The Five Most Important Questions About Artificial Intelligence
A brief guide for investors
The artificial intelligence news these days is confounding at best. Magic new algorithms, we are told, will revolutionize education, cure cancer and arrest climate change. They will, at the same time, toss most of us out of work, turn democracies into dictatorships and – at least slightly -- raise the odds of global extinction.
The corporate players driving much of this revolution gathered with President Biden last week and committed to proceeding more carefully with their head-snapping advances, but there are still far more questions than answers. In fact, it’s the companies that were not represented at the White House that will do most to shape AI’s impact as they tap its powers for their own growth.
And while it’s too soon to weigh in on the prospects for the human race, now is precisely the time for to explore how vast multinationals, small enterprise and fresh start-ups plan to incorporate this transformative wave of innovation. Answers that suggest “not much will change” should flag a stock to be sold, a chief executive to be replaced or a venture on the verge of unravelling.
1. How will AI boost sales?
Any business with lots of customers is already scooping up data to better understand their needs and more precisely target future sales. But most are already behind the curve given the promise of AI to tap into vast new wells of unstructured data and raise the art of market segmentation to new heights. New analytics will suggest you can sell more blue bracelets to people who bought tickets to the last Taylor Swift concert. It will tell you that you will boost profits with 10% discounts in Ireland and 5% in Germany. And it will let you know when these answers change.
2. How will it help control costs?
Any firm with large manufacturing processes and extensive supply chains will be thinking hard about how AI can keep control of operations that are already too complex for the brightest minds to run. AI tools can design new factories and streamline existing ones. It can flag weather trouble that may delay key part and automatically identify alternatives. And long before we have to worry robots running the world, AI tools can improve the robots we already depend on to keep costs in line.
3. Are you really so special that machines can’t replace you?
For many of us who are highly educated, possibly creative and strive to produce a tailored work product, the hope is that not even the most amazing algorithms will replace what we do. But what we often forget in these conversations is that most jobs include dozens of distinct tasks, many of which would clearly benefit from tapping into a much larger or more sophisticated database. A medical diagnosis from a high-flying physician has to improve with tools that can scour millions of images of tumors. An astute consultant will deliver better advice when armed with instant analysis of a new market. Even a humble “global market strategist” might benefit from a more granular outlook for Brazilian monetary policy. It’s not that these tools are unavailable now, but they will soon be at hand much faster and much cheaper. (Disclosure: ChatGPT checked this essay for grammar, spelling and weaknesses in the argument.)
4. Do you understand what AI can’t do?
For all the hype, there are limits and risks. Ultimately, artificial intelligence still depends on inputs from the digital world, whether it’s proprietary data on corporate activity, unrelated data sets it may tap or broad searches on the web. It cannot know what it cannot know. It doesn’t know about the fire down the street that hasn’t been reported yet. It doesn’t understand the racial bias embedded in the web. It doesn’t know how the political risks in Thailand may undermine growth prospects. Business managers will still have to account these unknowns somehow.
5. Do you really know what business you are in?
Key to answering these questions is understanding the core value a business really delivers. Famously, Kodak was trapped into thinking it was in the business of selling film and forgot that its customers were really paying for images. Is your business to supply automobiles or deliver transportation solutions? Are you investing to beat the market or to help someone with a pot of money understand what to do with it? Especially when these new tools are likely to transform different parts of a business so rapidly, keeping an eye what customers are paying you for may be far more than re-jiggering what you have always done.
If the future of humanity remains as uncertain as ever, the ultimate impact of artificial intelligence will depend on how the world’s firms choose to incorporate its tools now. The only sure thing is that those who are ignoring these questions won’t be around long enough to catch up.