Chintan is a serial entrepreneur and currently the CEO and founder of Brainvire.
Artificial intelligence (AI) and machine learning (ML), or AI/ML, are quickly becoming a crucial next step for business growth. Recent years have seen more and more businesses adopting this technology and witnessing significant benefits in several areas.
A McKinsey survey indicated that AI adoption rose from 50% in 2020 to 56% in 2021. As per another survey, “76% of organizations say they prioritize AI/ML over other IT initiatives, and 64% say the priority of AI/ML has increased relative to other IT initiatives.”
From ramping up dedicated budgets to increased hiring of data scientists, organizations have been making focused efforts to adopt AI/ML to stay ahead in the race. These initiatives are being viewed as a means to achieving top-line growth while maintaining bottom-line costs.
AI/ML Adoption: Opportunities And Barriers
Artificial Intelligence and Machine Learning are arguably the most influential technologies in the world today, with the potential to change how we live. For businesses, this level of disruption could translate to intelligent processes bridging the gap between humans and machines. By 2030, AI is projected to contribute as much as $15 trillion to the global economy, out of which “$6.6 trillion is likely to come from increased productivity, and $9.1 trillion is likely to come from consumption side effects.”
MORE FOR YOU
Having said that, a handful of early adopters and big players with vast resources are at the forefront of this wave and making the most of it. And even though AI/ML adoption is on the rise, many organizations continue to grapple with certain key challenges.
Governance: There are multiple facets to AI/ML governance and its implications on your brand and bottom line. Broadly, governance involves an overall process of how an organization deals with policy implementation, access control, risk and ROI evaluation. Despite being one of the top concerns among organizations, AI governance is yet to fully attain the strategic importance it warrants. If AI/ML is not prioritized, organizations will likely see adverse impacts on their AI/ML initiatives while also putting their businesses at risk.
Good governance can be the foundation for minimizing risk and improving ROI. One of the most effective ways to deal with governance concerns is to get all stakeholders on the same page in this matter. Effective collaboration and communication among teams are critical to good governance because that is the only way to avoid conflicts and ensure all concerns are heard and addressed. Governance has to be at the same level of importance as budgets and hiring.
Data-Centricity: One of the main factors that impact AI/ML implementation is the availability, quality, quantity and computing power of data. However, the problem with data is that its quality and quantity are rarely consistent due to the absence of standardized data practices across industries. Siloed, inconsistent or poor-quality data is a big challenge since many businesses are still leveling up and may or may not possess the technological sophistication to work with data at this level.
In their attempt to overcome these issues, businesses may see a delay in their AI journey. However, in order to stay on track right from the initial stages of implementation, a well-shaped strategy for sourcing and managing data can go a long way in successful implementation. The focus has to be on obtaining “good data” and protecting downstream algorithms from the impact of poor quality or biased data. Furthermore, it is important to take a forward-thinking approach to AI/ML, especially from an integration and scaling point of view.
Cost And Time: There are substantial costs and time involved in AI/ML deployment. Lack of resources and unfamiliarity with the tools and domain know-how can drive up costs. Moreover, investing in expensive smart technologies, updating existing systems and computational costs are additional expenses that may not be optional if AI implementation is underway.
Furthermore, as indicated in a survey, for 38% of the organizations, over 50% of their data scientists were engaged in deployment, and scaling can only make matters more time-consuming. Developing and upgrading software typically brings the risk of data loss and restoring it takes time. This is also applicable to AI/ML implementation.
Using well-designed systems and getting the necessary expertise on board can go a long way in mitigating the costs and time needed for integration. It is essential to understand that this level of deployment cannot work with a plug-and-play approach and presents numerous issues, including compatibility, software and hardware challenges. Constant monitoring, consistent upgradation and cross-functional team collaboration can go a long way in ensuring successful implementation.
As AI/ML continues to grow in importance, businesses should make informed decisions and appropriate investments no matter the size of the organization. A straightforward problem-solution approach may not be the best way of adapting to the changes. Comprehensive, long-term transformation strategies are the need of the hour to facilitate AI/ML adoption.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?