Reinforcing Rich Reputation, AI Is Proving Slow To Catch On With Small Businesses

Investment is expensive, expertise is scarce and the changes required to implement disruptive.

The question is whether this will give a few big companies an insurmountable lead - or whether market demand will eventually provide greater access because there is money to be made. JL

Sarah Castellanos and Agam Shah report in the Wall Street Journal:

The high upfront costs of AI tools, scarcity of people who can implement the technology at individual operations, and more pressing IT expenses have widened the gap in AI implementation. But from large technology vendors to startups, are coming up with tools that allow small businesses to use the technology without a data scientist on staff. Even with such tools, it can take time for any company, large or small, to implement a new technology into its business processes.
Artificial intelligence over the past decade has shifted from research theories to actual practices in corporate offices. But it isn’t within the reach of many smaller companies, yet.
The high upfront costs of AI tools, scarcity of people who can implement the technology at individual operations, and more pressing IT expenses have widened the gap in AI implementation. But a range of players, from large technology vendors to startups, are coming up with tools that allow small businesses to use the technology without a data scientist on staff.

Still, even with such tools, it can take time for any company, large or small, to implement a new technology into its business processes. The operational efficiency of an AI system, while desirable, is still far from a priority for many companies.
Potato farmer Andrew Mickelsen says he has observed how large agricultural companies are using AI to sort harvests and make smarter on-farm decisions. But the Roberts, Idaho-based farmer says he doesn’t have the time or money to make AI a priority on his 30,000-acre farm. Instead, the company relies on its up to 400 full-time workers to separate the rotten potatoes.
“We can’t afford on our own to go and spend all the time and money to put it together,” Mr. Mickelsen said.
Much of the challenge comes from how complex it is to implement AI and tailor it to a company’s needs.
Many small firms looking to use AI have to build everything from scratch, said Brad Fisher, KPMG’s U.S. leader for data analytics and artificial intelligence. Even then, the firms may lack the breadth of data to train and test these systems before they are deployed.
“Deploying AI technologies is a lot more complex than tech vendors would lead them to believe,” Mr. Fisher said.
“Many smaller organizations’ IT departments are understaffed, with IT professionals struggling to deal with more pressing concerns, such as updating aging hardware, securing corporate networks and supporting growing tech requests from end users,” said Peter Tsai, senior technology analyst at Spiceworks, a professional network for IT workers.
Only 21% of small businesses are using AI or plan to use it within the next two years, compared with 65% of firms with 5,000 workers or more, according to a Spiceworks survey. The results are based on a survey last year of 780 business-technology buyers at companies across North America and Europe, 65% of which had fewer than 500 employees.
Large technology vendors—from Microsoft Corp. MSFT -0.72% to Alphabet Inc.’s GOOG -0.36% Google—see an opportunity to make AI services and products accessible to small businesses.
Microsoft offers a set of AI models that already are programmed with data. Small companies don’t need to train the tool with more information and instead can build customizable AI-based applications that integrate speech and image recognition, without needing much technical expertise.
That’ll make it much easier for [small and midsize businesses] to embrace AI,” said David Carmona Salas, general manager of Microsoft AI.
Some of Google’s AI services don’t need machine-learning knowledge and can help companies reduce time spent on mundane processes, said Rajen Sheth, vice president of product for Google Cloud AI.
Meanwhile, startups also are coming up with AI-based options for small and midsize businesses.
It can be hard to cut through the noise when it comes to vendors touting AI services. Small businesses should make sure to ask what their definition of AI is and whether it matches the needs of the company, said Whit Andrews, a Gartner analyst specializing in AI.
Kim Taylor, chief executive and founder of Cluster Inc., an online marketplace to match engineering talent to industrial firms, says the company is taking a gradual approach to implementing AI to automate tasks and communicate with customers.
“It’s a behavioral shift. All these companies are under pressure to adopt technology they don’t understand. It can feel daunting,” she said