AI impact is still not showing up

"AI is everywhere except in the incoming macroeconomic data."

This killer sentence by Apollo chief economist Torsten Slok seems to show that despite the current hype, AI is failing to bring impact. So it seems that Solow's productivity paradox is applying to AI.

This is supported by this Fortune's article that states: 

"A study published this month by the National Bureau of Economic Research found that among 6,000 CEOs, chief financial officers, and other executives from firms who responded to various business outlook surveys in the U.S., U.K., Germany, and Australia, the vast majority see little impact from AI on their operations" (...) Nearly 90% of firms said AI has had no impact on employment or productivity over the last three years, the research noted."

 

We are still living the Hype

It is very interesting to note that while AI usage increased, confidence in the technology’s utility plummeted 18%, indicating persistent distrust.

From a personal perspective, this tells me that if we look at the Gartner Hype Cycle [fig. 1], we are not still in the plateau of productivity of AI. We are still in the "Peak of Inflated Expectations" and we still have the "Trough of disillusionment" and the "Slope of Enlightenment" before finally reaching the Plateau of productivity.
 

Figure 1 - Gartner Hype Cycle (Source: Wikipedia)

 

Like in the early days of internet there was all sorts of inflated expectations with .COM companies market value going through the roof while their "internet-related" topline was not showing up, you see now it happening with AI companies.

Some rumours say that some AI-surfing companies are trying to inflate demand on their products by aquisitions or paying its customers to buy its own AI technology. This shows that something is not really adding up to this hype cycle: AI-related results are being purposedly inflated to keep the money flowing in.

But when executives start to get tired of unfulfilled promises after looking year after year to their topline, reality may hit hard, and may have an AI bubble burst.

And then, from the ashes of the AI Bubble, will we enter a Slope of Enlightenment where we can rationally understand what was a hype and what really had the potential to provide real value. Only after that AI will be able to provide real, tangible value and enter the "Plateau of Productivity"

 

Taking irrational decisions is worse than not jumping into the AI bandwagon

Hype cycles can be identified by assessing if even people that are not familiar with a topic speak profusely about it. If we look around, all sort of people speaking about AI. AI is the topic these days. If you don't mention AI in your business pitch, you are out of the zeitgeist. But if you are experienced enough and have some knowledge in the field, you know AI has already a several decade-long development history. Although there were many scientific breakthroughs in the recent couple of decades, AI is not the solution for every business problem.

The problem with hype cycles is that they may cause susceptible decision makers to take irrational decisions. You have seen many companies firing employees only to find they needed them back or realizing that despite the investments made, 95% of AI pilots fail. 

Going all-in in AI and gen AI does not make sense (yet) for the majority of organizations. It implies capital expenditures, adoption overheads and unclear productivity gains.

There are, however examples where AI adoption is shining and delivering value. Software development is one of them (although technical debt risk and maintenance penalties may apply if the process is not carefully managed), fraud detection in financial services, document processing, health diagnostics, route optimization, etc. So beneath all the hype, there is value on AI.

 

The wise approach to AI

At the current time, unless you are playing in the AI frontier research field, keep investments grounded on solid data and wisdom.

What we always advise our customers to do is: "do not invest in AI without a solid business case and clear ROI perspectives". And to build a solid business case you need to get information about performance, risk, feasibility, costs and business alignment in the first place. And to get that, you should be first analysing the most promising use cases, building prototypes, testing technology maturity, testing market response and internal impact - gathering information that will allow you to develop a solid business case, instead of chasing hypes. This way you can take rational decisions grounded on real data, allowing you to take a high quality decision.

At Mourinho Solutions, we can support you through that process. We work with you so that real AI value is separated from the hype, crafting the best approach for your business and strategy and making sure that each cent of your investment yields the highest return for your business and operations. We help you to take high quality decisions, so that your business can thrive.