Can we go from ‘Zero to One’ in Real Estate?

Antony Slumbers / Midjourney

Peter Thiel wrote a book about ‘Zero to One’, a term that represents a leap from nothing to something, creating a novel product, service, or idea that did not previously exist. And he compared it to ‘One to N’ which is about scaling, improving, and distributing existing products and services.

In real estate we’ve traditionally done a lot of ‘One to N’ - in the future, and as a consequence of AI, we need to be thinking more ‘Zero to One’. Technology is so fundamentally changing the nature of real estate demand that we need to be thinking much more boldly about the products and services we deliver, and how we deliver them.

We need to be developing entirely new models of real estate usage and development, redesigning our urban spaces to accommodate future works patterns and lifestyles, and thinking how we can offer unprecedented levels of efficiency, convenience and sustainability.

Instead, and for understandable reasons, we are mostly looking at how we can digitise the past. Can this new technology, this PropTech, enable us to do X better, faster, cheaper; rather than considering should we even be doing X.

With most technologies this makes sense. That is exactly what they are designed to do. Allow you to do more of what you already do.

AI though is different, especially Generative AI. This type of AI is probabilistic, not deterministic. To say it is just a ‘stochastic parrot’, a ‘next word prediction machine’ is to underplay how capable it is, but it is not in the business of 2 + 2 = 4, or performing a database lookup. It is, fundamentally, a creative tool. And that is what people find hard to get to grips with. We are used to getting answers that are ‘factual and neutral’ - that is not what this technology is all about. This root characteristic is why it is much more likely to lead to ‘Zero to One’ scenarios than ‘One to n’. It’s not dull enough for that. If you want to add up, get a calculator.

If you want to really innovate though, this is the way to go. It can enable things hitherto impossible. It doesn’t lend itself to doing things for you, though it often can. Where it shines is when you use it as a sparring partner, a confidant, a consigliere, a sounding board. At best it acts as a Socratic Mentor. Where it encourages deep thinking, self awareness and logical reasoning. It empowers you to take ownership of your learning and decisions. It’s a powerful flywheel of intellectual curiosity and critical thinking. And that’s where innovation is going to come from.

Probably you are not using it like this. And likely have not considered it thus. More likely it has been promoted to you as helping you reply to emails, or taking notes in meetings, or summarising or writing reports. All of which are useful and worthwhile, but ultimately very thin gruel. This is ‘One to n’ territory.

Richer use cases might involve:

Transforming Tacit into Explicit Knowledge - codifying all your companies expertise, whether structured or unstructured into dynamic knowledge graphs. These would democratise access to critical expertise, especially amongst less experienced employees, allowing them to leverage the organisation's collective intelligence.

Virtual Mentorship Programs - Use LLMs to simulate the decision-making processes of experienced professionals, creating interactive scenarios where less experienced employees can engage in problem-solving exercises guided by the model. And then backed up by in-person human mentoring. We hear so much about the supposed problems mentoring younger employees in a hybrid working world that completely ignores the potential of AI to develop hybrid mentoring programmes that are orders of magnitude better than the norm. Done well we should be able to turn 2 years experience into 20.

Autonomous Insight Discovery - Anyone can work hard and discern the obvious. What we want is to uncover the non obvious, the unknown unknowns. Financial data, market trends, consumer behaviour, and operational metrics can all be mined at unprecedented levels, uncovering opportunities for cost savings, investment, and strategic pivots that were previously hidden. The machines can act autonomously but the ‘human in the loop’ uses their experience to guide and cajole. Much of our future value will come from our curatorial skills and wisdom. Oftentimes ‘we know more than we can tell’ - AI can help us make visible the previously invisible.

Innovation Catalyst Systems - Real estate is famously siloed. Left hand does not talk to right hand. We cut ourselves off from discovering novel combinations of existing technologies, processes, and business models. Machines have no qualms about who they talk to; and they’re stupendous at synthesising data. Set them free.

Hyper-Personalised Real Estate Services - How good are we at analysing client behaviours, preferences, and market dynamics in real-time? AI enables personalisation that humans cannot. But humans are definitely better at face to face. The key here is playing to each others strengths. Few of us have staff constantly briefing us about what we need to know about a given situation or people we are set to meet. AI can give us all this superpower.

Enhanced Decision-Making Frameworks - all of the above can and should feed into our decision making. A continuous feed of relevant data, predictive analytics, and scenario simulations available to all.

Two key points come from all of this (and the myriad of other advanced AI use cases):

First is that AI is our most useful assistant as we strive to go Zero to One. It simply offers capabilities that did not exist before, and ‘should’ enable more of us to reach a level of innovative thinking that maybe we never thought we could. Being extremely well briefed and informed is becoming much easier.

And secondly, it emphasises the critical role of us ‘humans’ in shaping the AI tools we use. Creating and curating the AI we want is down to human agency. WE need to do it. WE need to decide what we want, how we want things to work, what we want to prioritise and what we want to deprecate. Everything about how we want our lives, our companies, and our societies to function is OUR responsibility. And is in our hands. Fundamentally the User Experience of the world is down to us.

So instead of thinking what tech you can buy to do what you do now better, faster, cheaper think about what you could do if you could do a lot more. Because with AI you will be able to do a lot more. It represents a massive software upgrade to humanity. Most notably it represents a massive upgrade in intelligence; we can access more, we can utilise it better, and we can spread it around like never before. But this requires a great process of redesign.

Let’s stop aiming at being 10% better - let’s 10X everything.

The Impact of AI on the UK Real Estate Sector's Productivity

Midjourney / Antony Slumbers

Our starting point when thinking about how AI is going to impact productivity across the real estate sector must be to understand what it is that we are dealing with.

AI fits into two main camps; Predictive AI and Generative AI. The former is an analytical tool, the latter a creative one.

Predictive AI is all about ‘Predict, Cluster, Classify’ and relies on learning from a very large corpora of historical data to predict what will happen in the future.

Generative AI is all about ‘Create, Synthesise, Innovate’ and relies on ‘Models’ that have been trained on extremely large corpora of data to create a statistical model that can then be referred to in order to ‘generate’ new data, mostly in the form of language, imagery or computer code.

The difference is critical in real estate. It is only when we have very large quantities of relevant data that we are able to utilise Predictive AI, whereas Generative AI can be used with our own proprietary data but does not require it, and has enormous utility ‘right out of the box’.

Given its nature Predictive AI has powerful but limited ways in which it can impact productivity within real estate. For years people have talked about using ‘AI’ to predict price movements and market dynamics: these have largely failed because the historic data is not large enough to successfully apply Machine Learning to (ML being the bedrock of Predictive AI). This is unlikely to change. So predicting the future of real estate for investment purposes will remain more a data science than an AI project.

However, the potential for Predictive AI to improve the performance (‘productivity’) of actual physical real estate is enormous. Because buildings can be configured to generate a tsunami of data. Given the current and future potential to ‘sense’ the world around us there are endless use cases for utilising Predictive AI to predict what will happen to A given B and C. 

We’ve hardly started optimising how our real estate operates. Ultimately the purpose of real estate is to enable its users to be as happy, healthy and/or productive as they are capable of being. We need to be creating spaces and places that provide the perfect environment for people to do whatever it is that they need to do, when and where they need to do it. It is the user experience of space that really matters. And that is something Predictive AI can really help with. And great user experiences are highly conducive to optimal productivity.

Generative AI on the other hand is much more of a GPT - a General Purpose Technology. Meaning a technology that permeates everything, rather than one designed for a specific purpose. If any task involves language, imagery or computer code then Generative AI has a use. It can help us read, write, create, analyse and understand at a scale and speed hitherto unimaginable. We can all do many things faster, better, cheaper. As a personal productivity tool Generative AI is a superpower.

A way to think about these two types of AI is that Predictive AI is for the quantitative aspects of our industry and Generative AI is for the qualitative. Sometimes we’ll need one, other times the other, and often we’ll need both. Between them there are few areas of the real estate industry that will not be impacted by their existence.

As such, leaning into ‘AI’ is likely to reap huge benefits. At an individual level it sets one apart, at a corporate level it adds differentiation and competitive advantage, and at a national level it could make the UK a magnet for talent, and position our industry as the global thought leader and exemplar of cutting edge real estate. This position is there to be had; no country is standing out as yet.

The real estate industry faces two huge challenges: decarbonising the built environment AND providing the right buildings for the ‘industries of the future’.There is no doubt that AI is going to be a major factor in both. Or should be.

Historically the real estate industry has been a technological laggard. Going forward this has to change. We have very long project cycles. A large development can easily stretch to 5, 10 years. And a decade in tech equates to 100X increase in computational power. So we need to be futurists. What is needed and what’ll have value in the future? Where will we work, what work will we do, where will we live, how will we live? All these questions will be answered through an AI powered lens. Demand, and supply, will be impacted by AI. We will be living in an AI mediated world. 

Leaning into AI should massively increase productivity across the UK real estate sector. It would be hard for it not to. The question is whether the industry has the leadership to make it happen? 

———————-

Written as part of Bidwell’s ‘The Productivity Engine‘ Report. (https://www.bidwells.co.uk/insights-reports-events/productivity-engine-report/)

New Generative AI for Real Estate podcast!

Recently I had the real pleasure of taking part in a podcast for PropTech Denmark during their excellent symposium.

Links in first comment.

I genuinely learnt a lot. Hope you do to.

My co-presenters were:

* Rasmus Juul-Nyholm, Chair, PropTech Denmark & Co-founder, Home.Earth

* Leise Sandeman, Co-founder, Pathways AI

And our host was: Søren Vejby

It’s a very good primer.

Apple: https://lnkd.in/eVKxsvMH

Spotify: https://lnkd.in/eeAcGXqu

Generative AI, PropTech, and Human-Centric Real Estate

Antony Slumbers / Midjourney

I recently presented my ‘𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜, 𝗣𝗿𝗼𝗽𝗧𝗲𝗰𝗵, 𝗮𝗻𝗱 𝗛𝘂𝗺𝗮𝗻-𝗖𝗲𝗻𝘁𝗿𝗶𝗰 𝗥𝗲𝗮𝗹 𝗘𝘀𝘁𝗮𝘁𝗲’ talk, at CREtech London.

You can view it all here:

https://lnkd.in/eY7JiyuM

A rapid, 25 minute version. but we cover the What, So What, Now What? of the subject - what’s going on, why we need to pay attention, and what we need to do NOW.

When something improves 1000 times in just 8 years, you know something big is going down.

100% this is both a BUG and a FEATURE - we have agency, but we really do need to use it.

Hope you like it. Let me know. And of course if you would be interested in me presenting to your company, or at your event, just get in touch.

It is not 2nd best when the best would never have happened.

Antony Slumbers / Midjourney

It is not 2nd best when the best would never have happened.

I have ‘read’ 6 research reports this morning, relating to Skills, Science, AI and Foreign Policy.

A total of 533 pages.

Or rather I have read comprehensive summaries of each of these, the result of fairly elaborate prompts given to ChatGPT 4o.

‘But that’s cheating. Very much a 2nd best approach. You should read the reports A to Z.’

Some would say.

And they’d be right if the alternative was to read each of these cover to cover.

But that’s not the alternative is it? The realistic alternative is to read none of them, or maybe one in full.

Much of AI is about enabling things to happen that simply would not happen without it.

It’s not about replacing human skill, it’s about enabling humans to build bigger pies.

And bigger pies matter. Without bigger pies there won’t be much left for us to do. With them, the upside is infinite.

Something to think about the next time you read or hear about AI making us lazy and stupid.