THE BLOG
Two Years In: The Workforce Changes Reshaping Office Demand
Extraordinary AI capability gains are reshaping labour markets faster than most realise—with profound implications for which tenants will need more space, and which will need far less.
Executive Summary
Two years in, AI has leapfrogged in capability. Labour markets are shifting, with junior roles evaporating fastest. For CRE, this means some tenants will shrink, but others—those leveraging AI to expand—will grow, reshaping demand patterns.
Capability Development Phases
Two years ago I ran the first cohort of my #GenerativeAIforRealEstatePeople course. This week we’re starting the 12th.
I thought this would be a good time to review how the AI scene has changed over this period.
We’ve seen three phases of development:
Phase 1: Multimodality (late 2023 – mid 2024)
Focus: AI systems expand from text-only to unified models with images, voice, data, and search.
Nov 2023 – ChatGPT unified model (text, images, data, web search)
Nov 2023 – Arrival of Custom GPTs
Feb 2024 – Gemini expands to 1m token context window
May 2024 – ChatGPT 4o (advanced voice + vision)
June 2024 – Anthropic launches Claude Artifacts & Projects
Phase 2: Reasoning (mid 2024 – late 2024)
Focus: Models get better at structured thinking, coding, and sustained tasks.
Jun 2024 – Claude Code launched (dramatic leap in machine coding)
Sep 2024 – ChatGPT Projects launched
Oct 2024 – ChatGPT o1 (first “reasoning” model; advanced voice & search)
Nov 2024 – Google NotebookLM with audio overview
Nov 2024 – Lovable launched (“vibe coding” movement)
Dec 2024 – ChatGPT o3 (inference training + multi-pass consensus)
Dec 2024 – Gemini 2 & 2.5 Pro (“Thinking mode”, fully multimodal)
Dec 2024 – Deep Research arrives across models
Phase 3: Autonomy (2025 onwards)
Focus: Agents, orchestration, and early autonomous computing.
Jul 2025 – ChatGPT Agent launched (autonomous computing)
Aug 2025 – GPT-5 launched
Exponential Improvement Trajectory
Which means, in short, an enormous increase in the breadth and depth of capabilities.
When ChatGPT-5 launched recently the response was somewhat underwhelming. It was better, but not dramatically so, to the previous models.
However, if one compares the capabilities of GPT-5 with those of GPT-4, launched in March 2023, the difference is enormous.
It’s like the frog that’s been slowly brought to the boil without noticing.
Some have said AI progress is slowing down but, as David Shapiro has said ”AI is slowing down, insofar as most people are not smart enough to benefit from the gains from here on out!”
We have come a long way and have incredible power at our fingertips. All the time whilst, as Google announced last week, their energy cost of AI queries had dropped by 33X in just one year, and for any given unit of intelligence the ongoing trend of becoming 10X cheaper each year is holding steady.
Labour Market Impact Evidence
All of which is starting to have big consequences, even whilst it is still only circa 10% of ‘knowledge’ workers who use these tools every day as part of their workflows.
This week, three Stanford professors released a report, ‘Canaries in the Coalmine’, which examines early signs of how AI is reshaping the US labor market, especially for young and entry-level workers.
Key findings were:
Young workers (22-25) in AI-exposed jobs have seen a 13% drop in employment since late 2022, even as older colleagues in the same roles maintain or increase employment.
The decline is not due to industry-wide layoffs, but to firms quietly not backfilling entry-level positions as they become vacant, erasing the bottom rungs of traditional career ladders.
AI-driven job reductions are concentrated in roles where AI automates codified or rule-based tasks ("book learning"); jobs where AI augments human skills (especially those reliant on experience or tacit knowledge) are not impacted similarly.
Older and more experienced workers in the same occupations are not affected and may even see job growth, underscoring that this is age-specific displacement, not a general workforce reduction.
Wages across age groups remain flat, indicating that firms are reducing headcount rather than cutting salaries—for now.
Researchers controlled for industry and firm factors, showing that the phenomenon is linked specifically to AI-exposure rather than broader economic trends or unique company shocks.
Office Demand Implications
Meaning what for office demand?
Probably less people needing space for doing entry level work, and more space for more senior types capitalising on their ‘tacit’ knowledge.
Anecdotally one is hearing similar dynamics occurring across the professional services arena. The great intake of yearly graduates, populating the base of the pyramid, may be coming to an end?
Similarly, stories such as that recounted by Salesforce boss Mark Beniof last week, that he has sacked 4,000 out of 9,000 customer support staff, are moving from rarity to commonplace.
Now I am sceptical that this trend will last, and indeed my projection is that junior people will come to do well out of this new tech - see Issue #33 - but for now at least things aren’t looking good.
Either way, if younger people DO end up doing better than the research is suggesting right now, that might well be at the cost of more senior people. Same capability, but cheaper, has consequences.
For the office market, we’re going to need a lot of growth to support the market. The dynamic is 100% less people visiting, maybe more space per person for ‘collaboration’, and no mass return to 5 days a week. Without growth, this is a solid ceiling. If you don’t have the ‘right’ space, you’re in trouble.
In Pursuit of Tacit Knowledge
A key question for younger people is how to gain tacit knowledge? I think this might happen in two ways:
‘Someone’ will aggregate corporate tacit knowledge and publish it. This will happen across industries, and will be incentivised by the high value of this knowledge. Maybe ‘old timers’ codifying all they’ve learnt and passing it on. Or maybe through some sort of AI driven automated research tool, that gamifies in some way the passing on of knowledge.
Within corporates, Slack Channels, email, diaries, notetakers and the like will be able to codify all the tacit knowledge in this unstructured data. Hybrid working just makes this easier - as we increasingly work asynchronously, our ‘office’ becomes virtual and all our knowledge and information is contained inside this virtual ‘beast’. Getting to understand the actual workings of a company has never been easier. Tacit knowledge will become explicit just through the process of working.
And so either of these levels the playing field for younger people. And solves the problem of it being all very well getting rid of the cost of juniors but they’ll be no-one to enter the ranks of the seniors. Companies cutting down their intake lose the optionality that has historically given them.
‘Forecasting is hard - especially about the future’, said Nobel prize-winning Quantum physicist Niels Bohr. But…. I’d wager we’re going to see a different dynamic emerge anyway.
And this goes back to something we covered in #Issue4 and #Issue11 - https://www.flexos.work/trillion-dollar-hashtag
Strategic Imperatives
The future belongs to fast, agile, ultra-productive superteams and those designing their business ‘for a bigger pie’.
Incumbents planning their futures on doing what they’ve always done, just faster and cheaper, will be the death of the workforce. A mentality going nowhere and destroying all notions of a ‘social contract’.
Without new ways of doing, and aiming for growth, we’ll all be in terrible trouble.
So let’s not go there.
The lesson of the last two years of Generative AI is that it IS going to be increasingly easy to substitute machines for labour, and that many companies will be incentivised to do just that, ahead of any restructuring or redesigning of their businesses.
So, within CRE, we need to become very good at identifying these types of company, as they are definitely going to be needing less space. And frankly, are customers to avoid.
Focus on Two Things
Where we need to focus is on two things:
First, identifying those companies leaning in to the new technologies, coming up with new business models, and with aggressive growth targets.
And secondly, applying the same thinking to ourselves; as individuals, teams, divisions and companies. We need to be the ones ‘smart enough to benefit from the gains from here on out!’
In Preparation For Next Week
Next week I am going to build on this hypothesis: that we are moving towards an operating model where we’ll execute complete job functions from detailed prompts—define inputs, methodology, outputs, and let the model deliver results. And once optimised, these prompts will be evolved into autonomous agents.
The human role will transform into three functions: workflow design, data curation, and strategic intervention at critical decision points.
On the course we dig deep into where Humans + Machine works best. Where can we achieve AI Human Synergy - doing together what neither could do on their own?
This is going to require new types of working. And a very different mindset.
But, from the work I’ve been doing over the last few weeks, and based on the speed of progress since our first cohort, I’m increasingly convinced that this is not the pipe dream I thought it was a year ago, but actually the framework for what is definitely going to happen.
We shall see. Soon!
OVER TO YOU
Are you feeling the change? Are you thinking of moving job, redesigning your business, looking for different markets?
Do you think we can go straight for ‘a bigger pie’, or will we have to go through the painful transition first?
All my Blog Posts Are Now in My Weekly Newsletter
I now write a weekly newsletter - https://www.flexos.work/trillion-dollar-hashtag - and all my ‘blog posts are now printed there.
I list them all in the Index of the Archive on this site‘
I think you’ll like the newsletter - please do register and let me know.
AI And Bad Workmen
A new report has been caricatured as saying Generative AI is a failure - whereas it actually says that most ‘Enterprise’ users are simply using it badly
Don’t Believe Everything You Read
You might well have heard about a report that came out last week from MIT NANDA (Networked Agents and Decentralized AI) allegedly proclaiming that Generative AI was a busted flush and that 95% of enterprise pilots were failing.
The press, and a certain type of Linkedin ‘Expert’ were all over it. To the extent that, on August 19th, the UK’s Financial Times proclaimed that:
“US tech stocks sold off on Tuesday as warnings that the hype surrounding artificial intelligence could be overdone hit some of the year’s best-performing shares.”
Now, as Mark Twain quipped, one should "Never let the truth get in the way of a good story” but in this case I thought I better check on the veracity behind the cacophony.
And knock me down with a feather, it turns out that the report actually argues that Generative AI is a really useful tool that almost every large company is using incorrectly. Right there in the opening paragraph it states
“Just 5% of integrated AI pilots are extracting millions in value, while the vast majority remain stuck with no measurable P&L impact. This divide does not seem to be driven by model quality or regulation, but seems to be determined by approach.”
Not My Fault Guv’nor
Bad workmen are blaming their tools!
They are calling this the ‘Gen AI Divide’ - the stark gap between those generating considerable value and none at all.
And then they are very clear what the issue is:
“The core barrier to scaling is not infrastructure, regulation, or talent. It is learning. Most GenAI systems do not retain feedback, adapt to context, or improve over time.”
And the solution:
“A small group of vendors and buyers are achieving faster progress by addressing these limitations directly. Buyers who succeed demand process-specific customization and evaluate tools based on business outcomes rather than software benchmarks. They expect systems that integrate with existing processes and improve over time. Vendors meeting these expectations are securing multi-million-dollar deployments within months.”
In order to ‘address these limitations’ they go on to explain how Agentic AI systems are what is required because these can be configured with memory, adaptability, and autonomous workflow integration.
And all this is in the first three pages of the report, which you would not have thought would be too taxing for any ‘expert’ to read.
But then, writing up a report saying enterprises have an incredible new tool at their disposal if used correctly maybe isn’t as click worthy as ‘the bubble has burst - yet again the emperor has no clothes’.
The Enterprise Prerogative
Essentially, at enterprise scale, one has to ensure:
Integration with internal systems (CRM, ERP, lease databases)
Governance/traceability (audit trails, compliance logs)
Continuous improvement loops tied to business outcomes.
They need embedded, auditable, adaptive systems. And off the shelf, or even lightly customised LLMs don’t provide these.
So the report emphasises that enterprises should ‘Stop building static, non-learning tools that require endless prompting’.
What’s Wrong with Custom GPTs?
So why not build Custom GPTs?
The answer is that a Custom GPT can hold static context (instructions, tone, reference docs), but they don’t adapt dynamically from user corrections or workflow usage over time. If you correct an output today, it won’t automatically do it better tomorrow unless you manually retrain or rewrite the system prompt.
Which makes them configured assistants, rather than learning systems.
Big Is Boring - But Where The Money Is
WHICH IS FINE when using them for individual productivity, and within SMEs where the flexibility and adaptability is very much a feature rather than a bug, but in large enterprises you really need ‘process machines’.
It’s noticeable in the report that they say whilst most budget is currently pointed towards front-office applications (sales & marketing, customer operations etc) most of the ROI actually comes from building agents systems to manage back-office operations. Enterprises are process factories, not havens of entrepreneurial spirit. Frankly they need boring systems to do boring jobs that currently bored humans do.
2 out of 9 IS Bad
In their research covering nine industry categories, they found only two, Technology & Media and Telecoms, were on the right side of the ‘Gen AI Divide’. In these two industries, the impact of Gen AI was such that it was acting as a forcing function for structural change, and entire business models and operating models were being reimagined. For all the rest it was a world of bolt-on pilots, static tools and little ‘change’. In addition the levels of training were poor (only 40% paid for ChatGPT or the like Licenses), leading to a large amount of shadow IT, and a lot of ‘build-it-ourselves’ thinking was leading to botched implementations.
Where AI Is Working: Individual Adoption
Where these companies were having success was in areas they weren’t focussing on. Individual use by employees. In interviews they found that drafting, analysis, summarisation and outreach were very popular use cases at the edges, and adoption of consumer tools like ChatGPT was way ahead of anything handed down from above.
It Probably Does Not Apply To You Anyway
This report is very much focussed on larger enterprises, with upwards of thousands of employees. Mostly CRE companies don’t fit in this bucket. For instance the ‘large’ UK REIT, British Land, only directly employed 634 people as of mid 2024, and most developers, asset managers, and agencies operate with dozens to a few hundred staff. CBRE, JLL, Cushmans et al. are of course a lot bigger, but en masse CRE is an industry of SMEs. That often thrive by letting the bottom up approach reign.
As such, whilst (as we’ve written about before) we are going to see a lot of use of ‘agentic systems’ within our industry, we are not going to be constrained by the same imperatives that much larger companies have to operate under. And therefore should, as we already are, see a lot of Gen AI use at an individual or team level. People just making stuff happen, and pushing AI models to help them with whatever they have to deal with. No massive, centralised, bureaucratic plan needed.
And agility is a super-power in CRE. There are so many workflow/operating model innovations possible - leasing teams piloting deal-structuring agents; asset managers automating ESG reporting; development teams using AI for scenario modelling, and on and on.
The Bottom Line
Ultimately, the media's misreporting of this study perfectly illustrates the report's actual findings. The report describes a failure of implementation at the enterprise level, where structural change is needed but not forthcoming. We're seeing a lot of 'sound and fury... signifying nothing'. A complete bum steer.
Frankly this is not surprising - this technology is a ‘disruptive’ innovation, and that hurts, harms and annoys many people. So it is little wonder we’re seeing a degree of backlash right now. Who wouldn’t rather be safe and secure than throwing themselves into the rough and tumble of the ‘new’?
I suspect most readers of this newsletter are happy to be somewhat unnerved by all the change going on. We’re not the complacent type are we?
OVER TO YOU
What’s your greatest success with Gen AI? Where are you seeing how you work, or think, or act, change? Does this bother you? Let me know.
Real Estate's Great Rewiring: From Opaque Assets To Transparent, Self-Optimising Investments
From agentic facilities management to valuation 2.0, the next ten years will see real estate become cheaper to run, radically more transparent, and far more investable.
THIS IS A SPECIAL REPORT
It’s becoming clearer how AI and other technologies are likely to manifest themselves within the real estate industry over the next 10 years.
We have grouped 18 emerging trends into timelines: ones that are happening now, those that’ll develop over the next 2-5 years, and ones 5-10 years out. Several overlap but the trends demonstrate different use cases, representing meaningful workflows within the industry that we didn’t want to set aside.
If anything we are being conservative in these timings and they might occur faster, but it is safe to say that by 2035 we’ll be seeing an industry familiar and comfortable with a whole range of new technologies, workflows and behaviours.
As it is rather long, I’ve added this executive summary, which says it all in a nutshell. For the details, read the rest or skip to the trend you’re interested in.
Executive Summary
Overview
Over the next decade, AI and related technologies will transform commercial real estate from a fragmented, opaque sector into one defined by automation, transparency, and continuous optimisation. The adoption curve is compressing: many changes are already underway. By 2035, real estate assets will be cheaper to operate, higher performing for occupiers, and more transparent for investors—cementing the industry’s shift into a modern, data-driven future.
Key Timelines & Trends
Now – 2 Years (Already in play): Passkeys drive tenant app revenue; NABERS-style operational evidence replaces EPCs; provenance-first artefacts* accelerate deals; agentic FM triage cuts ticket times; AI-powered lease intelligence gains trust; controls optimisation pays back in <12 months.
2 – 5 Years (Scaling across portfolios): Budget-bounded AI agents handle procurement and property management; audit-ready, provenance-signed AI outputs become the norm; plant CapEx includes optimisation with M&V guarantees; live, signed data feeds flow directly to valuers and lenders.
5 – 10 Years (Systemic transformation): AI orchestrates O&M by default; portfolio-wide disclosure becomes mandatory; live operational data streams feed directly into DCFs; “assurable automation” emerges as a premium in secondary markets.
Why This Matters
Efficiency → AI agents cut portfolio operating costs by 20–40%.
Transparency → Provenance-signed, performance-based data shortens sales cycles and eliminates pricing games.
Capital → Assets with auditable automation and transparent operational evidence secure cheaper finance and higher liquidity.
Talent → Real estate finally looks like a modern, data-rich industry—attracting the calibre of people it has long struggled to retain.
————————————————————————————————————————
The Trends
Present/Near-Term (Already in play or with rapid impact):
1. Passkeys As A Revenue Lever
What/Why?
Passkeys, or device-based logins, eliminate passwords and reduce user friction. This in turn leads to more frequent sign-ins and higher engagement, as users authenticate via biometrics or device PIN. Less friction is highly likely to translate into higher ‘Monthly Active Users’ (MAUs) and more opportunities for personalised offers and transactions, effectively removing the "password tax" to unlock untapped revenue.
What to watch for:
Growing support from major tech companies. As of late 2024 20% of top websites support passkeys and Amazon reports 175 million customers use them.
Value hook:
It would not be unreasonable to see 15-30% more active users of a tenant engagement app with perhaps a 5-10% boost in revenue from increased amenity sales. This would transform identity from a security cost centre into a revenue lever, enabling "agentic" (AI-driven, personalised) interactions and improving security by virtually eliminating phishing risks.
2. Operational Evidence Beats EPCs in Due Diligence
What/Why?
Traditional due diligence relies on static metrics like Energy Performance Certificates (EPCs), which are based on design specs. However, investors and tenants realise these don’t tell the full story and will increasingly demand actual operational evidence, such as 12 months of utility bills, real consumption, and indoor environmental data, which is far more revealing of a building’s true performance.
We are already seeing this reflected in the increasing use of the NABERS rating system, which uses metered energy.
The great unlock is that providing a digital trail of operational evidence reduces uncertainty and would help speed up sale proceedings.
What to watch for:
Increased use of NABERS-style ratings but most especially investors explicitly pricing deals on in-use efficiency, with compliance to operational carbon targets influencing bid levels.
Value hook:
The lack of transparency of operational evidence costs time and money when it comes to sales, or investments. No-one trusts EPCs (rightly) and any good negotiator will chip away at the price based on lack of robust data.
3. Provenance-first Artifacts* Shorten Sales Cycles
What/Why:
A major time sink in real estate deals is verifying document authenticity and accuracy. Provenance-first artefacts are documents and data files cryptographically signed at the source, creating tamper-evident records. This immediately establishes trust in data integrity, collapsing the audit burden in due diligence by making cross-checking redundant. Knowing each artefact is "as originally issued" accelerates deal processes.
What to watch for:
Startups working on "blockchain-verified proofs" for private real estate data rooms, Real estate platforms using technologies like blockchain, digital notaries, or secure data escrow for due diligence, the emergence of ISO standards or industry protocols on digital building documentation, insurers or Big Four auditors requesting digitally signed building performance data, and sellers advertising a "verified data package" in their deals.
Value hook:
The core value is speed and confidence. By removing doubt one is effectively monetising trust.
4. Privacy-Preserving Occupancy Beats Camera Analytics (Total Cost of Risk)
What/Why?
Monitoring occupancy is vital, but camera-based analytics present high privacy and compliance risks (e.g., GDPR fines, legal liability, tenant backlash) due to collecting identifiable data. Privacy-preserving occupancy solutions (e.g., thermal sensors, infrared detectors) avoid collecting personal data by only counting signals, drastically lowering compliance overhead and risk. These solutions deliver necessary occupancy metrics without identifying individuals.
What to watch for:
Fewer employee complaints regarding surveillance, faster regulatory approval for sensor projects, with privacy-friendly sensors "sailing through" internal compliance, companies removing cameras due to staff backlash and Insurance or legal advisors recommending non-camera solutions to reduce liability.
Value hook:
The value is primarily in risk avoidance and operational savings. Face recognition is a GDPR compliance nightmare, involving throwing money down the drain. Avoiding it isn’t hard and makes little or no difference to how you operate.
5. Agentic FM Triage Under Spend Caps Cuts Cycle Time (Without Scaring Risk)
What/Why?
Facilities Management (FM) involves numerous small issues requiring quick resolution. Agentic FM triage uses AI "agents" to automatically handle routine work orders up to a defined spend limit, dramatically speeding up response times. These agents classify and action requests instantly, handling the routine 80% of tickets. Spend caps and rules keep risk in check, ensuring complex or high-value issues are escalated to humans, thereby lowering risk exposure.
What to watch for:
Early deployments in large portfolios as FM teams use their scale to re-orient around doing more with less. Change management programs looking at where the ‘human in the loop’ needs to be and how much leeway, in terms of budget, an AI can be allowed. Software companies will surely be building in such capabilities soon.
Value hook:
It’s a straight numbers game. If AI handles 30% of tickets automatically, FM teams can be 20–30% more productive, managing more buildings with existing staff. The budget caps keep everything manageable.
6. Controls Optimisation (Pick One Plant Loop) Pays Back < 1 Year
What/Why? Most large buildings’ HVAC and plant systems run sub-optimally. Controls optimisation applies advanced algorithms to run equipment more efficiently. While owners may fear costs, targeting a single plant loop or system yields such significant energy savings that the project pays for itself within a year. This "low-hanging fruit" approach is a bite-sized project with minimal disruption, as it involves improving existing controls rather than replacing equipment.
What to watch for:
Successful pilots, “Optimisation” becoming a dedicated budget line item in CapEx plans, and performance-based contracts (e.g., "No savings, no fee") for building optimisation becoming common.
Value hook:
The economics are compelling: energy costs in real estate are a significant line item, and everyone knows there is a lot of low hanging fruit in terms of badly configured systems.
7. Lease/Contract Intelligence with Citations
What/Why? Manual legal and financial analysis of leases and contracts is slow and costly. And the bane of everyone’s life in real estate. AI intelligence tools are now good enough to overcome the ‘trust problem’ and are "bankable" for decision-making by lenders, investors, and asset managers.
What to watch for:
More deals and workflows explicitly incorporating AI outputs, often noting they were "generated with AI and verified”, regulatory and court acceptance of AI-generated contract analysis as evidence, provided sources are verifiable, insurance or lender guidance accepting AI-flagged risks or waiving full attorney review if AI tools with citations have been used and major real estate brokerages and services firms partnering with AI providers for lease abstraction.
Value hook:
Dedicated lease extractions services, with direct citations, are good enough that no-one should be manually extracting leases. The cost savings are considerable, and the savings in time are important for ‘getting deals done’. Contract intelligence is similar though it’s likely a ‘human in the loop’ will be required. For now.
8. Converged Identity Extends to Visitors and Contractors
What/Why?
Traditional systems for employees, visitors, and contractors are often fragmented, creating inefficiencies and security gaps. Converged identity management provides one unified platform for anyone needing access (physical or digital), streamlining processes. This improves security and compliance by enforcing uniform policies and enhancing convenience with a single digital credential. It closes security vulnerabilities by automatically deactivating access when contracts end or IT accounts are disabled.
What to watch for:
New solutions from tech suppliers, policies requiring visitor and contractor compliance, regulatory drivers in sectors like healthcare and data centres, and visitor kiosks issuing credentials managed in the same system as employee badges.
Value hook:
Major gains are in security and efficiency. A unified identity reduces "orphan" accounts and badges, lowering breach risk and avoiding incidents. Possibly reduced insurance costs but significant efficiency savings in onboarding and compliance. Quantitatively, converged IAM can lower identity management costs by ~30%. It ultimately protects asset value and reduces costs from fragmented systems.
9. Performance-Based Disclosure (NABERS-Style) in Particulars & Term Sheets
What/Why?
Investors, tenants, and regulators are increasingly demanding actual performance data (especially energy and carbon) over design intents. In the near future, marketing documents for sales or leases will prominently feature performance-based ratings (e.g., NABERS, ENERGY STAR scores, actual consumption figures) as standard. This shift is driven by existing mandates (e.g., Australia), investor ESG mandates, and regulation (e.g., NYC’s Local Law 33), as stakeholders require real, measurable performance data.
What to watch for:
Appearance of energy use intensity (EUI) figures, operational carbon numbers, or NABERS/ENERGY STAR ratings on the first page of offering memorandums and broker brochures. Plus new industry standards (e.g., RICS, NCREIF), tenants requesting performance data during leasing, and "Particulars Plus" data rooms offering live data feeds supporting headline figures.
Value hook:
This enables market differentiation and pricing of efficiency. Buildings with strong performance disclosure will enjoy a "green premium" or better liquidity, while laggards might face a "brown discount”. Once performance-based disclosure becomes common, failure to be open will be punished, a shift that will be transformational for the market in data.
10. Insurers/Lenders Begin Pricing ‘Control-Plane Maturity’
What/Why?
Insurers and lenders are refining risk assessment. As buildings become smarter, their "control-plane maturity" (how advanced, reliable, and secure automation and control systems are) is an emerging risk factor. A building with robust, well-integrated, and cyber-hardened controls is deemed lower risk for property insurance and operational disruptions. Insurers already discount for IoT monitors; this will extend to pricing in the overall control environment.
What to watch for:
Insurers rolling out new products or endorsements for smart buildings, underwriting questionnaires asking detailed control-system questions, green financing criteria expanding to consider operational tech certifications or smart scores and rating agency commentary positively viewing portfolios with high digital infrastructure maturity.
Value hook:
Lower insurance premiums, attractiveness to investors and improved access to capital. This trend effectively monetises good operations.
Mid-Term (2–5 Years):
11. Budget-Bounded Agents Run Part of PM/PPM and Low-Value Procurement (2–5 Yrs)
What/Why? This is pushing on from trend 5. In 2–5 years, autonomous AI agents will handle a significant portion of property management (PM), preventive maintenance (PPM), and procurement tasks within defined budget limits. This extends beyond reactive triage to proactive management, like ordering consumables when inventory is low or scheduling routine maintenance within an annual budget. The "bounded" approach prevents overspending, ensuring financial risk is controlled while boosting efficiency.
What to watch for:
Formal policies authorising AI-driven spending, with CFOs setting expenditure limits for AI agents, startups marketing "auto-procurement bots" for spare parts or AI scheduling assistants for preventive maintenance, vendor consolidation of physical and digital ID systems allowing agents to manage contractors as just-in-time resources, and integration of IoT, inventory systems, and procurement, enabling AI agents to orchestrate sequences from fault detection to scheduling technicians.
Value hook:
This is when costs start to tumble. Automating 40–70% of procurement/PM tasks will substantially cut administrative overhead. Agents operate 24/7 with instant response, ensuring faster maintenance and preventing issues from escalating, avoiding downtime or damage costs. AI can reduce maverick spending and secure bulk discounts by consistently following plans. Every AI action is logged, providing detailed audit trails and consistency, reducing mistakes and mitigating risk.
12. Audit-Ready, Provenance-Signed AI Workpapers Become Normal
What/Why?
As AI increasingly informs business decisions, especially in regulated areas like finance and real estate accounting, there's a need to trust and verify its outputs. In 2–5 years, AI-generated analysis ("AI workpapers") will come with a robust audit trail, metadata, and cryptographic signatures proving its generation process, data, and integrity. This "provenance-signed" approach ensures auditors can confirm legitimacy, enabling "assurable automation".
What to watch for:
Auditors and regulators starting to mandate these provenance mandates, tech suppliers enabling them, industry consortiums defining standard formats, and insurers requiring audit logs.
Value hook:
The value is trust and compliance, directly impacting the bottom line by reducing the risk of compliance failures and improving decision quality. Audit-ready AI unlocks greater AI ROI by enabling confident automation. No more black boxes and immediate explainability.
13. Optimisation Becomes a Line-Item in Plant CapEx – M&V or Don’t Pay
What/Why? Historically, CapEx for plant upgrades focused on equipment, not ongoing optimisation. This is changing: in 2–5 years, it will be standard to include an "optimisation" line-item in plant CapEx, often under a "Measurement & Verification (M&V) or you don’t pay" performance contracting model. The industry recognises that new equipment needs continuous tuning (via analytics, AI optimisation) to deliver promised savings. RFPs will bundle multi-year optimisation agreements, linking payments to verified energy savings. See previous newsletter about ‘Outcomes as a Service’.
What to watch for:
RFPs and contracts using language like, "Vendor will provide continuous commissioning software... Payment tied to performance subject to M&V”, accounting treatment that allocates part of plant CapEx to an "opex-as-a-service" model, consultants recommending "optimisation or bust" for any plant upgrade, ASHRAE and CIBSE incorporating M&V and optimisation tuning periods into major plant retrofitting standards and insurance or lenders preferring projects with performance guarantees.
Value hook:
The value is ensuring projected savings are actually achieved and persist. An optimisation line-item, though adding 5–10% to project cost, can hugely improve ROI by guaranteeing 20–30% energy reduction. Making vendor payment contingent on M&V eliminates the risk of wasted CapEx on the optimisation part.
14. Data Rooms Offer Live, Signed Feeds to Valuers and Lenders
What/Why?
Real estate due diligence has relied on static documents. In 2–5 years, data rooms will commonly include API endpoints or live CSV links to key building performance data, cryptographically signed or blockchain-verified for integrity. This allows valuers and lenders to pull real-time data (e.g., hourly energy consumption, occupancy rates) directly from building systems, rather than relying on summaries. This enables dynamic analysis, speeding up diligence and improving accuracy.
What to watch for:
Inclusion of data APIs in deal marketing, valuation firms tooling up to consume API data from buildings, adoption of blockchain, regulatory moves requiring continuous provision of KPIs, and formalising the practice of providing BMS logins into read-only APIs with digital signatures in VDRs.
Value hook:
Value is in speed and confidence in transactions. Live, trusted data can reduce due diligence by weeks, enabling more precise and faster underwriting and faster closing for sellers. And be even more transformational than trend 9.
Longer-Term (5–10 Years):
15. Agent-Orchestrated O&M Becomes Default; Humans Manage Exceptions
What/Why?
Within 5–10 years, AI agents will coordinate most routine Operations & Maintenance (O&M) tasks by default, with humans managing only exceptions, oversight, and vendor relations. This is the logical culmination of earlier trends, where AI systems become the "autopilot" of building operations.
What to watch for:
Significant changes in staffing ratios (e.g., 1 facility manager per 50 buildings instead of 5), FM providers (e.g., JLL, CBRE) offering "AI-driven FM" solutions, case studies of buildings operating "95% autonomously”, and advances in robotics for building maintenance (window cleaning drones, floor scrubbers) as physical actuators commanded by AI agents.
Value hook:
Potential for huge efficiency gains from reducing human labour by 50–70% for routine tasks, leading to net savings of ~30% of O&M costs. Tenants benefit from a more responsive building, potentially leading to higher rents or retention. It is also scalable, allowing portfolio growth without linear OpEx increases.
16. Portfolio-Wide In-Use Disclosure Is the Market Norm
What/Why?
In 5–10 years, large real estate owners and REITs will routinely disclose portfolio-wide operational performance metrics (energy, water, emissions, occupancy) to investors, regulators, and the public as standard practice. This goes beyond individual building disclosure to aggregated, portfolio-level openness.
What to watch for:
Regulatory mandates requiring large real estate owners to measure and disclose operational carbon annually, major firms signing onto net-zero commitments requiring transparent progress reporting, tenants demanding owners share building performance data portfolio-wide and stock exchange requirements for ESG metric disclosure for REIT listings.
Value hook:
Can yield lower cost of capital and higher valuations for leaders. Transparent performance management often leads to a "transparency premium" from investors. Public data forces internal focus, spurring efficiency improvements and cost savings.
17. Valuation 2.0: Live Operational Inputs Flow into DCFs by Default
What/Why?
Traditional property valuation relies on static inputs. In 5–10 years, valuations will shift to a dynamic model where live operational data streams flow directly into DCF software by default. Appraisers will pull real-time data (occupancy, traffic counts, energy costs, air quality) from building systems, enabling more accurate and forward-looking valuations. This is driven by the emergence of supporting technology (APIs, digital twins, data standards) and industry calls to integrate sustainability and operational performance into valuation methodology.
What to watch for:
Valuation software (e.g., Argus Enterprise) integrating APIs for data ingestion, advisory firms advertising "data-enhanced valuations" using live building data, valuation standards (e.g., RICS Red Book) evolving to encourage use of operational data, digital twin platforms partnering with appraisal firms to feed them data and forward-thinking investors requiring actual sustainability performance data in cash flow analysis.
Value hook:
The most direct value is improved accuracy and confidence in valuations. This leads to better investment/lending decisions and less risk of overpaying/underselling or loan defaults. Live data helps catch declines in performance early, allowing for "real-time adjustment" of valuations and operational intervention to preserve value.
18. Secondary-Market Premium Emerges for ‘Assurable Automation’ Buildings
What/Why:
In 5–10 years, buildings with highly automated systems that are demonstrably assurable (reliable, safe, verifiable, cybersecurity hardened) will command a premium in the secondary market (sales, REIT/share price). This evolves from "green premiums," recognising the value of automation proven to be trustworthy.
What to watch for:
Brokers highlighting “SmartScore Platinum” in offering memos, actual sale comparables showing assets with verified automation selling at lower cap rates, ”Assurable automation" becoming a due diligence item, with buyers hiring consultants to audit a building's tech stack, and top tenants demanding fully smart-enabled buildings and paying more rent for them.
Value hook:
Quantitatively, an “assurable automation” building could have operating costs 10% lower and command 5% higher rent, significantly boosting NOI. It rewards buildings that are effectively "future-proof" and high-performing due to tech, leading to formal adjustments in valuations.
CONCLUSION
By 2035, most buildings will run largely on autopilot: self-describing, continuously optimised, and radically more transparent. The strategic stakes are clear:
Efficiency → Operating costs fall 20–40% as AI handles the bulk of routine work.
Transparency → Hidden inefficiencies disappear; pricing chips and data droughts become relics.
Capital → Valuations and financing will hinge on live, audit-ready performance data.
Talent → An industry once seen as slow and opaque becomes a magnet for modern, data-native professionals.
The winners will be those who act now - doubling down on today’s trends, positioning for mid-term automation, and ruthlessly questioning which assets and business models can survive in a world of transparent, self-optimising real estate.
OVER TO YOU
Which trend can you double down on today? Which one can you lean into for tomorrow? Which assets does this make you want to sell, and which to buy? And how might this change how you work?
It’s all doable, and will happen. And a lot of it already is. Where is your place in this world?
* Provenance-first artefacts mean documents or digital objects that are cryptographically signed at the source, creating tamper-evident records.
Short-Term Wins Are Still Wins!
The Hidden Alpha: Reasoning Models + Great Prompts
Last week we went into a deep dive about how to create long-term strategic advantage with commercial real estate. This week we’re going to look at short term tactical advantages.
As readers of this newsletter you probably are aware that OpenAI released its long-awaited upgrade to ChatGPT - Version 5, last week. And the incredible hullabaloo it created. People were expecting the moon and when the undeliverable was not delivered they got very animated. At times you’d think the world was about to end.
The advantage is bigger than we thought.
There was though one short tweet (X if you really must) that Sam Altman put out that contained some extraordinary, and important, information:
“the percentage of users using reasoning models each day is significantly increasing; for example, for free users we went from <1% to 7%, and for plus users from 7% to 24%.”
I was not surprised that almost no free users were ever using the ‘reasoning’ models but rather staggered that only 7% of ‘Plus’ paid users were. These models, notably o3, were so much more powerful than the default model, 4o, yet were going almost unused.
The biggest thing about GPT-5 is that for the first time hundreds of millions of people will have access to frontier reasoning models and, at last, will be able to see what these things can do! This IS a big deal.
Paid users, though, had access, but 93% were not using them. And perhaps this will remain the case for a majority at least.
So it seems likely that the gap between capability and day-to-day utilisation remains huge—and that gap is exactly where your personal and organisational alpha sits.
Most teams still use AI like search. They dabble in chat, copy/paste the output, and hope for the best. Which is such a waste because the edge comes from using reasoning-capable models with excellent prompting discipline, applied to high-leverage work. This is where GPT-5 shines—and where a small set of patterns will yield outsized results.
For now, this is a huge short-term win for you, your team, and your company—if you act. Below is some practical advice about how to make the most of this ‘capability gap’.
From “using AI” to “using it well”
I use a simple stack to explain why prompt quality matters:
WHO / WHAT / HOW
WHO: Do you want the AI to act as? Adding context is important. Be clear who you want to be interacting with.
WHAT: Do you want the AI to do? Think clarity, context and constraints. LLMs cannot read your mind - you need to tell them what you want.
HOW: Do you want the output to be returned? Length, format, structure. How do you want information presented?
These fundamentals turn a vague ask into a brief. They’re the front door to all good prompting.
The Importance of Patterns
And on these foundations you overlay certain ‘Patterns’ - types of prompt addressing particular circumstances.
Examples of these patterns include ‘Cognitive Verifier’, ‘Self-Refine’, ‘Self-Consistency’, ‘Flipped Interaction’, or ‘Outline-to-Synthesis’. These are reusable “moves” that force better thinking.
You then add ‘Verification’. Ask for a plan first, then a critique of that plan, then a final answer. This dramatically improves accuracy and defensibility for board-facing work.
And ‘Evidence handling’. For research, require citations, uncertainty labels, and explicit counter-arguments.
And finally, a ‘Deliverable spec’. Decide the end product upfront: one-pager, IC memo, risk log, table, or slide deck. The model writes to the format you define.
When you combine these habits in GPT-5, the model stops being a clever autocomplete and becomes a co-pilot for judgement work. The difference really is chalk and cheese, and normally amazes people when they first see it all in action. Expect this to go mainstream in the coming months.
Five GPT-5 patterns that change outcomes
Let me walk you through five GPT-5 patterns that change outcomes (with CRE use-cases).
1) Cognitive Verifier — Plan → Critique → Answer
Where to use it: IC memos, board papers, credit submissions, asset strategies.
Why it works: It reduces blind spots and shortens review loops by building a self-check into the workflow.
Example prompt: “Act as a senior investment committee adviser. First outline your plan to evaluate [insert asset/market information - more is better]. Pause. Critique your plan for missing assumptions, weak evidence and risks. Only then produce a one-page recommendation with mitigations.”
2) Self-Refine — Draft → Self-Review → Redraft
Where to use it: proposals, quarterly packs, client comms, investor letters.
Why it works: Quality rises across controlled iterations without human micro-management.
Example prompt: “Draft the [insert required document details - again more is better]. Self-review against these criteria: accuracy, clarity, evidence, senior readability. Produce a revision plan, then redraft and list the top five changes and why.”
3) Self-Consistency — Multiple paths → Converge on best
Where to use it: valuation sensitivities, strategy options, scenario planning.
Why it works: Running independent reasoning paths reduces idiosyncratic error; you see consensus and dissent clearly.
Example prompt: “Run three independent analyses of [insert deal details]. Use different assumptions/paths. [you could define this if you wish]. Show consensus, the strongest dissent, and a recommendation with what evidence would overturn it.”
4) Flipped Interaction — Socratic stress-test
Where to use it: lease strategies, transformation roadmaps, go-to-market plans.
Why it works: The model interrogates you, forcing clarity on constraints and hidden assumptions.
Example prompt: “Act as [insert the role of your interlocutor]. Ask sequenced questions to pressure-test my plan [insert plan details]. Don’t proceed until I answer. Be firm but fair. Summarise gaps and offer corrections.”
5) Outline-to-Synthesis — From bullets to a defensible plan
Where to use it: converting outlines + leases/EPCs/notes into an asset plan or ESG roadmap.
Why it works: You expand structure without losing the brief; sections stay evidence-linked.
Example prompt: “Using the outline and attached sources, expand into a six-page asset plan. For each section: decisions, evidence, owner, KPI, and 90-day actions. Include a one-page executive summary.”
And More!
And beyond these five it is worth considering the ‘Vision & “Reasoning-Plus” research capabilities of these models.
Vision. Ask the model to critique masterplans or dashboard screenshots and list “unknowns to resolve before pre-app”. You’ll be surprised how competent frontier models (GPT-5, Claude, Gemini) are at reading visuals in context.
Agentic research. Today’s models can plan, gather sources, synthesise, and self-check. Use this to compress the first 60-70% of a company brief or market note, then spend your time on judgement and negotiation. In particular Gemini and ChatGPT both have ‘Deep Research’ modes, and ChatGPT has a separate ‘Agent’ mode where it can actively visit websites and your own computer network (with permissions of course).
This is where you get an advantage others aren’t taking.
Where the edge shows up in Commercial Real Estate
And here’s how these capabilities directly translate into CRE workflows.
Underwriting & valuation. Multi-scenario cashflows, covenant testing and comps coherence, packaged as a defensible memo. (Verifier + Self-Consistency.)
Leasing strategy. Trade-offs (rent-free vs capex), anchor-tenant sequencing, adversarial planning for negotiations. (Flipped Interaction.)
Asset management. Quarterly plans that reconcile OPEX, tenant feedback, and BMS data into a single action map. (Outline-to-Synthesis + Verifier.)
ESG/retrofit. Pathway design with standards cross-walks, capex phasing and risk logs. (Outline-to-Synthesis + Self-Refine.)
Workplace strategy. Hybrid policy, utilisation analysis, change-management comms. (Self-Consistency + Flipped Interaction.)
The Window is Open (for now)
The above are just a starter set of tools you can leverage. There is almost no aspect of real estate where the application of excellent prompts will not yield strong results.
Yet, to repeat - ALMOST NO-ONE IS TAKING ADVANTAGE
Over time (I guess - 12-18 months) the default setting will shift towards reasoning; rate limits will rise, friction will fall, and more people will use these tools. But capability ≠ competence. Habits, libraries and governance are durable moats. If your top five workflows don’t yet have standard reasoning templates, you’re leaving results on the table. The easy pickings are there.
OVER TO YOU
A Personal challenge: pick one high-stakes task this week and make reasoning-by-default your norm—using a verifier step and a defined deliverable. Measure cycle time and revision count.
Team move: nominate three pattern owners to maintain a shared prompt library measure internal use of reasoning.
If you want to institutionalise these wins before your competitors catch up, the next #GenerativeAIforRealEstatePeople cohort starts on the 5th September and includes the full GPT-5 Prompt Library, with live practice on the patterns above.
The advantage is real—and still underpriced.
Short-Term Wins are still Wins!
But the window WILL close, so act now.