The combination of the SaaS business model and AI services could help bring AI to the masses without a heavy price tag. Could 2020 be the year AI as a Service takes off? I hope so—and to some extent it already has—but my excitement isn’t only because of how AI can lead to greater efficiencies and better customer insights for businesses in nearly every industry. In my opinion, the future of digital transformation itself depends on democratization of the latest technologies. In 2020, AI won’t just be nice to have, it will be a necessity.
Right now, advances in AI are racing ahead, and technologies like 5G and blockchain are making way for even greater developments—the IoT, real-time analytics, etc. Sure, some areas of AI are lagging (hello, speech recognition). But by and large, AI is breaking new ground daily. For example, just look at all the announcements at CES this week—everything is being connected, everything is creating data and all of that data will be used to power the future of customer experience or business optimization using the power of AI.
However, it’s important to note that last part about customer experience and business optimization is so important because those groundbreaking developments means nothing unless companies can actually use the technology in the real world. That’s what AI as a service is all about: making new technology accessible across the board, whether a company is big, small, tech-advanced, or on a budget.
Microsoft, Google, Amazon, and IBM have stepped up as major players in the AI as a Service race. Because they each have large public cloud infrastructure and AI platforms, they’re able to make AI as a Service a reality for those looking to use AI for everything from customer service to robotic process automation, marketing, analytics, predictive maintenance—you name it. Technologists have been talking about AI as a Service in varying degrees for a few years now. The following are a few reasons it will finally pop in 2020.
- Demand is growing. The secret is out. Businesses today can see the value AI is creating for their competitors, and the disruptions it’s causing around the world. They know they need AI to meet growing customer expectations. It’s estimated the global AI as a service market will hit U.S. $6-7 billion by 2023. I honestly think that estimate is low. Once AI is democratized, we’ll be seeing it pop up absolutely everywhere. And that’s exactly where we are heading, especially in areas like AI-managed customer service.
- The ecosystem is growing. For AI to be optimized, it requires an entire ecosystem of partners. A good example of this is how a player like AWS partners with NVIDIA to offer its GPUs in the cloud—AI as a Service. Other examples could be an industrial tech company partnering with an analytics firm to deliver AI/Analytics software platform as a service. Siemens and SAS have done this within Siemens Mindsphere offering by baking SAS analytics into its SaaS based Industrial IoT software platform. Businesses need their vendors’ apps to be compatible—which means they need infrastructure to be compatible—which means they need technology that speaks common language and works together for common goals. As more and more providers enter the AI space, more businesses from every industry will be able to find solutions that fit their specific use-cases, making adoption more global, which will make AI as a service even more popular.
- Companies are catching up. The legacy era is finally, generally, ending. Most companies understand that it’s time to move forward, and AI offers the types of efficiencies they’re looking for, be it through processing, data and storage, or cloud-based apps. Pretty soon, “AI as a service” will simply be what we know as “service” on every front.
- SMBs require it. AI as a service allows for advanced infrastructure at minimal cost. This is exactly what small and mid-sized businesses around the world need right now to compete against tech giants. AI as a service, in a way, is their only hope. However, cost can be prohibitive. Hiring the data scientists and standing up the complex infrastructure to build models and garner returns from AI will take significant resources that will be difficult for many large companies to compete, let alone small ones. Delivering AI in SaaS or PaaS layers whether an Autonomous Database that can self heal, self secure and drive like Oracle is touting, or a no-code/low-code dev environment like Microsoft’s Power Platform where companies can built custom chat bots for their website or automate a process without much (if any) real-world programming experience. These types of tools will be key to enabling smaller companies to consume the benefits of AI in the near term.
Basically, AI allows companies to use the power of AI without having to procure the expertise to manage it. In a time when there’s a shortage of AI experts and ever-increasing competition in the marketplace, that’s a huge workaround.
Yes, as AI as a service become more popular, it will also become more popular for hackers and other “bad guys,” creating complex security issues we all need to be more aware of. Still, there has never been a smarter way for today’s businesses to begin harnessing the power of things like machine learning, deep learning, and natural language processing—even if they’re just getting their feet wet to see which new technologies are worth investing in long-term. To me, AI as a service is one of the biggest game-changers we’ll see in digital transformation. Yes, digital transformation is hard. It isn’t always successful. But AI as a service could definitely improve a company’s odds of experiencing success.