Imagine searching through the hundreds of thousands of pieces of footage in the National Football League's history collection to uncover a certain image. More than 16,320 minutes, or 272 hours, of game film are produced in a single season. There is an apparently limitless quantity of film if you add coverage of every practice, halftime show, and postgame show, as well as every pregame, halftime show, and postgame interview. And it only applies to one season.

The NFL teamed with Amazon Web Services in December 2019 to utilize artificial intelligence to search and classify its video assets. This will make it simpler for staff members to produce highlight reels and other media from all of this content. The NFL's content development team had to educate the AI on what to look for as an initial part of the process.

The NFL's content development team had to instruct the AI on what to look for in the process's initial stage. Every player, team, jersey, stadium, and other visually recognizable content that the team wished to label inside its video collection received its own set of metadata tags. It then integrated those tags with the image-recognition AI system already in place at Amazon, which Amazon had trained on tens of millions of photographs. The AI used both sets of data to identify pertinent visual content in the video collection, and the content development team quickly approved each tag. Each video used to need to be manually searched for, located, and clipped before being stored in a repository and tagged with information.

In a previous HBR article titled "Collaborative Intelligence: Humans and AI Are Joining Forces," we discussed how some of the world's most successful companies are defying the stereotype that technology will make people obsolete by harnessing the power of human-machine collaboration to transform their industries and boost their bottom lines. Now, a number of businesses are using this strategy to not just out-innovate their rivals but also to fundamentally alter the entire essence of innovation as it has been practiced over the preceding ten years.

For instance, in the NFL's situation, AI sped up the image-recognition procedure, but the system would not have worked without staff members deciding which data needed to be submitted and then authorized. The NFL also didn't just delegate the task of creating highlight reels to AI; instead, content creators handled it, but they were able to do it more quickly and efficiently since AI has the unique capacity to swiftly filter through enormous amounts of data.

Assumptions regarding the fundamental components of innovation are altering as a result of the new human-focused approach to AI. Businesses like Etsy, L.L. Bean, McDonald's, and Ocado are rethinking how automation and artificial intelligence (AI) may connect a variety of cutting-edge information technologies and systems to allow dynamic flexibility and seamless human-machine interaction.

These innovative companies have historically made high investments in digital technology in an effort to meet the fast-changing needs of their customers and take on new operational difficulties. According to a 2019 Accenture study of more than 8,300 organizations, they have significantly boosted expenditures in cloud services, AI, and similar technologies, and they are producing income at a rate that is twice as fast as laggards. According to a second survey of more than 4,000 businesses in 2021, the 10% of businesses investing the most heavily in digital technology are outpacing the rest of the market by a factor of five with respect to revenue growth.

In order to succeed in a future where most businesses will owe their success to people rather than technology, we have converted the knowledge we've gained from this study into advice that business leaders can utilize. Five components of the developing technological environment are highlighted by our IDEAS framework: intelligence, data, expertise, architecture, and strategy. It may aid executives, both technical and nontechnical, in better comprehending those components and imagining how they can be combined to create potent innovation engines.

In this article, we utilize the IDEAS framework to analyze case studies of companies that have applied human-driven AI processes and applications to address issues in robotics, e-commerce, and other fields. You may act in a similar manner by mobilizing the abilities and knowledge of your own workforce.

Make AI less robotic and more human by using intelligence.

Artificial intelligence and human intellect work best together. Even the youngest people can learn, grasp, and interpret with an efficiency that no AI-powered computer can equal. If you drop something by accident and a one-year-old sees you reaching for it, they will pick it up for you. If you deliberately throw it down, the youngster will disregard it. In other words, even very young toddlers are capable of understanding the existence of people's intentions—a remarkable cognitive talent that appears to be almost hardwired into the human brain.

Not just that. Young children start to gain an intuitive understanding of physics at an early age.

They anticipate that things will travel in straight lines, endure, and collapse if left unsupported. They discriminate between live agents and inanimate objects even before they learn language. They have a remarkable capacity for generalization from sparse instances as they acquire language, taking up new words after hearing them just once or twice. They also discover how to walk independently through trial and error.

In contrast, AI can perform a wide range of tasks that humans, despite possessing natural intelligence, find challenging or impossible to do well. For example, AI can identify patterns in massive amounts of data, outplay the world's best chess players, manage complex manufacturing processes, and simultaneously handle a large number of customer service calls. It can also analyze weather, soil conditions, and satellite imagery to help farmers maximize crop yields. Most importantly, AI has made it possible for people and robots to collaborate effectively. And in contrast to predictions about the demise of jobs due to automation, such collaboration is generating a variety of new, high-paying opportunities.

Human employees are instructing a new generation of robot pickers at Obeta, a German electronics wholesaler whose warehouse is handled by the Austrian warehouse logistics business Knapp, on how to handle objects of various sizes and textures. The robots use a vision system, a suction gripper, and an industrial arm that is readily available. Importantly, they also include AI software from California-based start-up Covariant.

Knapp employees educate a robot by presenting it with strange items to see whether it can effectively adapt to them. When an attempt fails, it might modify its perception of what it is seeing and try new tactics. When it succeeds, a human-programmed reward signal is sent to it to help it remember what it learned. When a group of SKUs deviates significantly from