The idea of incorporating AI into your operations might seem like a daunting task at first. AI is something many of us may not be familiar with in practice, but that doesn’t mean we shouldn’t evaluate how it might fit into our operations. In the not so distant future, we’ll likely look back at this time and wonder how we ever functioned without AI.

But since AI is such a new technology, we’ve compiled some information to introduce you to AI, what it is, and how it can be utilized in a multifamily environment. This is the first of a two-part blog series.

Different types of AI

When it comes to AI capabilities, the industry has defined three different types of artificial intelligence, Artificial Narrow Intelligence (i.e. Weak AI), General AI (i.e. Strong AI), and Super AI. Currently, General AI and Super AI are theoretical and are not currently in practice, so we will focus on the functionality that is available in “Weak AI” tools. 

While it might be referred to as “weak,” that’s a bit of a misnomer as the time savings and efficiencies that can be achieved using these tools are immense. According to IBM, “[Weak AI] can be trained to perform a single or narrow task, often far faster and better than the human mind can.”

Inside of Weak AI, there are two distinctive functionality types, Reactive Machine AI and Limited Memory AI. Reactive Machine AI has no long term-memory, so the action they perform is based on the data that is currently available to them. As a result, the capabilities of Reactive Machine AI are very narrow. On the flip side, Limited Memory AI has access to previous data and can improve and get better over time. 

Lak Lakshmanan, Operating Executive, Data Science for Silver Lake described the evolution of AI like this, “You now have this evolution of going from something that is super simple and very straightforward, very rule-based to learning from data, to learning from unstructured real world information to actually explaining the decision.”

For the purpose of this ebook we’ll focus primarily on Limited Memory AI, which powers generative AI tools like ChatGPT as well as chatbots.

Efficient content creation with generative AI

Generative AI is a type of artificial intelligence that can be used to produce new content, images, audio, and synthetic data. It differs from predictive AI, which is used to identify patterns to improve data driven decisions, and conservational AI, which is used in virtual assistants and chatbots to respond to customer inquiries in a human-like way, in that it produces net new content based on parameters set by the user.

Generative AI isn’t something that is new (it’s been around in some form for years now), but as neural networks and natural language processing advanced, it became much easier to use. To put it simply, neural networks, large language models (LLMs), and natural language processing help you “talk” to the program like a human and it “thinks” and processes information like a human, and this is why it’s able to create content that is more accessible than previous versions of generative AI. This advancement is key because if the content created doesn’t read like a human wrote it, then it won’t be an effective way to communicate with residents and prospects.

With that being said, what are some of the key human characteristics that AI needs to be able to mimic? Lakshmanan noted that humans have senses. They can touch, taste, hear, see, and speak. When he says that AI is “machines acting like humans”, what does he mean? They have ‘senses’ so to speak. They don’t have eyes but they have cameras that give them sight, they don’t have ears or mouths but they have smart speakers that can “hear” and answer our questions. Taste is a little harder, but there are some AI programs that can look at something like the chemical composition of wine and tell what region the grapes were grown in.

Where AI is vastly superior to humans is it is constantly taking in new information and learning and getting better, according to Lakshmanan. When companies how to leverage those learnings to reduce friction and improve the customer experience, they’ll win/gain an advantage over the competition.

The potential benefits of AI

The potential of AI to replace repetitive tasks is significant. Jobs that require low cognitive input and are highly repetitive, such as manual data entry or copy editing, may be taken over by AI. Imagine a property manager who has to manage multiple ILS listings daily; AI can automate the process of listing creation, saving them hours each week. This shift can free up considerable human resources, allowing real estate professionals to devote their time to tasks that demand strategic planning and creativity and improve the customer experience.

“Automation is the goal,” said Lakshmanan. “You're saying that here is … a very manual process and we're going to speed it up perhaps by doing things in a more routine way. And one of the ways in which you can do things more routinely is through AI. You might say, ‘Okay, we're going to basically improve this thing.’ When there are multiple ways that you can make something much more routine and predictable.”

For example, it can help improve the brokerage process to help improve your listings. AI could scour data sources for lists of competing multifamily listings with similar rents, amenities, and locations as your properties and essentially conduct a market survey for the area, helping you to not only improve the quality of your listings, but do it in a fraction of the time if one of your staff had complete this task.

Moreover, AI tools have the potential to automate the asset management process, including property valuation and investment analysis, predicting property values based on various factors such as location, size, and local market trends, which can provide a more accurate and faster valuation process. This will aid property management companies in identifying properties that are not only ideal to acquire, but also when to dispose of assets based on specific market trends that are identified.

The role of AI in communication and marketing strategies is increasingly prominent. LLMs have proven their capability to generate diverse types of content, ranging from website descriptions that attract more leads and emails to property descriptions and social media posts. For instance, a property management company can leverage AI to draft engaging property descriptions that highlight key features, draft emails to potential clients, or generate creative captions for social media posts. This can significantly enhance the efficiency, consistency, and reach of a firm's communication strategy.

Beyond that, AI's capacity for data analysis can be a game-changer for identifying market patterns and risks. A property management company could use AI to analyze historical market data and predict future trends, thereby informing their investment strategy and mitigating potential risks. Future advancements in AI could even extend to rendering designs within minutes and predicting emissions levels and material procurement information, leading to environmentally-conscious and cost-effective building designs.

This is not to say AI is ready to take over all routine or repetitive tasks just yet. When using generative AI to create content, it’s important to review the copy that is created just in case there are inherent biases in the LLM you are utilizing. Additionally, you’ll want to check the sourcing and accuracy of the content. We’re still in the wild west days of AI, so it’s better to be safe than sorry to prevent the possibility of embarrassing the business.

Moving to an AI model that creates consistency and efficiency in repetitive day-to-day tasks lends itself to a centralized model that is more easily managed by a single centralized services director as opposed to multiple on-site property managers.

With AI in the mix, the primary focus would be on building and leading a team of remote customer service representatives that deliver a much higher level of service and performance than on site teams can deliver because they’re not burdened by being pulled reactively in many different directions to ensure residents' needs are being met. This team can focus strictly on completing the task at hand, and doing it much more efficiently because flows are triggered automatically based on behavior.

The key when implementing any new program like this is making sure you have baseline measurements and metrics that you can track against. This baseline allows you to see the ROI, time, and cost savings as they occur in real time. If your standards aren’t being met initially by AI, it’s not time to give up, it's time to reconfigure. Don’t be discouraged. The technology is new and it may take time to fine tune, but at the end of the day it will be worth it in the efficiencies you will gain.