Algorithms

- The New Project Fear?

2019 is now fully ensconced and I am beginning to relax when entering dates after having inadvertently returned to 2018 on a couple of occasions. In many ways it wouldn’t make any difference. Brexit isn’t much further forward as I write this, although the sense of panic is certainly more intense than it was last January. The talk of advancing technology is still largely about what might be, as January hosts outpourings from the CES event in Las Vegas. The toys are now being prepared for Christmas 2019 in the form of rollable, foldable screens and such like, alongside are the weaklings that will have sunk without trace before Santa’s elves even get to work. I have little idea of winners or losers but it seems that the elves forgot all about virtual reality headsets last month.

All that aside there are one or two things that appear to be coming along quite nicely, largely out of sight of the elves. Chief among these is 5G technology. This promises (reality check: much more promise than hardware right now) to be a massive step forward in mobile connectivity with implications far beyond getting that cute cat video to your phone more quickly. Both capacity and speed of data transfer without a physical connection will increase exponentially. The UK Government are firmly behind 5G rollout, demonstrated by the legislative power it has awarded the telecoms companies. There are noises coming out of the US suggesting that the country is determined to “win” on 5G. Sounds odd to me but there are lots of things I don’t understand about the US right now. So why does 5G matter?

There are two inventions that have been waiting for their moment to fledge: The Internet of Things (IoT) and Autonomous Vehicles (AVs). AVs are well publicised and should be familiar to you as concept images, if not as real objects…. yet. IoT on the other hand is a rather clumsy name for millions or, (more likely) trillions, of mostly tiny sensors attached to mundane everyday appliances, machinery, conduits or even people. Each of these devices collects and sends the measurements that it takes to anyone or anything that is connected to the internet. Scientists and engineers created sensors that didn’t need a power cable and shrunk them to such a degree that they could be placed almost anywhere they can find something to measure. Is there liquid in that pipe? Is it flowing? How fast is it flowing? Which direction and what is its temperature?

Demand ballooned and production costs fell to a point where the only major restrictions on their use were: 1. having enough addresses to identify each one and 2. how to collect the stream of information. The techies have now provided a way to expand the number of addresses to satisfy at least this first wave of expansion. That left the need for a way to get the data without wiring every sensor or manually reading it. Wifi is fine if you are inside a building that already has an internet connection, but it would be far better if the sensors can use existing, public networks to send their numbers. That way they can be located in out-of-the-way places or attached to moving objects.

That is where 5G comes in, the equivalent of railways from the coal mine. There will be a stream of numbers, way beyond the vast amount we already have, travelling through “mobile phone” networks. This is Big Data and, once gathered, all these snippets can be assembled to give insights into events, past, present and future, helping improve the product or use it more efficiently; whether “the product” is yours or not. To make that work you need an algorithm, a word which now seems to strike fear into the hearts of entire industries, even property. Big Data is pretty useless without algorithms and algorithms are unreliable without the data.

An algorithm is simply a bit of programming language which uses a set of rules to calculate or make a decision based on the information it receives. The greater the quantity of (good) information the algorithm receives, the better it will be at achieving its objective. This is the basis of all computing and therefore not new. Streetlights that supply information to help autonomous vehicles find places to stop or reduce their intensity when only autonomous vehicles are present?

The stars are aligning, even real estate will be moved by this force because all aspects of its operation and use can be monitored, assessed and adapted both globally and locally. The fundamental concepts behind property value will be put under a microscope like never before. Location will be re-evaluated, even if the final answer remains the same; but what are the chances of that? Our industry has changed very little in the last 40 years, so the scope for unearthing wrong turns could be considerable. It seems to me that commercial real estate has, so far, been about small data. The number of transactions, for example, is relatively low and the reliability of the resulting data has often been poor. The nature of transactions is overdue for a major overhaul, as is the quality of information. That was in my previous blog.

This gives context to two events that I attended during the closing weeks of 2018. The first was the annual Harris Debate at the RICS headquarters in Parliament Square. I can’t be sure what the invitation policy was but there seemed to be quite a number of empty seats in the imposing oak panelled auditorium. Notwithstanding a sizeable online audience, I would have thought that this was worth an hour or two of any professional’s time because the title of this year’s event was “What future for professions in a world of algorithms?” The annual debate series, which forms part of the excellent work of the Harris Foundation for Lifelong Learning, is themed around ethics and professional standards, bringing that discussion into the twenty first century. The debate that took place was appropriately existential in nature but there seemed to be little appetite for picking up the opportunities being revealed by the speakers and exploring how they could be used for common or personal advantage. Restraining excesses was very much the mood; algorithms are a problem.

The other event was a seminar organised by Revo (the retail property industry group, formerly known as the British Council of Shopping Centres) and The Investment Property Forum entitled “The future of the car and its impact on investment and property”. There were notable contrasts to the RICS event; there was standing room only in Hammerson’s conference suite and the questions did not pre-suppose that the AV genie could be pushed back into the bottle. This will bring opportunities and give us a reason to develop good algorithms. An unstoppable step-change in transport technology means that the world designed for cars as we currently know them is going to change and, given the extent to which that has underpinned virtually all our thinking since the start of the 20th century, it is a radical thought. A bit scary too.

The further tolling of bells caused by this week’s Christmas retail figures adds urgency to the need for lateral thinking in retail property. The elephant in the office is The We Company (formerly WeWork) with its apparently high-risk strategy of leasing long and letting short but with a strong branded approach growing on a previously unseen scale. This is a familiar idea pumped up by massive investment, but if it succeeds could result in property owners buying (or being bought by) hospitality groups. That would surely start to show a new attitude to fundamentals in the property industry and the use of proper Big Data should be central to that. Finally, residential property is heading towards its own renaissance with large scale, respectable investors working on providing homes designed to be rented on a long-term basis. The impact of raising the overall standard of management and adding branding in this process is key to future investment in homes. All of these activities will be driven by algorithms using data from those tiny little sensors and records of what people do and don’t like.

This is all potentially exciting, but it isn’t just a spectator sport. We need to evaluate the pros and cons realistically, but from a positive perspective. Look for reasons to do things rather than excuses not to change. These new technologies will shape how you, your clients or your businesses work, play and make decisions. Good luck in 2019 everyone – I reckon algorithms are not to be feared, they just need some understanding….and discipline, definitely discipline. What do you think?