Council Post: Big Challenges That AI Can Help Overcome And Steps Companies Can Take To Embrace It
Debanjan Saha is CEO ofDataRobotand a visionary technologist with leadership experience at top tech companies such as Google, AWS and IBM.
We may be on our way to a greater economic downturn as we look to 2023. Automation software and technologies that increase companies’ insights and efficiencies tend to be contra-cyclical (doing better in down markets) as companies look to do more with less.
AI-driven companies can exploit differences in time to action relative to competitors—even ones that have already created a strong business intelligence (BI) framework, according to a McKinsey Global Institute analysis. According to another McKinsey report, AI’s impact on the bottom line is growing, with the share of respondents reporting at least 5% of earnings before interest and taxes (EBIT) that’s attributable to AI increasing from 22% to 27% year over year.
Companies are already struggling to anticipate the movement of several economic drivers, which are hitting levels not seen in a decade or more. It will be crucial for companies to move beyond what has happened to predict what comes next. To this extent, companies are shifting their focus to more accurate predictions as the economy heads into new territory. That need, combined with increased access to data-driven insights, may increase the overall need for AI.
As I expect to see significant developments in artificial intelligence in 2023, here are five potential applications for AI and steps that companies can take to transition toward this technology.
Supply chain disruptions and shortages, which began with the Covid pandemic, continue to be a factor and are set to get worse as the economy moves into new territory.
For example, parts shortages, especially ones that crop up at the last minute, have crippled the production of some major companies. We can all remember stories from last year where numerous automakers had thousands of cars sitting outside their factories, ready to sell but for a missing chip. As supply chain issues intensify, manufacturers will be under pressure to avoid production mismatches, predict return rates to prevent overproduction and source alternatives.
Demand forecasting is the other side of the supply chain disruption coin. If consumer demand for a product outstrips supply, it leads to increased prices or shortages of that item and possible shortages for every component that goes into making it.
Executives used to view demand outstripping supply as a good problem to have. In the current environment, it’s still a serious problem. Creditors will not care that a borrower sold out of their products if that company is facing bankruptcy. As the complexity of products, supply chains and markets intensifies, so does the complexity of estimating demand, especially when you add confounding factors like seasonality and competitors’ promotions.
Not many industries like volatility the way the financial services industry does. The immediate impact will be to solve second- and third-order problems, reducing transaction costs through optimization of venue selection, price discovery, streamlining research distribution to clients and more—challenges that represent low-hanging fruit for cost savings and efficiency improvement.
It costs more to acquire new customers than to keep the ones you have, but it takes more than understanding to keep your customers.
Subtle signals hidden in things like purchase history, feedback, email frequency and other data show the state of a customer relationship. These patterns are often too nuanced to be detected by the human eye or detected in time to make a difference.
One crucial supply chain input is staffing, which has been a stress point for managers amidst labor shortages and the “Great Resignation” over the last year. Even with a significant economic downturn, managing seasonal staffing in human-centric industries like retail can be complex and challenging.
Data fuels artificial intelligence. For companies to embrace AI, they need an existing data fabric—an architecture for tracking, storing and analyzing high-quality data to make it useful to as many people as possible.
Fortunately, the data revolution is nearly two decades old now, and most companies already leverage some type of business intelligence within their organizations to derive insights from their data.
Beyond accessing data itself, companies need to take steps to move toward a culture of data as a central and mission-critical business asset.
Skilled data professionals are hard to come by, so companies should take steps to get the most value from the experts they already have.
As artificial intelligence becomes more intuitive and self-service oriented, intermediate-level data scientists—or “citizen data scientists”—will have the ability to solve more complex and higher-impact business problems using AI than in years past. These won’t replace the experts, but they will allow the most experienced data scientists in your organization to drive much larger projects, and accelerate time to value, that would have required more experienced team members.
We’re entering an uncertain period for business as the U.S. confronts what may be the end of the most prolonged period of uninterrupted economic growth in its history. But with uncertainty comes opportunity, even in a potential downturn.
The most significant impact of this widespread adoption of AI for making data-driven predictions—one that spans all the specific applications above—will be how it changes how intuition is used in business. Intuition has always played a critical role in helping executives bridge the gap between what is known and what experience suggests. Soon, that gap between knowledge and expectation will shrink.
In this new world, intuition will still be necessary, but companies will use it to discover new questions to ask, not to guess the answers themselves.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?