McKinsey’s latest AI research predicts it could create trillions worth of value … someday | Tech News

McKinsey Global Institute’s latest research on artificial intelligence focuses on how AI will affect business, analyzing 400 use cases — from airlines avoiding flight cancellations to online retailers recommending items — across 19 industries.

In aggregate, McKinsey projects that AI could eventually drive between $3.5 trillion and $5.8 trillion of annual economic value in those industries — a wide range, and with no timeline on that forecast.

Key points:

In general, AI — once it is further integrated into business — has the potential to drive growth. In 69 percent of the use cases McKinsey studied, “deep neural networks can be used to improve performance beyond that provided by other analytic techniques,” according to the report.

But it’s not a 10x change — at least for current lines of business. In the 19 industries McKinsey focused on for the report, it estimated the impact of AI somewhere between 1 percent and 12 percent of annual revenue.

For context, the report focused on ways that AI can add value in industries providing existing products and services, rather than inventing new businesses — so, for example, how a car company can get better at making the cars it’s already making, rather than launching an entire new AI-driven business like a fleet of self-driving cars. (We’d add: Those new businesses seem to be where the biggest potential for growth lies, but it’s harder to predict.)

Among business functions, AI could drive the most impact in marketing and sales and supply-chain management and manufacturing. For example, AI could improve personalized recommendations for an e-commerce company, or predict traffic patterns to reduce trucking costs.

Among the industries McKinsey analyzed, the one where AI could create the most value — as a percentage of total revenue — is travel, where it could potentially drive around 7 percent to 12 percent of impact as a percentage of annual revenue, according to the report.

Why travel? McKinsey Global Institute partner Michael Chui, the main author of the report, says that travel companies perform complex marketing, sales and operations, all areas where McKinsey sees AI creating the most value.

Other industries where McKinsey projects AI to have the biggest relative impact: High tech, defined as software, online and hardware companies; transport and logistics; and pharmaceuticals. It’s projected to have a lower impact on the public and social sectors, oil and gas, and chemicals.

Most companies are not already using AI. McKinsey cites its own research from last year, noting that “even among AI-aware firms, only about 20 percent are using one or more of the technologies in a core business process or at scale.” Key hurdles: Obtaining massive data sets for training, the threat of legislation and potential bias in data and algorithms.

Read the full study here.