First mover advantage will be significant for those that deploy AI at scale
ORGANIZATIONS understand the value of artificial intelligence (AI), and in the past 12 months, many, if not most, have attempted to deploy AI in one way or another.
According to a recent forecast by IDC, spending on AI systems is forecasted to reach US$97.9 billion in 2023, more than 2.5 times the US$37.5 billion that is expected to be spent in 2019.
IDC’s analysis reveals that spending on AI systems is currently dominated by the retail and banking industries, followed by discrete manufacturing, process manufacturing, healthcare, and professional services. Of course, spending on AI is growing rapidly in the media industry as well as among federal/central governments.
Despite the spending, it seems as though a majority of the organizations are running AI projects in pilots and are unable to scale them up effectively.
There’s no doubt that isolated AI implementations have yielded significant returns for individual teams or departments, but to unleash the full potential of AI, projects need to be scaled across the organization.
To better understand the need to scale AI projects and the advantages it offers, Tech Wire Asia spoke to IDC Big Data and AI AVP Christopher Marshall.
“AI to date has generally focused on specific narrow applications relevant to augmenting/improving the activities of individual workers or consumers.
“AI at scale leverages the learning of these narrow applications of AI to support many individuals simultaneously, say by sharing learning experiences from one doctor to many, or sharing security threats identified by one camera to the enterprise or nation as a whole.”
Of course, in order for organizations to truly scale AI, they need the right infrastructure in place — and a digital-first DNA to drive the change. It’s why champions or pioneers of enterprise-wide AI projects gain a strong competitive advantage in the marketplace.
Truth be told, the tools and technologies that help deploy and scale AI projects are becoming much easier to access and use — even in the absence of data science and AI skills. The ease of AI also means that there’s a definite first mover advantage that accrues to companies that dare and go full throttle on AI projects.
“Scale economies tend to accrue to first movers, and second movers find it difficult to catch up.”
However, Marshall believes companies need to understand that running AI projects at scale provides vastly increased business value for individual users, including reduced costs, greater revenues, and more insights.
“This further increases the extent to which these tools are used, and therefore increases the amount of data and the quality of the AI models used, and still further increases the value created by the AI models.
“This virtuous cycle is the holy grail of AI at scale and what make relatively minor improvements scale up to game-changing innovations.”
Of course, the benefits of deploying AI at scale bring along some challenges — but they need not hinder the pace of progress in any way.
“Data quality, privacy, ethics, transparency are important challenges in any large scale data-driven enterprise project and the tendency of AI to produce results that are complex to understand makes it difficult sometimes for end-users and customers to trust the AI models.”
The solution, according to Marshall, starts with acknowledging the problems and instituting enterprise strategies, especially around data and analytics to safeguard data quality and customers’ rights above all else.
Going into 2020, organizations that are planning to invest in AI need to remember to scale their projects quickly and capture the first mover advantage if they want to maximize returns and market share.