Almost any software you look at is currently touting their AI capabilities, but despite its ubiquitous nature, there’s a lot of mystery still swirling around artificial intelligence and how businesses can utilize it effectively.

So first of all, it’s best to define AI and some of the other buzzwords associated with it.

“AI is anything that artificially mimics decision-making,” says Michael Ding, founder and CEO of Bobyard, an AI-powered takeoff and estimation software. Ding did his AI research at Stanford University and is currently partnered with NVIDIA, Google, and Amazon cloud services. “It has been around for a long, long time. In the 1960s, we had algorithms to play checkers, technically that’s AI. These bots worked by computing a bunch of different possible moves and deducing the most optimal one. The downside was that if you threw something it had never seen before, it would not be able to ‘learn.’ For example, if you changed the rules, you would have to change the entire algorithm by hand.”

A layer deeper than that is machine learning, which is a system of algorithms that can improve itself. A common task machine learning models can do is image recognition like “Is this a cat or a dog?” Letting the model guess and informing it if it was correct or not for thousands or millions of times allows it to eventually decide if a new picture it has never seen before is a cat or a dog.

A subset of machine learning is deep learning where complex algorithms and deep neural nets train a model.

While GPS, self-driving cars and ChatGPT are all examples of different types of AI working in everyday life, there are also services promoting their solutions as true industry AI when it is just a wrapper solution on top of ChatGPT.

“It is important to remember that ChatGPT is generically trained on internet data at large, as opposed to serving the unique needs of specific industries, which requires deeper AI training on that industry’s dataset and careful fine-tuning to achieve the desired results,” Ding says.

Ding adds that the pressure on vendors to implement AI into their solutions is leading to some companies misrepresenting their AI capabilities. Customers need to ask deeper questions on what they are and aren’t getting with a new product.

“AI should provide customers instant results on their task and they are able to revise/edit the results as needed for maximum efficiency,” Ding says.

Use Cases for AI

AI is also far more than just ChatGPT, even though it’s probably the first example that comes to mind when thinking of an AI tool. Because ChatGPT is mostly text-based, it is most effective at automating simple/repetitive tasks, crafting professional messages, meeting summaries, and processing documents or design ideas.

“To truly optimize operations, landscaping companies should look beyond general-purpose AI solutions and explore industry-specific solutions,” Ding says. “For example, there are specialized AI solutions that can automate lead generation or answer customer calls 24/7 or analyze images for design or project planning — capabilities that ChatGPT can’t provide. By leveraging AI tailored to their specific needs, landscape companies can unlock new levels of efficiency and service.”

Ding says there are several overlooked areas where AI can be integrated and immensely benefit landscape companies.

One is improving employee training. By using AI customer personas, employees can practice role-playing or utilize real-time copilots to help them deal more effectively with difficult customer objections or situations during phone calls or email exchanges. 

It can also suggest optimal species based on local weather patterns or help identify and manage plant health and remedies.

“AI can help companies achieve their project targets with real-time adaptive decision-making to avoid the risk of labor or material variances,” Ding says. “Getting these alerts early allows teams to act quickly and avoid bigger problems down the line, saving both time and money. Even better, AI can flag potential issues before they happen, based on past trends, helping prevent problems before they impact the project or customer experience.”

Too often, AI is only considered for helping finetune emails, which is valid, but it can also help extract the details of a complex 100-page document, automate measurements for takeoffs, price materials and labor, optimize crew operations and provide proactive recommendations to avoid negative outcomes.

When considering what AI platforms to use, some of the key features to look for include data security and accuracy.

“For many of the tasks that you would want an AI to perform, it requires you to give it data from your business and you want to make sure your data is not uniquely identifiable,” Ding says.

When selecting a vendor to work with, it’s important to gauge the AI model’s accuracy. Poorly trained models could lead to employees spending more time fixing outputs than manually doing the work themselves. Allowing employees manual overwrite capabilities is a critical feature the software needs to provide.

Also, consider how easy it is to integrate the AI into your company’s existing technology stack.

“When it comes to implementing AI, choosing the right vendor is crucial,” Ding says. “You need a partner who has the know-how, capabilities, and flexibility to adapt to your specific needs. AI is evolving fast, so you want a vendor who can stay on top of the latest breakthroughs, offer custom solutions, and make sure that their tools keep up with your business as it grows. That way, AI becomes a dynamic asset, not just a one-time fix.”