Industry executives and experts share their predictions for 2023.  Read them in this 15th annual VMblog.com series exclusive.

Five Predictions for AI and NLP in 2023

By Luca Scagliarini, chief product officer, expert.ai

According to Forrester Research, companies using advanced data analytics to drive decisions are eight times more likely than beginners to see revenue growth of 20% or more. This makes communications to customers - contracts, emails, reports, digital archives - promising assets for process automation. Natural language processing (NLP) enables AI to understand and utilize unstructured language data. In doing so, gaps can be eliminated between humans and technology, enabling assets to be fully leveraged and new insights to arise.  

It's a fast-developing area and we closely track issues faced by companies implementing NLP models, industry trends and challenges on the horizon. Independent research commissioned by expert.ai surveyed current NLP practitioners in technical and business roles across Europe and North America. With this data, we've been able to get a peek at where things stand and where they're heading, resulting in the following five predictions for 2023.

Budgets to Get Bigger  

According to the survey, 77% of respondents said they expect spending on NLP projects to rise over the next 12 to 18 months. Just 13% felt spending would remain the same, very few anticipated a decrease in spending. For those expecting a bigger budget, the increases are sizeable: Nearly 40% project a 10% bump, 35% of respondents saying it could reach 20%. And there is a direct correlation between increased spending and NLP maturity. 

Deployments Growing

Four out of five organizations surveyed now have NLP models in production. Most projects are relatively new, with nearly half in production less than two years. Nearly 50% of respondents noted that data protection and governance (GDPR and PII) use cases are their primary focus. Over 70% reported they support multiple text analytics use cases. Deployment momentum is on an upswing and will expand considerably in the year ahead.

Challenges and Considerations  

Respondents were clear on the top hurdles businesses encounter when adopting an NLP solution. With both at nearly 40%, the two key issues are being able to align with stakeholders on which cases take priority, and justifying costs associated with NLP modeling and tools to show the value they deliver. How far along an organization is on NLP development impacts these challenges. While those evaluating NLP use cases are most concerned with data security and governance, businesses just starting to experiment and build NLP models feel choosing which AI approach will yield the desired model results (47%) is the most critical decision to make - and many will choose their direction in early 2023. 

Better Together

For NLP use cases currently in production, over 50% of respondents use a mixed machine learning (ML) and symbolic/rules-based approach. Nearly 80% employ ML with either deep learning or graph AI to drive their insights. Solutions that combine the best of respective techniques are reaping even higher quality, more accurate results. Not surprisingly, the majority of organizations are taking a hybrid approach, whereas just 6% are using ML on its own. Hybrid will emerge as the approach to follow in the year ahead and foreseeable future.

ROI and Responsible AI

Over half of those surveyed say their organizations measure return on investment (ROI) for NLP projects based on time to production (54%), efficiency improvements (53%) and cost reduction (53%). Efficiency improvement, in particular, can vary by geography. For instance, while 61% of companies in North America use efficiency as a metric, the figure is only 36% in Europe. Other measures used include competitive advantage, lowered risk and increased revenue. Still, in the year ahead, greater cost efficiency will be the overarching way to evaluate NLP ROI.

Yet, it's not always just about efficiency and money. More than a quarter of respondents said they strongly agree that when planning natural language projects, they consider how responsible the AI approach is going to be. Will results be explainable and unbiased? Will the process be energy efficient? Such concerns are on the rise and will come to the forefront as technology and practices continue to mature.

Fortune Business Insights projects the NLP marketplace will be worth $127 billion by 2028. Considering this boom, increased budgets, and that 80% of those surveyed now have NLP models processing tens of thousands of documents every month, the space will be very dynamic in the year ahead. Expect changes and challenges, but most of all, expect growth to continue in 2023 as NLP carves out the future.

##

ABOUT THE AUTHOR

 

As Chief Product Officer (CPO), Luca is responsible for leading the product management function and overseeing the company's product strategy. Previously, Luca held the roles of EVP Strategy & Business Development and CMO at expert.ai and served as CEO and co-founder of semantic advertising spinoff ADmantX. In his career, he held senior marketing and business development positions at Soldo, SiteSmith, Hewlett Packard and Think3. Luca received an MBA from Santa Clara University and a degree in Engineering from the Polytechnic University of Milan, Italy. He blogs regularly about the real-world applications of AI and cognitive computing for today and tomorrow.