Enterprises appear to be very optimistic about the potential of artificial intelligence in general and machine learning specifically to drive innovation and revenue in 2022.
This is borne out from new research comes from SambaNova, titled “The Race to AI Value: Scaling AI/ML Ahead of Your Competition”. The organization surveyed 600 global AI leaders about their AI progress.
The research reveals that while many businesses realise the potential that artificial intelligence can delivery, too many businesses struggle with scaling artificial intelligence initiatives. This includes challenges centred on customizing artificial intelligence models. In addition, there is insufficient computing infrastructure together with a distinct lack of trained talent.
In terms of how many companies are making progress in the area of artificial intelligence, 26 percent of top organizations indicate they have it figured out and their advanced technology initiatives. Moreover, the inference is that technologies with some forms of artificial intelligence have reached widespread production. This is up from 12 percent when a similar group of companies were surveyed in 2021.
As to why there are so many slow adopters, the data finds that 35 percent of enterprise leaders cite insufficient computing infrastructure to handle intensive artificial intelligence.
Even for those who can handle the current generation of artificial intelligence, 53 percent of respondents indicated they will run out of computing power within the next decade. To address this, new computing architecture is required.
Another barrier to progress is with workloads, either in terms of time, financial resources, lack of skilled personnel, or insufficient people to assist with the necessary delivery of the technology.
The issues of skills is captured in the fact that 28 percent of enterprise leaders are blaming the lack of trained talent for holding back progress in introducing artificial intelligence into the firm.
More specifically, 42 percent of companies say they either lack enough artificial intelligence knowledgably engineers on staff or they lack an adequate pool of potential applicants or they are suffering from some combination of both.
In terms of moving forwards, the design space is called out as important and this area appears to be in better control. Here, 50 percent of enterprise leaders note customizing models for unique needs as the top challenge when scaling artificial intelligence and machine learning initiatives.
These companies also report they are gaining a competitive edge, moving ahead of firms that have not adopted new technologies.
