Industry observers say we’re about to enter the next AI Winter. (Yes, it’s a thing.)
Artificial Intelligence emerged more than 60 years ago, and in that timeframe we have seen many such seasons. While I don’t believe the coming winter will be as severe as the others, by all accounts, it’s coming.
The first AI Winter happened in the 1970s, when, after more than a decade of heavy funding of academic research by the U.S. Department of Defense, the government pulled back. It became clear that advancing AI would be much more challenging and expensive than originally foreseen. The government curtailed funding and triggered a massive slowdown.
Years later, it was AI Spring again. Funding from the private investment community and governments went into high gear in the late 1970s through most of the 80s, including by the U.S., the U.K., Japan and others. Several years later these investors became impatient again, sending AI into its second cold season. This one was driven by failures within specific technologies and application sets, and cancelation of new spending on AI by the Strategic Computing Initiative in 1988.
A key turning point came around the year 2000, as machine learning emerged, and investor interest in AI was rebooted. The blending of AI and ML with advanced supercomputers and the mountains of raw data we’d collected over time were the kindling that enabled the next wave of growth to blaze.
You see, when we first made the switch from analog to digital, we collected tons of data, but had no way to sort through it all and use it. That is what AI, ML and IoT have done for us in the last 20 years. IoT enabled us to gather data pretty much everywhere all the time. ML allowed us to pull information and insights from the data. And AI gave us the tools to activate and apply our learnings.
A high-profile turning point for many of us was the 2011 introduction of IBM’s ( Chart International Business Machines Corp. - $147.41 0.11 (0.00075%) ) Watson on Jeopardy. (Many viewers thought it was an incredible new technology but didn’t realize it had been with us for decades already.)
Going forward, I believe each new AI Winter will be shorter and less severe, until they are no longer an issue. Remember: Data never disappears. As time passes, we find new ways to work with it — separate signal from noise — find the relevant needles in the haystack.
One harbinger of the coming winter was IBM’s sale of Watson Health early this year to Francisco Partners. Only a few short years before that, countless partner companies joined IBM in Las Vegas at the World of Watson trade show. The excitement seems to have died down considerably.
If and when we do enter the next AI Winter, I believe companies in the Artificial Intelligence and Machine Learning space will continue to advance their work. It will just be harder for them to garner investment capital and media attention. They will have to learn how to punch their way onto the map and fight for a place on the radar of the investment, media and general marketplace. Meanwhile, many investors and governments will be keeping a long-term eye on the future.
Jeff Kagan, a telecom, technology and wireless industry analyst and consultant, is an Equities.com columnist. He covers 5G, AI, IoT, the metaverse, autonomous driving, healthcare, telehealth, pay TV and more. Follow him at JeffKagan.com, and on Twitter @jeffkagan and LinkedIn.