These facilities benefit from being in clusters – e.g. automakers building battery plants near their assembly plants to streamline logistics, or robotics companies locating near top engineering universities. From a real estate standpoint, this means there is value in assembling “advanced industry parks” where multiple synergistic operations are neighbours (a battery plant, an EV assembly, a robotics supplier, and maybe a testing track, all in one campus). We would seek to facilitate such clustering, which increases the value proposition for tenants (they gain an ecosystem).

Example: The fund could develop an “Advanced Robotics Campus” outside a major metro – comprising high-spec industrial bays for robotics prototyping, a section for drone testing (with a netted enclosure or adjacent open field), and offices for associated software teams – effectively a one-stop hub for autonomous systems development. Such a project might cater to several companies (tenants) and perhaps a shared demo facility. The specificity of the infrastructure (flooring that can accept bolting of robot rigs, indoor GPS systems, etc.) becomes a selling point.

Notably, policy initiatives are heavily boosting advanced manufacturing in many countries, not only providing demand but also funding. For instance, U.S. federal incentives (CHIPS Act, IRA) and similar European programs have unleashed a manufacturing construction boom. This reduces risk for us as investors – public money often complements private capital in these projects (through grants, tax breaks, tenant covenants), meaning our developments can secure anchor tenants with support already in place.

New Mobility Infrastructure (EV Charging, Vertiports, etc.): Supporting the exponential growth in electric and autonomous mobility will require new or upgraded infrastructure integrated with real estate:

• EV Charging Hubs: While individual charging stations are small-scale, we foresee the rise of large EV charging forecourts or “charging lounges” on highway corridors (akin to petrol stations of the future) and in urban centres (perhaps integrated into parking garages). These have real estate plays – e.g. redeveloping portions of underused parking lots to host fast-charging pods with amenities for drivers. Power access is the key constraint – sites might need their own substations or energy storage (which itself could be a real estate asset: battery storage facilities to buffer the grid).

• Autonomous Vehicle Depots: If autonomous taxis or delivery vehicles proliferate, they will need fleet depots for maintenance, cleaning, and overnight charging. These depots differ from today’s car garages by being optimised for automation – yard layouts that accommodate self-parking, wireless charging pads, and advanced telematics. Real estate wise, these could be retrofitted warehouses or new builds in logistic zones at city peripheries. They may not be huge drivers of rent, but as part of a broader portfolio, having some mobility infrastructure assets can complement an ecosystem (for instance, an autonomous vehicle depot next to a robotics industrial park).

• Vertiports for Air Mobility: Companies developing electric vertical takeoff and landing (eVTOL) aircraft (air taxis) anticipate the need for networks of “vertiports” in and around cities. These are essentially small-scale airports – a typical vertiport might consist of a takeoff/landing pad or two, a passenger terminal or lounge, charging stations for the eVTOLs, and maintenance hangar space. They could be on rooftops of large buildings, on repurposed parking structures, or on piers/greenfield sites at city edges. CRE opportunity: While experimental now, if air mobility takes off, owning strategic vertiport locations in major cities (either directly or via partnerships with infrastructure players) could be valuable. These would operate somewhat like airport real estate (with revenue from takeoff fees, retail in terminal, etc.).

Space Infrastructure Facilities: As the commercial space sector grows (satellite constellations, space tourism, etc.), it drives demand for terrestrial facilities:

• Launch Sites / Spaceports: Historically government-run, now there is a trend towards commercial spaceports (e.g. in Cornwall UK, or Florida’s commercial pads) that cater to private launch companies. Real estate aspects: large land area, regulatory clearances, integration with logistics (to bring rockets and fuel), and range safety infrastructure. A fund likely wouldn’t own a whole spaceport (governments often involved), but could invest in components like the integration facilities or visitor centres around them.

• Assembly and Test Facilities: Rockets and spacecraft require huge assembly buildings (think SpaceX’s giant hangars in Boca Chica or Blue Origin’s factory in Florida). Satellite makers too need cleanrooms for assembly and testing (often in industrial parks). These are akin to specialised manufacturing discussed above. Given the niche nature, any such investment would probably be tenant-specific (built for a particular company under long lease).

• Downstream Space Data Centers: A subtle one – as thousands of satellites launch (for earth observation, communications), there’s growth in ground stations and data centres that receive and process satellite data. These sometimes are in remote areas (to catch satellites overhead) but networked to central facilities. It’s a convergence of space and data infrastructure.

Overall, the specialised CRE needs can be summarised as infrastructure-intensive, high-specification, and often requiring co-location with talent or resources. They are not commodity buildings; each often requires tailoring to industry-specific standards. This complexity means fewer developers deliver them, contributing to supply inelasticity. As JLL observed about advanced industries like manufacturing and data centres: “All have highly specialised facility requirements not easily met by existing stock… new construction is often a necessity… these sectors compete for the same specialised trade services”. For investors, this means that when you deliver a well-executed specialised asset in a growing market, you face limited competition and can attract top-tier tenants who have few alternatives, enabling premium rents and long leases.

It is worth emphasising the infrastructure intensity and ecosystem dependency of these assets: success is not just about the building, but its integration into a wider network (power grids, transport links, talent pools). For example, a data centre must have not only its structure, but also guaranteed power supply (sometimes involving co-investing in substations) and fibre connectivity. A life science campus benefits from being near a university (perhaps involving partnerships or providing incubator space to start-ups spinning out of that university). Thus, our approach to development and acquisitions will often involve a broader development plan and stakeholder collaboration (with utilities, local governments, universities, etc.). This adds complexity but also defensibility – once these pieces are in place, it’s very hard for a competitor to replicate a similar integrated environment quickly.

5. Market Opportunity Size and Investment Outlook (10–15 Year Horizon)

The global addressable market for specialised CRE serving exponential industries is immense and set to expand dramatically in the next decade. We can quantify this opportunity by looking at both the demand side (the growth of the industries themselves, which drives facility needs) and the supply side (current stock vs. required stock of suitable real estate):

Share of the Economy & Growth Trajectory: The 18 arenas of tomorrow are projected to increase their share of global GDP from ~4% today to between 10% and 16% by 2040 . That implies these sectors will triple or quadruple in relative size. In absolute terms, as noted, their revenues could reach ~$29–48 trillion by 2040 , with $2–6 trillion in profit. Even half of that growth occurring in the next 10–15 years will mean trillions of dollars in new economic output needing physical space. For context, the earlier generation of “arenas” (2005–2020 period) saw their GDP share triple and they captured half of global corporate profit by 2019 – and those included e-commerce, cloud, etc., which spurred an unprecedented wave of CRE development (warehouses, data centres, etc.). The coming wave could be even larger . This translates to an unprecedented building boom for advanced infrastructure.

Volume of Investment in Facilities: We are already witnessing massive capital commitments to build out these industries:

• Data Centers: A Boston Consulting Group analysis finds that industry players “are readying a massive deployment of capital – $1.8 trillion from 2024 to 2030 – to meet data centre demand” . This figure (if even roughly accurate) is staggering; it likely includes the entirety of global data centre-related capex. Even a portion of that available to real estate investors (land, shell construction, power infrastructure) represents a multi-trillion dollar asset creation. By 2030, the global data centre market is expected to reach ~$650 billion annually , growing ~11% CAGR. Importantly, near-term tightness means occupancy and rents will remain strong – Goldman Sachs projects data centre utilisation peaking >95% by 2026 before new supply catches up. High occupancy plus high investment typically yields high NOI growth for existing assets (we’re already seeing double-digit rent increases ). Compared to traditional CRE classes like office or retail which are struggling with high vacancies and modest growth, data centres offer both growth and income stability – an attractive combination.

• Life Science Real Estate: In the U.S. alone, the top life science clusters had over 40 million sq. ft. of lab space under construction or in planning as of 2024. Global investment into life science properties has risen sharply in recent years (in Europe, life science real estate is now considered “the asset class of the future”, attracting new investors ).

While there has been a short-term oversupply in some U.S. markets (vacancies ticked up to ~18–20% as a flood of new space delivered during a brief slowdown in biotech funding ), the long-term demand drivers – aging population, healthcare R&D spending, biotech innovation – point to robust absorption ahead. Indeed, life science employment is at record highs and big pharma is flush with cash for acquisitions and expansions, which will require more lab and manufacturing space.

We anticipate the current inventory will be digested by expanding occupiers and that by late-decade new waves of demand (e.g. for AI-integrated labs as discussed) will spur further development. Scale perspective: A single major pharma might lease a 500k sq. ft. R&D centre; a big biotech manufacturing campus can be 1–2 million sq. ft.

We expect dozens of such large facilities globally. The global life science real estate market size, though hard to pin down, is in the hundreds of billions (and growing at double-digit CAGR according to market analyses). For example, an industry report projects global life science real estate could reach on the order of $100+ billion by the early 2030s (though figures vary widely).

What matters is the relative growth: few other CRE segments have the combination of low vacancy (historically) and secular growth that life science does. And notably, life science tenants often pay a premium for quality space (in some clusters, lab rents are 2–3x Class A office rents) and sign longer leases, boosting yield potential.

• Industrial/Manufacturing Space: The reindustrialisation trend (especially in North America and Europe) tied to advanced industries is driving record construction of manufacturing facilities. The US has over $200–300 billion in semiconductor fabs announced or underway, 120+ new battery gigafactories globally by 2030 are needed (with investments well over $100 billion), and EV assembly and supply chain facilities tens of billions more. For instance, just one EV company, Tesla, invested over $5 billion in its Berlin Gigafactory; Ford and SK Innovation committed $11+ billion for twin battery plants and an EV plant in Tennessee/Kentucky. Each of those projects is multi-million square feet. According to SIA, global semiconductor industry capex will total $2.3 trillion in 2024–2032 (not all real estate, but a large chunk is facilities). Bottom line: the next decade will see trillions in advanced manufacturing facility construction worldwide, supported by government incentives and corporate necessity. The opportunity for our fund is to participate in this via development partnerships, sale-leasebacks (many corporates will seek to recycle capital after building – we can buy their facility and lease it back to them), or providing needed infrastructure (like an adjacent supplier park).

• Emerging Areas: For newer segments like space, quantum, etc., the absolute numbers are smaller but growth rates are very high. The space industry is expected to grow to $1 trillion by 2040 (per Morgan Stanley), up from ~$350 billion now – that implies dozens of new facilities (factories, launch pads) will be built. Quantum computing, while still in R&D, has seen >$1 billion in venture funding annually; if it commercialises, we could see specialized data centers for quantum by the 2030s. We treat these as high-upside options.

• Comparison to Traditional CRE Asset Classes: Traditional sectors like office, retail, and even standard residential are growing much slower and in some cases shrinking. Global office demand is stagnating due to remote work (with global office vacancy at ~16% and many markets struggling). Retail has bifurcated, with e-commerce pressure reducing need for physical stores (and much retail CRE being repurposed or facing vacancy). In contrast, the exponential industries present a new frontier for real estate – effectively creating new sub-asset classes (data centres were not a mainstream asset class 15 years ago; now they are). Investor capital is already rotating accordingly: alternative CRE (which includes life science, data centres, etc.) has grown from a niche to a significant portion of institutional portfolios. For example, specialised tech real estate platforms (like life science REITs, data centre REITs) have generally outperformed their office peers in recent years, thanks to higher rent growth and occupancy. Alexandria Real Estate (life science REIT) and Digital Realty (data centre REIT) have seen far higher revenue growth than most office REITs, illustrating the premium yield and growth potential. Cap rates in these sectors have historically been lower (reflecting higher investor demand), yet investors are willing to accept that because of perceived safety and growth – life science and data centre assets in prime markets often traded at 4-6% cap rates in the low-rate environment, versus offices at 5-8%. Even with recent interest rate shifts, the spread in rent growth means total returns can outstrip others. Furthermore, tenant stickiness and long leases in specialised CRE mean cash flow stability is often better than in office or retail where tenant churn is higher. A lab tenant might sign a 10- to 15-year lease to amortise lab fit-out costs; a cloud operator might commit for 10+ years in a data centre . This long-term income can underpin a core-like risk profile, but with growth that rivals opportunistic investments – a very attractive combination for a 10–15 year fund.

• Defensibility and Yields: Because creating these assets requires expertise, the competitive pool of landlords is smaller. If we establish an early-mover portfolio of, say, key life science campuses and AI computing centres in global innovation hubs, we have a defensible position: high replacement cost assets, deeply embedded in tenant operations. The yield premium (in terms of total return) comes not just from initial cash yield but from superior rent escalation and appreciation. Many exponential industry leases have escalators tied to inflation or fixed 3%+ annual bumps, given the lack of alternatives and high tenant investment. There’s also often the possibility of participating in growth through profit-sharing or performance rent in certain cases (for instance, some data centre leases include usage-based components). Even at stabilisation, we foresee these specialised assets achieving cap rates that, when adjusted for growth, provide spread over traditional assets. For example, a data centre at a 5% cap with 3% annual rent growth is far superior in value creation to an office at a 6% cap with flat rents. Additionally, on exit in 10-15 years, we expect strong institutional demand for such assets (given the secular trend, there will likely be even more dedicated funds/REITs wanting to buy stabilised exponential infrastructure assets from us).

To put a rough number: If one tallies the major components – data centres ($1.8T by 2030 in new investment ), manufacturing (several trillions across chips, batteries, EV, etc.), life sciences (hundreds of billions in new labs/factories), plus other infrastructure – it’s plausible that $3–5 trillion of new specialised commercial facilities will be developed globally in the next 10–15 years.

Even capturing a small slice of this market would mean deploying a fund of tens of billions.

For a more concrete addressable sub-market: consider data centres + life science + advanced manufacturing facilities as an investable universe. This already now rivals the size of the traditional office market in some regions.

In the US, for instance, the industrial (much of it advanced manufacturing and logistics) development pipeline is nearly 370 million sq. ft. under construction, and life science under construction is nearly 10% of existing inventory in top markets (which is significant growth). The direction is clear: investors who pivot to these themes are positioning for where growth will be, not where it used to be.

Premium yields and rents: An illustrative example – a biotech manufacturing facility might cost $1,000/sq.ft. to build (due to cleanrooms) and lease for perhaps $100/sq.ft. triple net, whereas a regular warehouse costs $150/sq.ft. and leases for $10/sq.ft. The specialised asset yields a higher absolute rent and often at a similar or only slightly higher capex multiple, meaning higher return on cost if executed well. Similarly, a data center might lease its power capacity at rates that equate to much higher per-square-foot revenues than a traditional property. Tenants pay up because the property enables their billion-dollar business. Moreover, tenants often sign long initial terms with built-in growth, giving the landlord the benefit of compounding rental income – for instance, a 10-year data centre lease with 2% annual escalator means by year 10 the rent is ~22% higher than year 1, all else equal, which significantly boosts yield on original cost.

Finally, defensibility comes from the fact that these properties, once developed, are hard to relocate or replicate. A competitor can’t easily build another semiconductor fab next door without also securing the talent, power, water, and billions of investment; similarly, if we own a prime life science campus integrated into Cambridge’s ecosystem, a new entrant can’t recreate the Cambridge cluster out of thin air. This embedded value provides insulation against competition and economic downturns. In downturns, traditional CRE often sees tenants downsizing or defaulting; in contrast, in something like a data centre or lab, the tenant’s core operations (drug research, cloud services) are so tied to the site that they will cut many other costs before they cut their facility lease – it’s mission-critical space.

In conclusion, the quantitative and qualitative evidence builds a compelling picture: the next 10–15 years will see an unprecedented expansion in specialised CRE to support exponential industries.

This is not a niche side story; it is becoming a central theme of global real estate development. By executing a focused yet flexible strategy now, our fund can become a leading provider (and owner) of the “advanced infrastructure” enabling the industries of the future.

We stand to benefit from high growth, strong income fundamentals, and the tailwinds of both public and private investment flows.

In a world where generic real estate is increasingly challenged by technological shifts (e.g. remote work hitting offices, e-commerce thinning retail), our strategy is to invest in the real estate that technological shifts create rather than the real estate those shifts obsolesce.

This alignment with innovation, underpinned by data-driven insights and strategic agility, is what will drive superior risk-adjusted returns for the fund over the next decade and beyond.

Sources:

• McKinsey Global Institute – “The Next Big Arenas of Competition” (2024) – on 18 arenas of tomorrow (revenues, profit potential, and characteristics of high-growth arenas).

• McKinsey Global Institute – arenas defined by step-change innovations, escalatory investments, large addressable markets .

• Innovation Leader summary of MGI report – arenas dominate profit and growth, mostly in US/China, illustrating scale .

• JLL Research (2024) – “Growing industry sectors face unique construction challenges” – notes that advanced manufacturing, data centers, life sciences have specialised requirements not met by existing stock ; manufacturing construction boom (spending doubled since 2021) driven by policy (CHIPS Act, etc.) ; data centers facing labor and power constraints amid high demand ; life sciences integrating AI, needing on-site compute infrastructure .

• CBRE Research (2023) – “Revolutionizing Biomanufacturing” – highlights clustering and unique needs of cell/gene therapy manufacturing facilities .

• PERE News / Industry commentary – sticky nature of life science tenants (due to high fit-out investment) ; rising power requirements in AI-driven pharma labs (strain on power grid, need for new property capabilities) .

• GI Partners (2021) – Launch of Tech & Science real estate fund focusing on data centers, life science, “always-on” R&D facilities – underscores investor interest in this thematic and the 24/7 mission-critical nature of these assets.

• S&P Global Ratings (2024) – “Data Centers: Risks and Opportunities” – notes 50 GW of new US data center capacity 2023–2028, record leasing in 2024, strong demand from AI ; also that hyperscale tenants sign ~10 year leases and invest heavily, making them sticky .

• Goldman Sachs (2023) – “AI to drive 165% increase in data center power demand by 2030” – quantifies massive growth in data center usage (50% by 2027, 165% by 2030) and tightening supply (occupancy >95% by 2026) .

• BCG (2025) – “Breaking Barriers to Data Center Growth” – cites $1.8 trillion planned global data center capex 2024–2030 .

• Market data via CBRE, JLL: record-low data center vacancy (~3.7% in US) with double-digit rent growth ; U.S. industrial pipeline (369 million sq ft under construction) showing scale of new logistics/manufacturing builds ; Life science construction (~8–9% of stock) in top markets indicating growth despite temporary higher vacancy .

• Axios/Moody’s (2024) – global office vacancy at all-time highs (~20% US, ~16% globally) , reflecting oversupply in generic CRE – a contrast to our targeted sectors where undersupply is more common.

• McKinsey (2022) – Battery 2030 report: need for 120–150 new battery plants globally by 2030 , emphasizing huge factory development.

• Semiconductor Industry Association / McKinsey – $200B+ in semiconductor fabs announced in US alone ; industry on track to $1T market by 2030 requiring many new fabs (Nasdaq analysis) .

• Global X (2025) – “Why Multi-Theme Investing? – highlights need for dynamic rotation and that converging megatrends demand an unconstrained approach .

• (Additional data points are embedded throughout the text in the relevant context.)

Antony Slumbers 11/8/24 Antony Slumbers 11/8/24

The New Glass Bead Game: AI and the Search for Synthesis

In The Glass Bead Game, Hermann Hesse presents an idealised world devoted to intellectualism and synthesis. Set in the fictional province of Castalia, the novel follows Joseph Knecht, a scholar immersed in a sophisticated game that combines music, mathematics, art, science, and philosophy into a single, harmonious pursuit. This “Glass Bead Game” represents the ultimate intellectual challenge, demanding mastery across disciplines and the ability to draw connections between them. Castalian scholars devote themselves fully to this intellectual life, aspiring to reach a level of understanding that transcends individual knowledge and creates a universal language of human achievement.

However, Hesse also warns of the limitations of pure intellectualism. Knecht eventually questions Castalia's detachment from the real world, sensing the dangers of a life solely devoted to abstract thought. This tension—the pull between contemplation and engagement—resonates strongly in our AI-driven world. Like Castalian scholars, we have access to tools that could elevate our understanding, expanding human knowledge and creative potential to new heights. But, as with Castalia, there's a risk of disengagement from real-world challenges if we use AI solely to streamline tasks rather than as a means to enhance our cognitive and creative faculties.

In today’s world, our relationship with AI is often imagined as a “centaur model”—where human and machine work together but remain separate entities, with humans retaining control over the intellectual journey. However, Hesse’s work points toward an even deeper integration. By adopting a “cyborg” mindset, we could think of AI as part of our extended cognitive system, seamlessly blending it into every aspect of our intellectual and practical lives. AI, as a fully integrated thinking partner, has the potential to expand our minds, help us solve complex problems, and push us beyond the limits of our individual intelligence. But this relationship requires that we approach AI not just as a tool for efficiency but as an intellectual companion—one that actively supports and strengthens our ability to think, create, and innovate.

One of the key challenges in achieving this integration is the human tendency toward shortcuts. Rather than using AI to think more deeply, there’s a risk of using it simply to offload cognitive work. This approach undermines our agency, leaving us passively dependent on AI instead of actively collaborating with it. When we treat AI merely as a shortcut, we diminish our own intellectual engagement and risk losing control over our cognitive journey. To truly benefit from AI, we must resist this impulse to delegate thinking and instead embrace AI’s potential to enhance our cognitive abilities, helping us approach problems with greater depth and insight.

To retain agency over AI, we must cultivate a mindset that values curiosity, critical thinking, and human-centered skills. By focusing on problem-solving, analytical thinking, and data-driven insight, we can engage with AI in a way that enhances, rather than diminishes, our cognitive capabilities. This requires a deliberate effort to integrate AI as a collaborator, one that challenges us, refines our thinking, and encourages us to approach the world with curiosity and creativity. In this way, we uphold the spirit of the Glass Bead Game—striving to blend art, science, and philosophy into a unified intellectual pursuit that enriches both human and artificial intelligence.

Ultimately, by approaching AI with a curious and disciplined mindset, we ensure that we retain control over our intellectual journey, creating and curating AI companions that reflect and amplify our deepest values and aspirations. This path allows us to live up to the ideals of The Glass Bead Game—where knowledge, creativity, and human insight converge in a harmonious pursuit of truth and understanding.

Antony Slumbers 10/3/24 Antony Slumbers 10/3/24

Real Estate as Maven

Since 2015 I‘ve been talking about how ‘The Real Estate Industry is no longer about Real Estate’. In the sense that office occupancy, even back then (it’s not just a post Covid phenomena) was hovering around 40-50% and a similar figure stated that their offices failed to enable them to be productive. These statistics, from Leesman, were screaming from the rooftop that the office ‘Product’ needed an upgrade, and it wasn’t just about the real estate.

My talks at the time ended with two slides. One saying ‘The future of real estate is in creating places of inspiration’ and the other ‘The office is dead - long live the Imaginarium’.

This was soon to morph into years of espousing for #SpaceasaService - spaces that ‘however procured provide the spaces and services appropriate to the ‘job to be done’ of every individual, as and when they need it.’

Nearly a decade on, all of this needs an update. And that update is to move beyond thinking about ‘Space as a Service’ to thinking of ‘Real Estate as Maven’.

Within the real estate industry we have a requirement to divine the future. What matters technologically today matters less than what will matter ‘one development lifecycle’ ahead. Given that any significant project can easily take 3-10 years we have to consider how the world might be some way off.

Over the last half century that hasn’t been so hard, as technology has been moving fast (Moore’s Law will be 60 next year) but from a low base. Doubling tiny amounts of computational power doesn’t make that much difference. Today however we are doubling high levels of compute power, so each doubling is suddenly highly meaningful. In fact it is much more meaningful than this because, as Jensen Huang (CEO of Nvidia) stated recently, we are almost running at Moore’s Law Squared at the moment. The power of his companies chips (which are the foundational tool for AI) has increased 1000X in just the last 8 years. The amount of data we have to work with has grown 10X over the last decade, and the size of AI models has increased 10X every year for the last 10 years.

So, in a typical Moore’s Law decade we can expect to end with 100X the computational power that we started with. But over the next decade that might well be 1000X.

This is non-trivial.

Not much stays the same at these rates of change.

What we now need to do is decide what we believe won’t change, what will, and how to accommodate both.

From my own experience I am increasingly convinced that we are moving from a technological Battenberg Cake to an Eton Mess. In the former, ingredients, and characteristics, are neatly defined. This square is X, and that Y. In an Eton Mess all is mixed up. There is not X or Y, there is just Z.

As this relates to how we will work with AI and other advanced technologies, the point is that we are likely to become less Centaurs and more Cyborgs. Instead of us doing A and ‘the machines’ B, we will intertwine our capabilities and do things together.

However, and this feels paradoxical but is true, this means as technology develops exponentially we need to develop our own human skills exponentially. We both need to become ‘more human’ and more at home with working intimately with ‘the machines’.

Because there is an AI Creativity Challenge. Based on current research, AI does improve everyone’s creativity and ability to generate ideas but this comes at the cost of lower variation and novelty. AI makes us all smarter, but in the same way. AI also tends to lead many to be lazy, and unwittingly homogenous. The AI’s ideas or answers are more than ‘good enough’ so one does not need to expend too much cognitive energy. Unwittingly people are producing better ideas but losing their differentiation. So, in effect, becoming dumber.

What is both a necessity, and a great opportunity, is to develop new thinking skills and behaviours whereby we can leverage AI for truly novel creativity. We need to prioritise critical thinking and synthesis skills.

Which also means we have a need for environments that not only help us foster distinct ideas but also actively cultivate our human cognitive abilities. We need to evolve our environments, our education and our working practices to complement AI, not become slaves to it.

Which is the genesis of the "Real Estate as Maven" Concept.

It is ‘an innovative approach to real estate that transforms physical spaces into dynamic, intelligent ecosystems designed to actively facilitate human and organisational success. It integrates cutting-edge technology, adaptive physical environments, and human-centred services to deliver measurable outcomes in productivity, wellbeing, innovation, and connectivity.’

It involves moving beyond passive space provision to active facilitation of human potential. It provides a vision for how real estate can evolve to meet the complex needs of a rapidly changing world, where the lines between physical and digital, work and life, are increasingly blurred.

For instance: "Imagine a product design team working in a Maven space. As they brainstorm, AI systems analyse their conversations and sketches in real-time, providing relevant information and suggestions. However, unlike traditional AI assistants, the Maven's systems are designed to highlight divergent thinking. When it detects that team members are converging on similar ideas, it might introduce unexpected stimuli - perhaps changing the room's lighting to mimic a different environment, or displaying images of nature that are tangentially related to the problem at hand. These subtle interventions are designed to nudge human creativity in new directions, while AI continues to support with data and analysis. The result is a blend of AI-enhanced efficiency and human-driven novelty.

‘Real Estate as Maven’ has 13 core characteristics:

Active Facilitation: Unlike traditional passive real estate, it plays a proactive role in enabling the success of its occupants. It anticipates needs, suggests connections, and provides resources.
Outcome-Focused Design: Every aspect of the space is designed with specific, measurable outcomes in mind, such as enhanced productivity, improved wellbeing, or increased innovation.
Adaptive Environments: Physical spaces are highly flexible and can rapidly reconfigure to meet changing needs and support different modes of work or interaction.
Technology Integration: Cutting-edge technology is seamlessly embedded throughout the space, enhancing capabilities and experiences rather than being a separate component.
Human-Centred Services: A suite of services focused on supporting human needs and potential, from wellness programs to skill development opportunities.
Connectivity Catalyst: The space actively fosters connections between people, ideas, and resources, both within and beyond its physical boundaries.
Continuous Evolution: Built-in systems for gathering feedback and data analytics enable the space to learn and adapt over time, staying ahead of user needs.
Holistic Approach: Considers the whole person and the entire organisational ecosystem, not just work-related activities.
Sustainability Integration: Incorporates sustainable practices and technologies as a core part of its design and operation, not as an afterthought.
Experience Curation: Actively curates experiences (e.g., events, collaborations, learning opportunities) that add value for occupants.
Boundary Blurring: Blends traditionally separate realms such as work, learning, wellness, and social interaction.
Community Building: Promotes a sense of community and shared purpose among its occupants, whether within a single organisation or across multiple entities.
Human Potential Enhancement: Actively works to enhance human cognitive and creative capabilities through environmental design, technological integration, and tailored services.

And these core characteristics might manifest themselves in things like this:

Human-AI Collaboration Hubs:

Spaces designed for humans to work alongside AI systems, with intuitive interfaces for human-AI interaction
Areas equipped with advanced visualisation tools for complex data interpretation and decision-making
"AI sandboxes" where people can experiment with and develop new AI applications

Creativity and Ideation Spaces:

Environments designed to stimulate creative thinking and problem-solving
Tools and technologies that enhance brainstorming and idea development
Spaces that can capture and digitise spontaneous ideas for later refinement

Skill Development and Learning Centres:

Flexible spaces for continuous learning and skill development
Virtual and augmented reality training facilities
Areas for hands-on experimentation and prototyping

Wellness and Human Performance Optimisation:

Spaces and services focused on physical and mental wellbeing
Advanced diagnostics and personalised health optimisation programs
Stress reduction and cognitive enhancement facilities

Cultural and Artistic Expression Zones:

Areas dedicated to showcasing and creating art, music, and other forms of human expression
Spaces that celebrate cultural diversity and facilitate cross-cultural understanding
Facilities for performances and creative collaborations

Community Building and Social Impact Centres:

Spaces designed to foster a sense of community and shared purpose
Facilities for developing and implementing social impact initiatives
Areas that encourage intergenerational interaction and knowledge transfer

Now some of this might sound similar to what you’ve already seen or heard about but the distinctiveness of the ‘Real Estate as Maven’ concept lies in:

The depth and sophistication of implementation
The integration and interconnectedness of these elements
The shift from passive provision to active facilitation
The focus on measurable outcomes beyond just work productivity
The emphasis on continuous, data-driven evolution

In a later essay we’ll elaborate on this but the central point is whilst individual elements might not be entirely new, the combination of all these characteristics, implemented at a high level and working together as an integrated system, is what makes the "Real Estate as Maven" concept distinctive. It represents a paradigm shift from real estate as a product to real estate as a service and facilitator of success.

The ultimate goal of all of this is to enhance uniquely human qualities and capabilities. These ‘Maven’ spaces will be where the best companies, with the smartest employees, will want to be. They will be the antidote to the cognitive laziness that is likely to be a major feature of business a decade hence. They will be places that enable people to be happy, healthy, productive, and connected.

And this is a transformational opportunity for the real estate industry. Where the real estate provider becomes an integral part of their tenants' success strategies, moving far beyond the traditional landlord-tenant relationship. This model could create spaces that are not just desirable to come to, but essential for personal and professional thriving in the modern world.

That is what one would be selling and that is where the magnetism will lie. Is it the whole market - of course not. But is it the ‘best’ market to be in, undoubtedly yes. Real estate just as real estate has no future. It needs to be more. It can be more. And people will love it.

‘Real Estate as Maven’ - who’d have thought it even possible?

Antony Slumbers 9/14/24 Antony Slumbers 9/14/24

A blueprint for the future of rental living

Explore the convergence of technology, sustainability, and community-centric design in the evolving rental living sector over the next 5-15 years

This is based on a research project undertaken for Say Property and Hyperoptic, to coincide with the launch of their Managed Wi-Fi product.

This conversation is courtesy of Google’s ‘Audio Overview‘ technology.

Listen to the ‘podcast‘

Antony Slumbers 8/22/24 Antony Slumbers 8/22/24

The Consequences for Offices of Artificial Intelligence

Introduction

Making predictions is a mugs game, so they say. However, extrapolating out core trends, not fads, isn’t as precarious. Just as with #SpaceasaService, which I first wrote about over a decade ago, there are certain fundamental drivers of change which can only really end up in one direction. Timing might, often is, be a tricky nut to crack but certain inputs really do necessitate certain outputs. And so it is, I think, with the ongoing developments around Artificial intelligence.

Here we are going to look at what the key drivers of change are, and how they ‘will’ manifest themselves across the real estate industry, primarily focussing on the office market. My time focus is 4-7 years, the length of many office development projects. So we’re considering what might be towards the end of this decade.

To be clear, these drivers and consequences are what I believe will apply to the top end of the market. Maybe top 30%. In much of the market, change will come a lot slower, maybe even after complete obsolescence has set in. Of the offices and occupiers.

I am talking about the spaces and places leading companies, from start ups to multinationals, will want to occupy. Buildings that will support companies that are creating the future, that are leaning in to AI and other new technologies. Companies that are early adopters, to early majority. Companies that in 2030 will be where most are in 2040.

So let’s start with the key drivers that will impact the office market.

Key Drivers

1. AI-Enabled Decision Making:

Increasingly we will witness a democratisation of insights and decision-making processes. AI tools, often mediated through natural language interfaces, will provide data-driven insights to employees at all levels, not just top management. This will enable faster, more informed decisions across organisations and raise the potential for more innovative problem-solving as diverse perspectives are empowered by these AI driven insights. This may, nay should, lead to more agile and responsive organisational structures.

2. Automation of Routine Tasks:

AI will take over repetitive and some managerial functions, whilst freeing up, and augmenting, human workers to focus on more complex, creative, and strategic tasks (with an emphasis on those where humans retain primacy). This could lead to significant changes in job roles and required skill sets. There is great potential for increased productivity and efficiency in many business processes, as individuals and teams restructure their work to concentrate on where they truly can add value.

3. Enhanced Data Processing and Analytics:

We will be making more informed, real-time decisions. AI can process and analyse data much faster than humans (which is obvious) but more importantly AI can provide deeper insights and identify patterns that might be missed by human analysis. AI, or at least a major branch of AI, is all about prediction. Everywhere we should be able to raise the quantity, and dramatically lower the cost, of high quality predictions. This will both lead to more data-driven cultures within companies but also significantly raise the value of judgement. Machines are great at prediction - advanced humans likewise with judgement. Though note the caveat - advanced.

4. AI-Powered Collaboration Tools:

Now there is a huge market for them, a great deal of money and attention is being directed at developing improved, often AI powered, collaboration tools. Distributed, remote and hybrid working is only going to get easier, and therefore more effective. Anything done entirely remotely will be easy to distribute globally. Location will be meaningless for these tasks, though in practice most jobs will require meaningful human connection - after all few humans are so skilled that they alone can generate complete sets of value. And jobs that don’t require human connection will ultimately all be done by machines. AI will enable deep virtual connection and then it’ll be about ‘what else do we need’? One should assume that no-one needs to be located three feet apart: why might they want to is the key question?

5. Flattening of Organisational Hierarchies:

We’ll see a large reduction in traditional management layers, as AI takes over some traditional middle management functions like scheduling and performance monitoring. AI should enable more direct communication between top leadership and front-line employees and this could lead to more autonomous, self-managing teams. Ronald Coase wrote ‘The Nature of the Firm’ in 1937, and he focussed on ‘transition costs’, primarily:

Costs of negotiating and drawing up contracts
Costs of gathering information
Costs of monitoring and enforcing agreements

All of these are bullseyes for AI, predictive and generative. If machines can do them, there is no requirement for human managers to do the same. Being in this middle zone is a very dangerous place to be. As it will be done away with, or at least hugely reduced in scale.

6. Reimagining of Business Processes:

The mechanism for point 5 above will be the ‘Unbundling and rebundling’ of work processes. AI will allow us to fundamentally rethink how work is done and value is created. In fact it it won’t be until we undertake this process that real value will be derived from AI. We need to stop iterating on the past, powered as it was by certain technologies, and turn our eyes towards redesigning work around what these new technologies enable. The direct analogy is the introduction of the electric engine which replaced steam power - great productivity gains didn’t manifest themselves until factories were fundamentally redesigned to take advantage of multitudes of standalone electric powered machines. This process took four decades - with AI it’s likely to take one at most. One has to assume and plan for rapid change on this front.

7. Continuous Learning Requirements:

There is going to be a huge need for upskilling and reskilling. AI is developing at such a pace that we all need to over index on continuous learning, with a very strong emphasis on adaptability. Currently organisations are not investing in nearly enough training. For every CEO proclaiming their companies AI initiatives only about a third of them are actually running the necessary training programs. Talk is cheap - training has to be taken seriously. The leading companies do, and this will result in significant competitive advantage.

8. Sustainability Imperatives:

This is already happening but the imperative will only intensify. As each year passes they’ll be an increased focus on energy efficiency and sustainable practices. Predictive AI should be a major influence here as it is very good at optimising resource use, reducing waste and energy consumption. This is a rapidly developing area and we ‘know’ what needs to be done, and largely how to do it. A growing area is generating and accessing renewable energy. Expect to see a lot of investment in energy infrastructure, particularly solar power married to battery storage. Self sufficient buildings has to be the north star being aimed at.

9. Evolving Employee Expectations:

We are changing. Always have of course, but it does seem that the global pandemic did act as a catalyst for many people. Demand for work-life balance and meaningful work environments has definitely grown since 2020.

Employees increasingly seek purposeful work and better integration of work and personal life. There’s a growing importance of company culture, values, and social responsibility and an expectation for more flexible work arrangements and personalised work experiences. Some say ‘wait for the next recession and all of this will disappear’ but that ‘feels’ unlikely. Adapting to this change in zeitgeist is a major requirement for the real estate industry.

Companies need to lead, but work happens within real estate.

10. Balancing Intangible and Tangible Business Aspects:

Not everything is going virtual. Yes much of modern business is focussed on intangibles but there still exists an awful lot of physicality. Life science and biotechnology space is very fashionable an area today, but there are many other areas where physicality is important. Healthcare, advanced manufacturing, aerospace and defence, the creative industries, education, training, personalised hospitality. Key is understanding which industries have these ‘special’ characteristics that demand very bespoke space. Their requirements are particular but at least they do actually ‘need’ office space.

11. Changing Economic Models:

And finally they’ll be a strong driver of change as work becomes ever more atomised. We’re already seeing larger companies make increasing use of ‘contingent’ workers, and a thinning out of full time, fully integrated workers. Tech companies have laid off large numbers over the last few years, with seemingly no impact on productivity and instead rising share prices. Sure they over hired during Covid but nevertheless it does seem like less people are required to maintain a steady state. As early adopters of AI perhaps we are seeing in tech companies the consequences of individually more productive individuals. One simply needs less people when each person operates at 5-10X what their peers did a few years ago.

Consequences - Anticipated Changes and Developments

All of the above is very likely to occur. It will evolve over time, and some companies will operate like this much faster than others, but betting against it happening is surely high risk.

So what will be the consequences of Artificial intelligence for the real estate industry, especially those dealing with still the sectors largest asset class, offices. What will be the nature of demand in the years ahead?

1. Shift to Flexible, Adaptive Spaces:

We simply do not know how space is going to be used in 3, 5, 10 years time. 10 years equates to 100X more powerful computers. How do we design for that? So much depends what becomes possible. People will use space based on that, and that is ….. just guesswork.

So we will definitely need to create spaces that that maximally flexible and adaptable. Reconfigurable with modular furniture and movable partitions. Spaces that can easily transform from individual work areas to collaborative zones. Where exceptional integration of technology allows for quick reconfiguration of digital and physical resources. We will of course be using AI to assist in all this space optimisation, that will create new configurations having learnt from usage patterns.

2. Increase in Collaborative Areas:

Much more space will be dedicated to teamwork and cross-functional collaboration. Which will require a variety of meeting spaces catering to different group sizes and work styles, with integrated technology to support both in-person and virtual collaboration. Informal gathering spaces to encourage spontaneous interactions and idea sharing will be a top focus. Most of the space will be the ‘water cooler’. Every space will be designed to catalyse ‘Human’ skills and capabilities. We are in the office to do what the machines, the AI, cannot.

3. Emphasis on Learning and Development Spaces:

A key ‘want’ in future offices will be exceptional areas dedicated to continuous skill development. In-house learning centres or corporate universities will become common. These will integrate AI-powered learning tools and simulators, and will be flexible spaces that can accommodate various training formats (lectures, workshops, hands-on practice). In many ways our places of work will become places of learning.

4. Enhanced Technology Infrastructure:

Our buildings are going to incorporate a lot of AI. We are moving to a world of intelligent, self monitoring and self optimising assets. All of which requires a lot of data, transmitted at very low latency. So every building will need advanced digital infrastructure to support all of this. Robust, high-speed networks capable of handling increased data loads. Edge computing, where analytics and AI inference (the running of AI applications) is done locally on devices incorporated into the buildings fabric, or attached to it, is likely to become a big thing. For a variety of reasons: Reduced latency, improved privacy, bandwidth savings, reliability and energy efficiency. All of this requires a high spec building.

5. Integration of AI-Human Interaction Spaces:

Our buildings will have dedicated areas for employees to work with AI systems that are designed for privacy and focus. With specialised equipment and interfaces for optimal AI-human collaboration, including immersive AI environments using AR/VR technologies. We’ll see the emergence of new spatial design principles for these AI-human workspaces. All of this will be a key reason to come to the office - to use equipment not available anywhere else.

6. Data-Centric Infrastructure:

Despite, or indeed because of, the focus on human-centricity, we’ll become far more data centric. We might even start to see the Integration of data centres within office developments. Buildings will be designed with enhanced cooling and power capabilities to support data infrastructure. There is potential for data centre heat to be repurposed for office heating, improving energy efficiency and this may lead to new approaches in building design that treat data infrastructure as a core component.

7. Reduction in Traditional Management Spaces:

As discussed above middle management is set to diminish so we will need less offices for this layer. Former executive spaces will be converted into collaborative or multi-purpose areas. With flatter organisational structures companies have the potential to design more open, democratic workplaces.

8. Decentralised Decision-Making Hubs:

Specialised and distributed spaces equipped with AI tools for decision support will emerge. Large companies may have their own but smaller companies will rent them as needed. These will take the form of 'war rooms' or 'decision theatres' with advanced data visualisation capabilities and integrated AI assistants to facilitate rapid, data-driven decision making.

These spaces will be designed to support both in-person and remote participation in decision processes. AI is really good at ‘modelling’ and ‘simulating’ and over time we’ll all become conversant with how to run these operations, but they will require bespoke setup and operational support.

9. Reimagined Workflow Layouts:

Offices will be redesigned reflecting the new realities of unbundled and rebundled work processes. Spaces are likely to be organised around work functions rather than traditional departments with the creation of 'neighbourhoods' for different work modes (focus, collaboration, learning, socialising). As happens today but more so. We will see the integration of AI-powered workflow management systems into the physical environment. Examples being smart meeting rooms, intelligent desks and workstations, AI-assisted collaboration spaces, automated asset tracking, personalised productivity environments and Intelligent wayfinding. Overall, layouts will become much more fluid but much more intelligent.

10. Focus on Employee Experience and Wellbeing:

Enabling people to be as happy, healthy and productive as they are capable of being will be the primary aim of any place of work. What this means is #SpaceasaService - spaces that provide each individual with the best possible environment to do whatever it is that they need to do, whenever they need to do it. Personalised, optimised, customised. Offices are likely to be designed more like hospitality spaces, prioritising user experience.

11. Sustainability-Driven Design:

Obviously sustainability is a non negotiable for offices of the future. Fortunately AI-powered systems will be massively helpful in this area, as they are excellent for predictive management and optimisation, as well as interacting with internal and external data sources. The potential for AI to manage complex sustainability initiatives across entire buildings or campuses is great. We know what needs to be done here and AI is very much our friend in getting it done.

12. Hybrid Work Support:

Everywhere spaces will be designed to integrate in-person and remote workers seamlessly. We will create 'Zoom rooms' and other tech-enabled spaces for virtual collaboration, and provide hot-desking and hoteling systems to manage flexible office usage. AI-powered scheduling and space management tools will be widely adopted and we’ll have enhanced audio-visual technology throughout the office to support impromptu virtual connections. As this form of working becomes deeply embedded we’ll use AI-driven tools to maintain company culture and employee engagement. We’ll spend years thinking about hybrid work but eventually it’ll just become ….. work.

13. Evolution of Urban Office Hubs:

We’re going to see a lot of the transformation of traditional office buildings into mixed-use spaces and the integration of residential, retail, and recreational facilities within office complexes. Some exist already but 'vertical villages' in urban skyscrapers will become commonplace. As will the repurposing of excess office space for other uses (e.g., vertical farming, educational facilities, sports venues, whatever we can imagine).

There is great potential for AI to optimise space usage and transitions between these different functions. This will start to strongly influence urban planning and zoning/planning regulations, leading to much more integrated city designs. CBDs will die off (with very few exceptions) as we gravitate towards more decentralised, polycentric urban development. AI will be at the heart of how we redesign our cities, but the guiding principle will be human centricity.

14. Rise of Mixed-Use Developments:

Everywhere we’ll be integrating office, residential, retail, and recreational facilities with a view to creating self-contained ecosystems that support diverse needs of workers and residents. This blending of work and living spaces to support changing lifestyle preferences will utilise AI for community management and service optimisation as these will be complex environments. There is potential for new types of leasing or ownership models in these mixed-use developments and this may lead to the emergence of new property categories that defy traditional classifications.

15. Location Strategy Shifts:

But where will we work. Yes, still in our newly forming poly centric major cities, but also in areas with strong AI talent pools and research connections. Traditional CBDs will have to compete with these emerging tech hubs or university adjacent areas. At a national level companies will give consideration to locations with advantageous AI and data regulations. There will be a balancing of physical accessibility with digital connectivity in location decisions. All of this may lead to the emergence of new tier-2 city hotspots with good quality of life and strong tech ecosystems and could result in a redistribution of office demand across wider geographic areas. Anywhere that can fulfil the technical and talent requirements, and is a ‘nice place to be’, will thrive. Agglomeration remains important but ultimately the internet is the master agglomerator. We’ll have more optionality and qualitative factors will become increasingly important.

16. Security and Privacy Considerations:

The more AI mediated our world becomes the more vital will be the implementation of advanced cybersecurity measures. Spaces that are designed to ensure privacy for confidential AI-human interactions will be at a premium. There will be pervasive integration of physical security measures with AI-driven surveillance and access control systems. It is likely we’ll see the emergence of new standards and certifications for AI-secure office spaces

17. Globalisation of Workforce:

Today the era of globalisation is receding but this’ll be short lived. Office spaces will need to facilitate international collaborations and virtual team management. The creation of 'global collaboration hubs' with advanced communication technologies and that have 'follow-the-sun' work models will impact on current office usage patterns. Particularly in major cities, where multinationals tend to congregate, offices will become more 24 hour. A 15 minute City, with a 24 hour lifestyle is rather wonderfully local and global at the same time.

Conclusion

The future of the office is a fascinating paradox. As AI automates routine tasks and reshapes traditional work patterns, it simultaneously amplifies the need for spaces that prioritise human connection, creativity, and well-being.

The offices that will thrive in this new era won’t simply be vessels for technology; they'll be vibrant ecosystems designed to cultivate uniquely human capabilities. They will be flexible, adaptable, and deeply integrated with AI, yet always centred around the needs and experiences of their human occupants.

This transformation represents both a challenge and an unprecedented opportunity for the real estate industry. By embracing AI not as a replacement for human-centric design, but as a powerful tool to enhance it, we can create workspaces that are not only more efficient and sustainable, but also more inspiring, engaging, and ultimately, more human.

And, I may add, more valuable.

Agree? Disagree? What have I missed? What have I got wrong? Let me know in the comments.

And thanks for getting this far!

Antony