"Automation" is the golden word in business.

The more you can automate, the faster you can move, the bigger you can grow, the more efficiently you can run, and the more effectively you can communicate. Without automation, you're left doing everything by hand, manually operating every aspect of the business--from A to Z.

But automation is really only the tip of the iceberg. The whole idea behind automation is being able to take pre-determined sequences and repeat them over and over again--a rear-view mirror approach, if you will. The next step, however, requires some element of artificial intelligence and machine learning, wherein past behavior can influence future steps.

In short: AI can take it upon itself to do the next thing, without you needing to tell it to do so.

I was recently reading through this viral story about how Facebook had to shut down two of the bots it was developing, when it realized these AI agents had developed a language of their own. As FastCompany summarized, bots are programmed to find the fastest, most effective way to get from point A to B, based on a specific goal or reward. Over time, these bots "learned" that English was not the most efficient means of communication, and slowly started developing their own language to achieve their pre-programmed goal.

Obviously, this story rampaged all over the Internet, yielding declarations from the masses, "I, Robot is here. Run for your life."

I tend to be fairly optimistic about the future of technology, so I chalked this tiny misstep up to trial-and-error learning. But it got me thinking about how AI will continue to evolve, and the business applications it will soon bring to various industries.

I was chatting about this with Greg Pietruszynski, CEO and co-founder of Growbots, a machine-learning sales tool. We were talking about some of the unsolved-for challenges in AI and machine-learning right now, and seeing as Growbots just raised another $2.5 million, putting their total funding at $4.2 million, these are opportunities he is very aware of.

"I think the challenges with AI have 2 layers: the data you use for training your algorithms--and the algorithms themselves. Without the relevant training data, all of the algorithms are useless. AI researchers say that getting data is 90% effort and 10% building algorithms. Many people wonder why Tesla was the first company to introduce the autopilot feature to their cars, and the answer is quite simple: they've been collecting data from all of their cars on the road, while Google has only been collecting data from a few prototypes," said Pietruszynski.

This is the same mindset Growbots has applied to their own AI capabilities, learning from customers' actions to improve upon targeting algorithms. How it works is the platform works as a specialized CRM. As a business, you put in the qualities of your ideal customer, and then Growbots scours millions of company websites, every single day, to find contacts that match those defined characteristics. Users of the platform are responsible for writing the content of the emails, etc., but all of the follow-up emails and sales sequences are automated.

And clearly there is a demand for it. According to TechCrunch, Growbots as a company is experiencing 10 percent growth, month-over-month, with $4 million in annual sales.

"There is a lot to be excited about in terms of the capabilities of artificial intelligence," said Pietruszynski. "When you think about it, connecting with potential customers is the biggest challenge for any company. So, unless you automate that process, what other options do you have? You're either going to spend millions of dollars on ads, or build very expensive sales teams. AI and machine learning is very quickly going to allow smaller companies to compete the big players and their enormous budgets."

The interesting twist to all of this, of course, is what the general public often worries about the most, which is the morality behind AI, bots, machine learning, and more. We're already in an age where highly targeted ads feel somewhat intrusive. Personally, I find it annoying when I call a company's help line, talk to an automated voice, and hear "pretend typing" while they "think" about my statement--as if hearing the sound of a keyboard will make me believe the robotic voice I'm talking to is somehow more human.

There absolutely is a moral and ethical component to technology and innovation, and unfortunately it's a topic that tends to get pushed to the side. But one of the moral challenges tech companies will face is actually in deciding when to tell customers they're speaking with a real person (via website chatbot, for example) or a bot.

"In our case, it's all about saying honestly what is automated and what's not," said Pietruszynski. "While the word 'automatic' might sound like 'lower quality' to some people, it's really quite the opposite. Our algorithms can analyze millions of data points in seconds, so that you can be sure you're only contacting perfect-fit customers and that they get relevant content included in the emails sent to them. Machines are great at analyzing data, so we want machines to do what they do best: exchange data and suggest potential matches so that people have more time to talk to other real people within the sales process."

So, as hot as marketing automation has been over the past 5-7 years, if you want to stay ahead of the curve then where you need to look next is AI. As the adage goes, "If you can automate it, you can scale it."