In our last issue, I showed you why I believe AI is getting much closer to something that resembles human-level intelligence.
Which makes what’s happening inside corporate America all the more surprising.
By now, you’d think companies would be replacing workers with AI agents just as fast as they could build them.
After all, today’s AI can solve elite math problems, conduct scientific research and write sophisticated software.
Instead, many businesses are discovering that intelligence alone isn’t enough.
The Intelligence Paradox
Getting a technology to work is only half the battle. Figuring out how to use it profitably can be a completely different challenge.
The first automobiles worked long before there were highways, gas stations or traffic lights.
The internet existed years before businesses figured out how to build profitable online companies.
And I believe artificial intelligence is at a similar stage.
AI models are becoming more capable with every new release. But figuring out how to turn that intelligence into a positive return on investment is proving to be much harder.
Uber (NYSE: UBER) learned that lesson earlier this year when the company’s engineers embraced AI coding assistants so enthusiastically that Uber exhausted its annual AI budget in just four months.
Part of the reason was that Uber encouraged employees to use the tools as much as possible.
The more they used AI, the more the company paid. And as usage exploded, so did Uber’s bill.
Uber’s problem wasn’t adoption. If anything, employees adopted AI faster than expected.
The challenge was proving that all those AI-generated solutions were creating enough value to justify their cost.
That’s an issue many businesses are struggling with today.
PwC recently surveyed more than 4,400 CEOs across 95 countries about their AI investments, and 56% of them said AI had produced no measurable increase in revenue and no meaningful reduction in costs.
Another 22% said AI had actually increased their costs.
In fact, only about one in eight CEOs reported seeing both higher revenue and lower costs from their AI investments.
Those numbers might seem surprising given the growing evidence that AI can make individual workers more productive. In fact, I recently showed you new research demonstrating that AI helped customer-service agents resolve more problems per hour and allowed consultants to complete their work faster.
So there’s documented evidence that AI can boost employee productivity. But once businesses enable AI to make decisions instead of simply assisting employees, the margin for error becomes much smaller.
Starbucks (Nasdaq: SBUX) recently ran into that problem.
Last year, the company began testing an AI system designed to count inventory in its North American stores. Using cameras and artificial intelligence, the system was supposed to recognize products automatically, saving employees time.
Instead, the AI often confused similar-looking items and still needed workers to step in and correct its mistakes.
After less than a year, Starbucks stopped rolling out the system and returned many stores to counting inventory the old-fashioned way.
Pizza Hut (NYSE: YUM) is facing an even more expensive AI-related issue.
A franchisee operating 111 restaurants sued the company after being required to adopt an AI-powered delivery management system called Dragontail.
According to the lawsuit, the software’s delivery predictions encouraged DoorDash drivers to delay pickups while waiting to combine multiple orders.
The franchisee claims the result was colder food, unhappy customers and more than $100 million in damages.
Pizza Hut disputes those allegations. But regardless of who ultimately prevails in court, both of these stories point to the same challenge.
AI can be incredibly smart. But if it doesn’t work reliably in the real world, businesses won’t trust it.
Businesses need AI that produces the right answer thousands of times in a row. It also has to understand company policies, follow regulations, work with existing software and know when to ask for help.
In other words, businesses are running into a different kind of AI bottleneck.
Dependability.
Artificial intelligence is already smart enough to perform many individual tasks. The next step is to turn those capabilities into dependable systems that businesses can trust.
And just as important, AI has to cost less than the employee it’s replacing.
That economic reality is becoming impossible for investors to ignore.
Over the past few weeks, many of the biggest AI stocks have sold off even as the technology continues to improve.
That’s because Wall Street is becoming less interested in benchmark scores and more interested in business results.
Companies are spending hundreds of billions of dollars on chips, data centers and AI infrastructure. Now investors want proof that all that spending will eventually translate into profitable products and services.
Benchmark scores can tell us which AI is smarter.
But they don’t tell us whether AI is ready for the messy reality of running a business.
Here’s My Take
Yesterday, I argued that we’re getting closer to AGI with every advance in reasoning, research and autonomy.
But intelligence isn’t the only quality that matters in the workplace.
The average employee isn’t valuable simply because they’re smart. They’re valuable because they show up every day, understand the business and can be trusted to get the job done.
That’s the standard AI is ultimately competing against.
Businesses are still trying to figure out where AI can create genuine economic value.
But here’s the thing…
That’s not all AI’s fault.
Tomorrow, I’ll show you a chart that reveals just how much of today’s AI potential businesses are leaving on the table.
Regards,
Ian KingChief Strategist, Banyan Hill Publishing
Editor’s Note: We’d love to hear from you!
If you want to share your thoughts or suggestions about the Daily Disruptor, or if there are any specific topics you’d like us to cover, just send an email to [email protected].
Don’t worry, we won’t reveal your full name in the event we publish a response. So feel free to comment away!
