There is an explosion of AI tech across all sectors. Marketing, tech, and equities trading have been using algorithms for a long time. AI is impacting these sectors already.

Real estate has been famously resistant to technology. Could AI change that?

Can AI deal with something as complex as the development process? Manage fundraising, approvals, and negotiations. Will it create products that human people love?

Can AI provide a competitive advantage if every company uses the same models?

Or will the models move every company to be building the same buildings on the same sites?

Let’s compare equities to real estate. Three things differentiate the two.

1. People. No one lives in equities, but people live in real estate. Cities are the primary way we experience the world. Are equities subject to people’s feelings, expectations, and desires the same way City projects are? Absolutely not. You do not need to consult a zoning review board to purchase shares in TSLA. You just need a Robin Hood account and conviction. Real estate deals are subject to the most fickle variable there is: People. How does an AI compute that?

2. Locality. Also unlike equities, every real estate deal is subject to localized and deliberately nuanced legislation. There is no “one size fits all” federal law for real estate—each county can have a wildly different approach to the same problem set. Take laws in Orlando, Florida, vs. nearby Winter Park for instance—it’s much easier to get a building permit in the City of Orlando but cross the street into Winter Park—and you could be held up in approvals for twice/three times as long. Not to mention when legislation overlaps, or you want to build something that requires an exception or approval.

3. Relationship. AI does data, but to build real estate you need human relationships. No AI in the world can make up for the relationship network that has been established and worked for years.

Building a relationship with a landholder and working out a deal that everyone wins from is an inherently human undertaking.

An AI can not tell you that you can do a JV with the owner of the retail strip—building multifamily units above his holdings, and splitting the proceeds without ever buying the land underneath.

Or getting the local community group to agree to turn their vacant parking lot into affordable homes.

These are the deals that make developers great at what they do—seeing the opportunity where no one else can and building a human connection to make it happen.

Contrast this to the anonymous, high-frequency trades executed computer to computer on Bloomberg terminals.

Even though we see the unique aspects of real estate, at Giraffe we still think there is a place for AI within the sector. We’re helping the community of City Makers (developers, planners & consultants) navigate the transition to digital—what is worth investing time and energy into? And what is not?

Here are our four pillars of digital transformation that we’ve identified as capable of making real change to the bottom line immediately, without wasting time & effort on underdeveloped technology.

1. Consolidating Acquisitions Research

Acquisitions research involves lots of data points, usually trapped in isolated geospatial portals, such as the Government GIS tools, a comp database, a parcel tool (like Regrid), and your own Google Maps tracking of site locations.

With Giraffe, you can consolidate all this data into a single pane of glass & collaborate across all offices to build a pursuit map. This can greatly reduce wasted time on research & eliminate double handling of sites.

2. Paramatising Feasibility Assessment

Feasibility assessment can and should be a mostly self-contained process. Of the hundreds of developers and government agencies we’ve spoken to, far too much of this workflow is outsourced—making it slow and costly to test a site beyond a back-of-the-envelope style calc.

Most multifamily developers do something like this in their heads…

“3 acres at 20 dwellings per acre is 60 dwellings. 60 times an average of $2,500 per unit, less a 25 percent OpEx is $112,500/month or $1.35 million per annum. At a cap rate of 6.5 percent that’s a $20 million stabilized value. Less my costs of $288k a dwelling, that’s $3 million. Less a 10 percent margin that’s $1.3 million to buy the land.”

Anything beyond that “gets done by the civil.” It’s an expensive way to live—and as tools to be laser accurate now exist, that $1.3 million is gonna get outbid by the guy who knows he can build 21 dwellings acres per dwelling within 15 minutes of finding the site.

3. Reducing Soft Costs via Design Standardization

The most significant opportunity afforded by digital tooling is the standardization of designs & the opportunity for prefabrication. In the case of multifamily, consistent unit layouts dramatically reduce soft costs. If this can be solved at the feasibility stage—you can plan with increased forecast accuracy. Not to mention the reduction in soft costs as designs are already accessible and reusable.

In Giraffe we have a design library that can store your optimal building footprints & configurations. Tied to real costs, these instant  “drag and drop” elements can take a parameterized feasibility and link it directly through to detailed design files. Clients we’ve worked with to build out design libraries have reduced their soft costs substantially.

4. Surgical deployment of AI to augment (not replace!) the human process of city making. Taking the above opportunities sets up the use of AI to augment your processes – not replace them.

Building data sets. Ai is perfectly positioned to create new data sets where none already exist. Take the example of OSM GPT – an open-source natural language way to engage with the Open Street Map data set. In downtown Orlando, instead of hunting for a data set that I may never find, I can ask OSM GPT to map ‘all bus stops’ and away it goes. I can now combine that with my other Giraffe tools to build some analysis around the downtown core.

Generating Designs. While some design decisions are best made by a human (for instance how will a project respond to the surrounding environment) some are just best done by a computer. Like car parking or configuring unit layouts. These tasks are tedious—but will make or break your project. An algorithm can sort them out with no problem. Take the below examples as a starting point.

Lifelike renders. Communicating your project is also difficult—but 3D gives a great advantage. Tools like VISOID can take that 3D and a simple prompt to give you life-like rendering—in a minute with a monthly subscription fee instead of exorbitant rendering fees.

These four pillars are how we at Giraffe are approaching the AI revolution. We’re embracing the opportunity while keeping in mind what makes real estate the biggest asset class in the world. It’s the world we live in and experience every day.


By admin

Leave a Reply

Your email address will not be published. Required fields are marked *