Penn Medicine's Brian Wells: A strategy for AI adoption

AI
As the major AI manufacturers woo consumers, developers in a corporate setting face challenges in deploying the technology.

Artificial intelligence (AI) technology is hitting its stride for consumers, but developers have some work to do before it becomes viable for corporate use, according to Brian Wells, associate vice president of health technology and academic computing at Penn Medicine.

The bleeding edge of healthcare technology has already begun to engage with elements of AI via assists from cognitive computing platforms. The potential for medical robots that could free up time, energy and attention for practitioners has attracted attention to the technology, not to mention funding.

RELATED: At HIMSS 17, IBM's Ginni Rometty talks trust, values in a cognitive computing era

The technology’s maturation in the consumer space brings AI closer to general adoption, but it also raises thorny problems for independent developers and corporate users, Wells writes in an op-ed piece for HealthcareITNews. 

As the current big players in the consumer AI market fight for share, Wells cautions that developers must understand the benefits and limitations of all of them before attempting to use them in a custom solution. Using voice interaction, particularly, means “you must either choose a horse to ride or port your solution to three or four different platforms,” he writes. A standardized application programming interface (API) providing a unified back-end development environment would go a long way toward mitigating this issue.

RELATED: Healthcare industry looks to future of APIs

Making matters more challenging, consumer platforms tend to make use of existing user accounts in existing technology ecosystems (think Amazon, Google, Apple, Microsoft), which doesn’t necessarily translate to a corporate environment. Account security in a corporate environment, in which a system might need to discern among different users, logging them on and off the system smoothly all day long, presents a more complicated use case than that of a domestic consumer account.

Addressing the challenges around easier development and more efficient corporate use would truly unlock the potential of the technology and drive greater adoption, Wells writes.