Despite significant hype surrounding the use of AI most enterprises are not actively pursuing its use in the short term. A new survey by chipmaker AMD found that 86% of businesses plan to prioritise the technology in five years but only 14% plan to introduce it within six months. However, two thirds say AI-enabled tech will make employees more efficient.
While AI has been part of the enterprise landscape for decades, the past year has seen the hype surrounding it grow significantly. This is in part in response to the success of ChatGPT from OpenAI and companies like Microsoft, Google and Salesforce rapidly adapting high profile products to include generative and foundation model AI.
For the survey AMD questioned 2,500 global IT decision-makers to get a picture of attitudes towards AI in enterprise. The report authors found that while IT leaders recognise its potential and are ramping up investment, more say AI is moving faster than enterprise can handle.
Many are working on mid-to-long-term integration of AI across businesses, with five year and longer plans well advanced. The big gap is between current hype and the reality of doing business at scale. According to AMD, IT decision-makers “lack a clear roadmap for implementation,” particularly around its use in hardware infrastructure.
Enterprise grade AI not yet a reality for most businesses
It is expected the market for enterprise AI will grow exponentially in the coming. A report published in April by Market Research Future suggests it will hit a value of $155.2bn by 2030 with a 34.6% CAGR. A separate report by Bain and Company published in June found that system integrators and marketing companies were among the earliest adopters of generative and foundation model AI.
But AMD’s research shows this level of spending is still some way from coming to fruition. It found that more than half of those surveyed said their organisations didn’t have the IT infrastructure to effectively handle AI workloads. Some companies like IBM, through its watsonX platform, as well as Databricks, are offering AI compute on company data, while hyperscalers like AWS and Azure offer full stack AI solutions.
More than half of UK-based IT decision makers, and just 46% of global leaders working for larger enterprises with more than 5,000 employees said their organisations weren’t willing or ready to adopt AI. This is in part due to the speed of change in the technology. Just 19% say their organisation plans to prioritise adoption of AI within the next year.
There were three main reasons IT leaders gave for showing caution in rolling out AI in their organisations: potential security risks, lack of infrastructure and the burden of training staff.
Of those working for organisations already deploying AI solutions, 90% reported increased workplace efficiency. “There is a benefit to being an early AI adopter,” said Matthew Unangst, senior director for commercial client and workstation at AMD. “IT leaders are seeing the benefits of AI-enabled solutions, but their enterprises need to outline a more focused plan for implementation or risk falling behind.”
Read more: Is AMD narrowing the AI gap on Nvidia?
