Fascinated by computers since his Army experience in the mid-1980s, Julio Barros is now an AI engineer and advisor and coordinated a regional ‘AI in Production’ conference in July of 2024. 

What we now call AI was formerly referred to as “machine learning,” and has been around for about 60 years. Back then, a group of computer scientists gathered to discover how it works and what it can do, and according to Barros, scientists have been working on it ever since. 

Over the years, the definition of AI has meant different things. Some AI abilities like spell- and grammar-checkers were once a big innovation, and are now woven into daily life for just about everyone.

Recently, and largely referring to generative aspects, Barros believes the definition of AI is much bigger than that. It includes a lot of planning, optimization, logic, and more sub-disciplines, some with an emphasis on helping computers make decisions with imperfect information.

“The real strengths of the generative aspect of AI are language models and their ability to transform information,” Barros said. “It can be a great way to summarize notes or pull out points to highlight. It mainly affects the text and image portion of a business, and every business has those parts incorporated into it.”

This type of technology can be useful to businesses in any sector, but are more practical in some areas than others. Clearly, AI cannot be used in addressing any physical work that needs to be done. It can go a long way in helping companies better understand patterns of invoices and inventory, forecast a variety of data sets, and process user requests in marketing, sales, and customer support departments. It can now be easily used in generating messaging towards consumers and target markets.

The current number one use of AI in business is developing new software codes. Even if the generated code is not perfect, it helps projects move along faster, though not all computer programmers would agree, according to Barros. Other popular uses include the automation of customer support as well as helping staff find answers to questions in document sets without using specific keywords.

“AI is getting really good at trying to understand what a human is asking for and trying to find where in the documentation that answer may be,” Barros said. “Whether businesses expose that to their customers, to the public, or use it for internal support, the ability to find answers to questions really quickly can be very useful.”

While these systems are powerful, Barros believes many are not yet being used to their full potential. Some people criticize generative technology because answers generated are not always correct. This can cause frustration for those expecting only correct responses, but it can be helpful if users can utilize select generated information to spark additional ideas and questions. 

Computer scientists are continuing to figure out how to combine multiple types of AI systems and encourage more reasoning abilities for users to get more out of these tools. While these may not be incorporated by all types of businesses, those on the cutting edge will likely work with these more advanced systems in the near future.

Right now, many AI users are asking basic questions without realizing less helpful answers can be augmented. Answers may be mediocre because they are reproductions of the average answers to these questions, but a human can ask for the answer to be reframed. 

“You can ask AI to make an answer shorter, funnier, for adults only, or whatever you want it to be,” Barros said. “The more you ask for and the more specific you get, the better your answer will be. You can ask it to help in brainstorming and creating plans, and you can chain these things together to come up with a list of things to do, start addressing these items, and actually accomplish them. This will likely become a more popular and more useful part of business because it can process full plans in multiple stages, then people can loop back and adjust accordingly.”

The wide variety of uses and so many possibilities may seem like added complications, but Barros said it can simplify some operations allowing businesses to do more. Creating efficiencies can lend itself to increasing capacity and goals, but should simplify operations in the long run as people become more familiar with all the different capabilities.

“There are so many use cases in so many specifications, it’s hard to say where AI will go,” Barros said. “By becoming familiar with the current tools and what they can and cannot do, as different tools come out, it can become easier to integrate into your life, and you’ll be in better shape to really take advantage of it.”