Taking AI Adoption in High Tech to the Mainstream
The high-tech industry has been a leading adopter of artificial intelligence (AI), so much so that it influences the digital transformation trends in most other industries. Technology giants are not just building AI solutions, they are acquiring smaller AI companies to build more capabilities and finding new use cases outside the IT department, in our offices, hospitals, and homes. Yet, as per leading research reports these firms are still lagging in the overall AI adoption when compared with other digitization efforts.
This paper attempts to find the factors behind the high-tech industry’s failure to scale and the best practices the industry can adopt to encourage mainstream adoption.
Untapped potential of AI in high-tech
It has been almost seven decades since Alan Turing first envisioned a ‘thinking’ machine that could potentially carry on a conversation with a human that was indistinguishable from a human-to-human interaction. Since then, AI has seen several periods of crests and troughs. The last two decades, since IBM’s Deep Blue beat reigning world chess champion, Garry Kasparov, can be considered the beginning of AI’s golden age.
The Mckinsey Global Institute’s report titled ‘Artificial Intelligence the next Digital Frontier’ points to the lead role of high-tech industries when it comes to AI adoption. The leading tech giants and digital native companies including Google, Apple, etc., have collectively invested several billions of dollars in AI over the years. This includes investments in powerful supercomputers, dedicated teams to build highly sophisticated algorithms, and data technologies to support machine learning.
Not surprisingly, digital native organizations have emerged as the primary providers as well as the biggest consumers of AI technologies. For digital-native organizations, technology is not just an enabler, but the primary driver, irrespective of the sector that they belong to. For example, Tesla is more of a technology company than an auto company. It is the same for companies such as Airbnb, Uber, Netflix, etc. Most of these companies have been consistently adopting AI and automation in their enterprise IT for the last few years, although they are at differing stages of the evolution journey. Yet, when we look at the actual impact of AI today versus its tremendous potential, the progress has been disappointing.
AI’s failure to scale in high tech
Despite this sustained push on AI, we find that its on-ground adoption is not very high, and certainly not proportionate to the hype and investment, in terms of the scale or volume that one might expect. There are several reasons for this discrepancy. One is that a lot of the investment in AI in the high tech industry has been directed to improving internal performance, rather than on the customer or business aspect. The inability to demonstrate business impact has also led to lack of interest from stakeholders in some cases. Most companies are stuck in the first or second phase of evolution and have been unable to move forward. Also, AI cannot flourish in isolation unless it is supported by humongous computing muscle, powerful analytics and plentiful high-quality data. The relatively low uptake of these has impacted AI adoption.
The lack of joint business and IT sponsorship for AI projects has been an important impediment to AI adoption. Complex business cases require involvement from multiple stakeholders across multiple function / organizational boundaries.
Sometimes, the deterrent isn’t really to do with AI technology per se. For instance, people might resist AI adoption since they are not confident in their own ability to make an impact. They might even worry about job loss as a result of AI adoption. In some instances, the lack of cultural alignment can impede adoption. The organization may be ill-equipped to embrace change. Or it may just lack the presence of a visionary leader who is championing innovation.
Best practices for a successful AI initiative
In general, here are a few considerations that can help run a successful AI initiative for high-tech industries:
The right questions
Asking the right questions around the automation strategy, technology solution, governance of AI is very important, such as:
- How do you plan to operationalize your Robotic Process Automation (RPA) Centre of Excellence?
- How do you use AI to enable the humans in your enterprise?
- How will you integrate RPA into your IT ecosystem?
Clarity on when to implement AI has tremendous influence on the outcome of automation. Change management to ensure that changes to Standard Operating Procedures are done in time and automation is effectively adopted on the ground by the relevant team or department is an important success factor.
Clearly defining the scope and having a line of sight into costs is crucial. Similarly, having a list of processes to automate, estimates of cost of implementation, and confirmation on metrics such as productivity, error rate reduction, customer experience improvement, regulatory benefits is key to the business case.
A thorough understanding of impacted applications, provisioning of robot IDs, policy exceptions for robot access, access to infrastructure in the development, test and proof environment are also crucial.
Organizations should have procedures and controls in place to carefully evaluate the changing needs in business processes as a result of automation. Also, the harmonization of processes is needed to facilitate automation. More standardized a process is, the better and higher is the benefit leveraged from automation.
Sometimes, a process may not yield expected results even if it is completely automated, if process owners and their teams do not embrace it wholeheartedly. Client buy-in is an important part of the process too since certain aspects of automation will need to access ERP and IT systems simulating as end users. In addition, hosting of the solution in the client network can be a challenge unless there is client buy-in.
Today, chatbots like Siri and Alexa are part of our daily lives. IT automation is becoming an active area of adoption in enterprises to handle security concerns, production management and user technology problems. Natural language, face and speech recognition technologies are emerging as the backbone for many AI applications.
As AI gains more acceptance and comfort amongst users, demand for AI-backed offerings will increase and drive more adoption. High-tech companies can find a huge opportunity here to drive new revenue streams and achieve new dimensions in customer experiences. However, unless there is a conscious effort to address the challenges, mainstream adoption at scale will not be a reality for a long time to come.