Turbulent take-offs stymie enterprise AI, ML adopters

Dive Brief:

As artificial intelligence and machine learning capabilities bring greater business intelligence, security, cost savings and innovation to businesses, many continue to lack fundamental building blocks, such as talent and digitization, that enable value generation at scale, according to recent reports.
Businesses that want to realize AI's potential need to scale its application across the business units and functions and build a strong digital backbone for the enterprise, according to McKinsey & Company. A lack of clear strategy and talent and functional silos were identified as the biggest barriers to AI adoption. Putting key technology enablers in place — such as top-management sponsorships, efforts to tighten talent gaps, a clear strategy and a portfolio of AI opportunities — can help businesses establish foundational practices for valuable AI.
Early adopters are making machine learning a priority in the budget: 26% of these companies reported that ML takes up 15% of the total IT budget, according to a Google Cloud study in conjunction with the MIT Technology Review. Early ML adopters are using the technology for cybersecurity, risk and fraud analysis; asset management; and customer services. But the majority of executives, 82%, report that predictive analytics is having the most impact today, according to Google Cloud studies with IDG and Harvard Business Review.

Dive Insight:

From cybersecurity and data analytics to R&D, early adopters are touting the promises of AI and ML in enterprise settings. While most businesses have at least a strategy or early implementations, the technology is far from a meaningful, enterprisewide application across industries.

Cloud computing is helping companies overcome the high barriers to AI and ML. The cloud helps executives deploy AI and ML with integration capabilities for new tools and platforms, more flexibility in process and vendor choice as well as faster deployment and iteration, according to a Google Cloud, MIT survey.

In operations, cloud-based workloads are making processes more efficient, driving down costs and improving productivity and time to market.

The real power behind cloud ML environments reaching new maturity is that it allows organizations to take advantages of AI technologies without having to have a data scientist, according to Daniel Castro, VP of the Information Technology and Innovation Foundation, in an interview with CIO Dive.

Most companies are not using AI extensively, but accessibility will help steady experimentation and adoption.

In addition to the foundational building blocks laid out by McKinsey, businesses also need to continue focusing on data quality. For newcomers and experienced businesses trying to embark on the AI journey, "data is the Debbie Downer of any AI project," according to Forrester Research.

Investments in information architecture will continue as businesses that experiment with AI realize that the promise of the new technology can only be met with investments in the data environment.