Famed British philosopher Alfred North Whitehead once remarked about the First Industrial Revolution that “the greatest invention of the nineteenth century was the invention of the method of invention”. The new industrial age was unlike all preceding eras of history, not just because of the technologies it brought, but the processes that accompanied them. Whereas breakthroughs in science and engineering, from the rifle to the printing press, would once arise more or less at random, and proliferate unevenly, now there was a stable and predictable trajectory from ideation to development to deployment that would occur. From this was born the term used so often today: innovation.

Innovation, then, is not simply invention. Developing a new idea, theory or product is a necessary, but not sufficient, condition for innovation to occur. Missing are the practical and scalable components that transform invention into innovation. The practical component comes from the need for innovation to improve something. Whereas one can invent a grill for car exhausts, this would hardly count as an innovation. If something new improves on something existing for which there is demand, it then needs to be able to scale in order to satisfy that demand. Inventions that can improve the world, but which lack the means to be implemented and used, fail to qualify as innovations.

To encourage meaningful innovation, it is important to look beyond technologies themselves, and ask how they can be embedded into the environment. Breakthroughs in fields such as artificial intelligence and biotechnology constantly draw attention, but very few people’s lives have felt transformed by their existence. The capacity for these inventions to bear the fruits of innovation has not yet fully materialized.

One of the reasons why is a knowledge gap. New technologies take time to make an impact because there are not enough skilled people to actually implement them in the real world. Managers who would love to use these technologies, similarly, have little idea what to look for when hiring someone, or how to effectively place the technology in their workflow. It took decades for electricity to make a real impact on the American economy, but by the time it did, it transformed the nation far more than anyone anticipated. Whereas when the technology appeared, people believed its main benefit would be to save money, after a while, many managers realized the ability for electricity to transform their workplace environments and organizational processes to fully leverage its potential. This took time and education to understand.

Another barrier is technological complements. Most technologies do not exist in isolation, but rather interact with each other to deliver results. Take for example current machine learning methods that dominate artificial intelligence research. The most effective machine learning tools that businesses can use require large datasets and high amounts of computational power. Innovations in hardware, such as the introduction of the Graphics Processing Unit (GPU), designed for a completely different application, were essential in bringing down the cost of research. While the cost of computer hardware continues to decline, and quality increases, it nonetheless is still too cost prohibitive for most businesses to use effectively. Another innovation, cloud computing, which has enabled businesses to store their data at lower cost, also allows them to rent the necessary computing power, as well as access the necessary software via the internet, to leverage AI techniques at a much lower cost. As 5G, the latest, and fastest, iteration of wireless broadband gets deployed, the quality and adoption of cloud computing will only increase, making AI even more accessible, and allowing its benefits to spread.

Finally, policy is essential for transforming invention into innovation, both from private and public actors. Standards-setting bodies help ensure that the same technology can be used in different places and still integrate with the wider ecosystem essential for it to deliver benefits. The process of developing standards, however, requires a high degree of coordination and a sufficiently mature research environment, to be both useful and not lock companies into outdated technologies. Government needs to provide companies the room to experiment with different standards, allowing them to make mistakes at smaller scales, so that better technology standards can be enacted as innovations spread.

Taken together, it is clear that technologies cannot blossom or make an impact in isolation. The method of invention requires that an ecosystem is forward-looking, facilitates relevant transfer of knowledge, and allows for advances in various fields. Policies targeting specific technologies fail to deliver results for the innovation ecosystem by constraining trial and error. Only when understood to be a broad phenomena involving many actors in loose coordination can invention become the desired innovation.