McKinsey: Applied AI Is a Top Tech Trend in 2022
- By Paul Mah
- September 01, 2022
A new report released by McKinsey identified Applied AI as one of the 14 significant technology trends unfolding today, with an investment of USD165 billion made in 2021.
McKinsey’s findings were outlined in the “McKinsey Technology Trends Outlook 2022” report, with investments in Applied AI outstripped only by clean energy, Web3 technologies, and mobility.
Growing industry relevance
As defined by McKinsey, Applied AI is the use of intelligent applications to “solve classification, prediction, and control problems to automate, add, or augment real-world business use cases.”
In a nutshell, Applied AI is about solving business problems and delivering actionable insights that make a difference. And as AI technologies push the frontiers of innovation, business adoption will continue to grow, says McKinsey.
And business leaders should pay attention, as AI has already made itself felt where the bottom line is concerned. Indeed, 27 percent of respondents surveyed in 2021 report at least 5% of EBIT being attributable to AI. Two-thirds (67%) of respondents reported a revenue increase through AI adoption, while 79 percent report a cost decrease via AI adoption.
Technology-centric industries are leading adoption for now, with product and service development, service operations, and marketing and sales being the business functions leading the adoption of AI, says McKinsey.
What is breathtaking is the sheer breadth of industries that are impacted by AI since its early days of dubious applicability. Today, AI impacts anything from high-value industries such as automotive and assembly, pharmaceuticals, and aerospace, to sectors that might not intuitively make sense such as agriculture, healthcare, and education.
Some ways AI is set to make an impact:
- Agriculture: Optimize processes through capabilities such as productivity forecasting and driverless tractors.
- Healthcare: Enhance healthcare services through automated pathology recognition and diagnosis decision support.
- Education: Offer personalized learning based on a student’s progress.
- Public sector: Leverage AI/ML to expedite delivery of key services.
Barriers to AI
But while the increased access to AI and ease of implementation will likely lead to new technologies and practices that will make it more readily available, various risks and uncertainties might yet slow down its adoption.
Specifically, the high up-front investment in talent and resources threatens to create a high barrier of entry to developing ML workflows for production, says McKinsey. Moreover, cybersecurity and privacy concerns are growing, too, with data risks and vulnerabilities occurring across the AI workflow.
Indeed, 55% of survey respondents cite cybersecurity as a leading risk in their business in 2021 and are actively taking steps to mitigate it, notes the report. AI ethics is another consideration – and which we have covered more than once, and which encapsulates issues such as responsibility, equity, fairness, and explainability.
Finally, increasing regulation could yet hinder the development of AI. Already, the EU had proposed more regulations with heavy restrictions on a range of “high-risk” AI systems last year, while China has unveiled ethical guidelines governing AI algorithms to guide the use of AI technology by Internet technology giants and enterprises.
Implementing AI
Of course, implementing AI in a production or commercial environment is quite different from that of a trial. For businesses, this means not just hiring (or training) the right expertise, figuring out where AI can benefit the organization, developing ML models, and implementing it in production systems. That is easier said than done.
“The biggest shift affecting AI’s broad adoption is tied to more mature tooling and the emergence of a canonical tech stack that is drastically simplifying how AI solutions are engineered and integrated with other digital applications,” noted Jacomo Corbo, a partner at McKinsey.
“AI is quickly becoming more consumable, and solutions that use AI are accessible even to organizations with few to no AI engineers of their own,” Corbo said.
Paul Mah is the editor of DSAITrends. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose. You can reach him at [email protected].
Image credit: iStockphoto/Shutter2U
Paul Mah
Paul Mah is the editor of DSAITrends, where he report on the latest developments in data science and AI. A former system administrator, programmer, and IT lecturer, he enjoys writing both code and prose.