You may have seen a cartoon with the theme “When everyone digs for gold, sell shovels” very frequently on your LinkedIn feed. Though I used to see that cartoon every day in my LinkedIn feed, I could not find it when I wanted to grab it for this post this morning. So, I recreated it below.

I was motivated to write this article by a Bloomberg news article I read this morning. The article titled Nvidia Becomes Tesla’s Successor as Market Flips From EV to AI postulates that investors should not see this intense focus or interest in AI as limitless. The article compares the current hype with the EV hype in 2017. I agree that we should be cautious about the AI boom being “limitless.”

I was never an EV enthusiast because even in 2017, you could see how the segment, with its limitations and technology, will fare a few years from then. I shared this sentiment in many of my posts and articles, and eventually, as years progressed, it became a common sentiment. With AI, it is different.

In one of my “Think About It” episodes a couple of weeks ago, I warned that if we keep focusing on just building more and more powerful algorithms, and do not match the fervor with an explosion in applications in business and industries, we may be looking at the third AI winter. While the good news is that even if a third AI winter comes our way, it will not be soon, the bad news is that there is actually a probability of a third AI winter.

Last two AI winters were not just about hype. They were also a result of hype not matching with hardware capabilities. The meteoritic rise of NVIDIA is due to its capability to provide the “shovels”, the chips, that are powering the AI race. The technology to support the vision exists at this time. We must ensure that the current AI race between large tech companies eventually percolates as impact-creating capabilities in businesses and industries.

Capability and solution are two separate aspects. What leading technology companies have built are capabilities. Their mega AI capabilities are robust. The next step is to turn these powerful capabilities into solutions across industries and functions. As a natural first step, these companies are working on integrating these capabilities into their existing solutions to turbocharge these existing solutions. However, to match the current hype with applied solutions, there needs to be a need to massively percolate these solutions.

Hype lives and breathes within the end users of technologies. These are not the people fueling the demand for NVIDIA’s chips. The demand is being driven by those who are looking to mine on the hype. And their strategy is prudent. The current race to build capability is with the hope that the capability will be leveraged to address the hype. The time to start addressing the hype though, is now !

While improving the capability of AI technologies is imperative, another key imperative is to push these capabilities to the end users in the form of solutions. And that means solutions that end users actually find useful. As with every hype, no matter how astute marketing is, if the end user eventually does not see sustained value, the hype dies.

While on the one hand, large technologies companies are chasing AGI, 80% of companies are failing to implement even ML solutions. And this disconnect is what may lead us to an AI winter. Not at the scale of previous two AI winters but there is a probability of a third one.

Luckily, this time, we have the technology available to make dreams a reality. The only path that can take us to the third AI winter is the one where these massive capabilities that big tech is building, does not translate into thousands of applications in business and industry soon.