No – not yet.
I’d say it’s more like 1996.
In the development of the next bubble, we are at the Yahoo or Netscape going public moment, which was in 1996.
But, yes, another bubble is in the early stages of brewing. And it’s around data and machine learning – with some social and mobile possibly thrown in.
One sure sign that a bubble might be forming comes from the changing fads in how companies like to brand themselves.
‘Data’ is becoming what ‘dotcom‘ was back in the 1990s – the label that every company is suddenly using no matter their core business.
For example, Intel announced at its recent investor day that they are *not* a semiconductor company. No. They are a *data* company.
“The way we look at the world is a bit different,” Intel CEO Brian Krzanich said. “We are a data company.”
Never mind that semiconductor sales generate $30 billion in annual revenue for Intel with a fantastic 35% operating margin.
Another bubbly sign comes from M&A activity.
Are large ‘uncool’ companies feeling left out of the data dance party? Are they paying a lot for a ticket?
Yes. Check that. Intel – the data company – yesterday announced the acquisition of autonomous vehicle technology company Mobileye for $15.5 billion. That’s an eye-popping valuation – something like 40 times sales. No doubt Mobileye raised that valuation by announcing recently that they too are not what they appear. They are not an autonomous vehicle tech company. Oh no. They are a *data* company.
When will this bubble fully develop?
No one knows.
History rhymes. It doesn’t exactly repeat.
Are we 3 years away? 5 years? 10 years? The only accurate answer can be “I don’t know.”
Bubbles typically only form around real technologies with real potential. We aren’t prone to Tulip bubbles in Silicon Valley (that was one of the first recorded bubbles a few hundred years back in Holland).
Machine learning is indeed an important new development with huge potential. Magic can and will happen.
To be clear, I also thought the Internet back in 1996 – and 1999 for that matter – had huge potential. My first keynote speech came in 1996 at the “Surfing the Web Potential” conference in Silicon Valley (gotta love the name of that conference). A year later I was on the startup team of one of the first companies that Amazon then purchased (at a healthy valuation, which I quite appreciated). And I have continued to work in the Internet world ever since.
But, in the late 1990s, I also produced a report that Business 2.0 featured suggesting we were in a bubble – and that there would be a correction. I, of course, had no idea how bad it would get or how much it would affect my company or Business 2.0 for that matter, which went out of business. I naively thought I’d benefit. Oh well.
Then several years after the dotcom crash, with the help of Whitney Tilson and others, I anticipated the housing bubble. Fortunately, I was able to be a bit smarter about that one.
While we can’t know when this bubble will fully materialize, we can easily tell once the bubble has formed. Despite what then Fed Chairman Alan Greenspan said, bubbles are easy to see when they are raging.
How do we know when the next bubble has arrived?
I have a simple test I’ve created over the years.
I call it the ‘taxi driver’ test (perhaps should be updated to ‘Uber driver’ test).
You’ll know when we are in the next bubble when taxi drivers talk to you about machine learning, robots, and AI. In 1999, taxi drivers talked to me about dotcom stocks. In 2007, taxi drivers talked to me about buying homes (often multiple homes). Of course, if self-driving cars replace taxi drivers, I’ll need a new test.
While there are signs today that a bubble is brewing, one has not yet fully formed. Some taxi drivers may be upset with Uber but they aren’t talking to me about machine learning or know the name, Mobileye. Yet.
I believe machine learning and related AI technology will indeed change most everything in the coming decades. But from testing current versions of Tesla’s Autopilot, Amazon’s Alexa, and other products I can assure that we are very early in this change. These are not ready for mainstream consumer use (yes, true even of Alexa, which for many buyers has turned out to be a very good but expensive cooking timer or music player).
So – what should companies do today?
Certainly, invest in and experiment with machine learning. No question. Maybe even buy a business (hopefully at a more reasonable valuation than Mobileye’s).
What should leaders of those companies do?
They should get smart about machine learning. They should experiment in developing new products that incorporate data and algorithms.
At the same time, they should stay focused on the reality of what their customers really need at this moment.
Most importantly, leaders should get out from behind their desks once in a while and observe customers interacting with their products.
There’s no replacement for spending time with real living breathing human beings. That simple act #acuts through industry hype, confirmation bias, and arrogance. Seeing the world from the point-of-view of customers also helps leaders create strategies that protect from the worst effects of bubbles.
And, of course, everyone should go ask Alexa to play Prince’s 1999 because as of February 12, 2017 Prince is finally available on Amazon Unlimited.
– Phil
P.S. If you’d like to hear more from me, sign up for my monthly newsletter.
