Cloud accelerates AI adoption

June 12, 2019

A blog post Jeff Loucks, executive director, Deloitte’s Center for Technology, Media, & Telecommunications, Deloitte Services LP

Advances in machine learning and deep-learning neural networks have opened up myriad ways for companies to improve their operations, develop new offerings, and provide better customer experiences. However, until recently, most organizations lacked the expertise and resources to take full advantage of AI. Only companies with access to large data sets, scarce expertise, massive data centers, and specialized processing power were able to uncover the best AI use cases, create customized solutions, and scale them.

As a result, it has been mostly the technology titans of the world, capable of investing billions into technology and talent, that have reaped the benefits of AI to date. That’s poised to change, however. The pioneers that have perfected their own AI solutions at scale have launched services for mass adoption. Meanwhile, large enterprise software companies are integrating AI capabilities into their cloud-based enterprise software and offering it to customers. In addition, a host of startups are sprinting to market with cloud-based development tools and applications.

These innovators are making it easier for more companies to access AI capabilities through the cloud without making big upfront investments. Deloitte Global predicts that, in 2019, companies will accelerate their use of cloud-based AI software and services, resulting in more full-scale implementations, increased investment and returns, and democratization of AI benefits.

A faster track to AI adoption

Deloitte surveyed 1,900 executives whose companies could be considered early adopters of AI for its recent global state of AI in the enterprise report. These respondents revealed that their top obstacles to AI adoption are data issues (accessing quality data, cleaning data, and training AI systems), integrating AI into existing processes and workflows, and AI implementation.

In addition, these early adopters said they struggle to recruit the necessary talent for their AI initiatives. Forty-one percent said that a moderate skills gap hindered their AI initiatives, while 27 percent called their skills gap major or extreme.

Cloud-based software and platforms can help companies sidestep some of these technology and talent issues, enabling them to benefit from AI even if they lack the expertise to build and train systems or manage data on their own. Deloitte Global predicts that, by 2020, 70 percent of AI adopters will do so through cloud-based enterprise software, and 65 percent will create AI applications using cloud-based development services.

The easy way: AI-Infused Enterprise Software

Enterprise software with integrated AI has proven to be the most popular path to acquiring AI capabilities. According to the survey, 58 percent of respondents are using this approach. This software is typically cloud-based through either public or private deployments. Deloitte Global estimates that, by 2020, about 87 percent of AI users will get some of their AI capabilities from enterprise software with integrated AI.

This adoption method has several advantages. Companies don’t have to develop their own AI applications: The cognitive capabilities simply run in the background of the software. End users don’t need specialized training to run these applications. The IT organization doesn’t have to develop new user interfaces, which can be a challenge when developing AI from scratch. In addition, the vendors offer continual upgrades and new functionality.

The marketplace of AI-infused enterprise apps is expanding, giving companies even more options, such as services aimed at specific business functions including HR and marketing and industry-specific cognitive-enabled apps. One of the biggest benefits of the enterprise software approach, however, is also one of its biggest drawbacks: Use cases are strictly defined. On the one hand, companies don’t need to develop their own use cases. On the other, the software may offer only limited customization, which could curb competitive advantage for customers.

The customized way: AI development services

That’s where cloud-based AI development services come in. These offerings, such as automated machine learning and data science modeling tools, do require companies to have in-house AI and data science talent. However, they can help companies get the most out of these professionals by providing access to proven models and accelerating key processes. That can enable companies with some technical AI expertise—but not enough to develop their own AI systems from scratch—to create a higher volume of AI services quickly and at scale.

There are multiple steps to creating an AI solution: building models, training them, evaluating their performance, and tuning them for optimal results. It’s a labor-intensive process that AI development services can help to automate. Automated machine learning, for example, can select the most robust model out of a given set and fine-tune it 100 times faster than a human data scientist can. Some AI development services are getting so intuitive that developers don’t even need much specialized knowledge, such as coding experience. As with enterprise software, customers can access these services in the cloud.

By the few, for the many

As these cloud-based AI capabilities have become more pervasive and early adopters have gained experience with them, they’re beginning to produce results. Respondents to the Deloitte survey spent an average of $3.9 million on AI in 2017, and these early adopters reported an average return of 16 percent on those investments. Even more telling is the fact that 63 percent of them said AI is very or extremely important to their company’s business success today, and 81 percent said it will be within two years.

It’s clear that AI adoption will continue to accelerate as more cloud-enabled services come to market—from prepackaged enterprise AI solutions to development tools. IT leaders who want to take full advantage can take several actions:

• Follow AI trends closely. The market is changing rapidly, and smart IT leaders should stay on top of emerging options. Just as the competition for market share is driving advances among technology companies, many AI early adopters are experimenting with these capabilities to leapfrog their rivals.
  
• Take advantage of off-the-shelf options. There’s no need to reinvent the AI wheel in all cases. IT leaders can first assess ready-made cognitive solutions to see if they meet their needs.
  
• Don’t outsource everything. AI-savvy CIOs should make sure to hire some AI experts. Enterprise software and cloud-based development platforms provide an effective gateway to AI, but companies should have their own expertise to ensure AI services meet business needs, to develop and customize algorithms using development platforms, and to keep business expectations in check.
  
• Focus on the business need. The answers AI-enabled software provides are only as good as the questions it receives. IT professionals who understand both data science and the business context can help ensure the IT organization is investing in the AI solutions most likely to deliver desired business outcomes.
  

This article first appeared on the WSJ.