2016 was supposed to be the year the tech bubble finally burst.
Much like the dot-com bubble of the early 2000s — an industry implosion marked by high-profile flops such as online grocery delivery startup Webvan and pet supplies retailer Pets.com — skeptics pointed to less VC funding in 2016, stratospheric valuations including Uber’s $69 billion, the sales of once-pricey companies such as One King’s Lane, and sky-high rental and real estate prices.
And contrary to tech insiders who largely remain bullish on the industry, some even saw smaller signs of a bubble in the hours-long bumper-to-bumper traffic on the US-101, a highway that meanders its way down the peninsula to tech-laden cities such as Menlo Park, San Jose and Mountain View.
But after more than six years in Silicon Valley collectively, I’m convinced there isn’t one big bubble these days, but rather a series of smaller “bubbles” within tech that balloon and swell until they burst, taking with them the droves of copycat derivatives and poorly managed companies all trying to capitalize on the latest, frothiest trend.
Ask just about any venture capitalist at this moment, and they’ll tell you they’re seeing a glut of artificial intelligence and machine learning startups flow their way angling for cash, employing increasingly complex algorithms across a wide range of industries.
While some of these new companies may fulfill actual needs, there may simply be more AI startups than the world needs.
Of course, some AI startups are more promising than others. Andreessen Horowitz general partner Vijay Pande told Yahoo Finance he is particularly bullish on companies such as Freenome, which the firm invested in last June. The Palo Alto-based startup uses machine learning to help detect different types of cancers from a blood test rather than from a tissue sample — a process that detects cancer long before more traditional methods can. Another startup Pande invested in, the health tech startup Cardiogram, is promising because it makes sense of and analyzes large amounts of user data to provide actionable insights that could ultimately save lives.
Some A.I. ventures are trying to shake up other long-standing industries, like the San Carlos, Calif.-based Farmers Business Networks, a social network for, well, farmers, that relies on machine learning to improve data results around seed performance and pricing. And there are many, many more.
While it’s too early to tell which of those startups will evolve into viable businesses and which won’t, it’s relatively easy to look back over the last decade now to see past “micro-bubbles” for what they actually were.
Alex Mittal, CEO and co-founder of the FundersClub, an online VC firm which invests in promising tech startups, agrees Silicon Valley has found itself swept up in macro-trends over the years that come and go in predictable cycles.
“Every time there’s a focus on a technology that’s new, it gets overhyped, and the hype reaches an extreme,” Mittal told Yahoo Finance. “The pendulum always seems to swing too far, and there’s some sort of correction. Sometimes, it literally was just hype. There’s no substance, and then it goes away. But sometimes, there’s something really there.”
The 2008 financial crisis, interestingly, marked the first micro-bubble, marked by the “sharing economy,” a business model based on the idea that assets or services are shared between people through the internet or mobile. Airbnb, founded in 2008, singlehandedly legitimized the idea of couch-surfing as a hotel alternative, by easily letting people rent out a room, an apartment or a home; Uber in 2009 upended the crusty, old taxi industry by creating a network of private drivers reachable with just a few easy taps on the smartphone.
But while Airbnb and Uber have become bona fide global businesses, many more sharing economy upstarts failed to catch on. Remember the Uber copycat Sidecar? Shut down in 2015, because Uber and Lyft had more money and an easier-to-understand user experience.
How about laundry delivery ventures Prim and Washio? Shuttered in 2014 and 2016, respectively, due to low profit margins and high infrastructure costs. Even businesses like Homejoy, an online marketplace for cleaning services, hit the skids — despite many a venture capitalist crowing about what a promising business it was — apparently due to a lack of repeat customers and slew of lawsuits.
On the heels of the sharing economy bubble came a slew of e-commerce startups like online design store Fab.com. It burned through a significant chunk of the $325 million it raised in aggressive attempts to expand globally, acquiring similar sites in Germany and England, before a spectacular crash-and-burn that few in Silicon Valley, including its investors, will forget anytime soon.
Meanwhile, once-promising home furnishings site One King’s Lane, which failed to differentiate itself enough from the glut of flash-sale sites, sold for just $12 million last August to Bed Bath & Beyond — a serious markdown from its $900 million valuation of yesteryear — and online furniture retailer Dot & Bo used up its $20 million in funding before shuttering last September.
The most recent micro-bubble to burst? On-demand food delivery startups. No less than a dozen food delivery startups have shuttered over the last 18 months, with names like Bento, Spoonrocket, Din, Kitchit, Kitchen Surfing, and the creatively-named Take Eat Easy. Others like Munchery, Zesty and Sprig, trudge on, but with considerably downsized workforces. Because, while people certainly enjoy good dining, there were too many startups for San Francisco locals for them to keep track of and not enough interested mouths to feed. Indeed, Din founders Emily Olson and Rob LaFave pointed to an overly crowded market as a key reason for closing the startup in a postmortem interview with SF Eater in October.
Many of the venture capitalists and founders I’ve spoken to in recent months are hopeful that this latest boom in A.I. and machine learning startups isn’t part of another micro-bubble in the way many sharing economy and e-commerce startups came and went in the past, largely because these technologies can ostensibly benefit and improve any industry, from health care to agriculture to consumer-focused virtual assistants. (Hello, Alexa.)
“A.I. is probably more accessible than it has ever been before,” contended Peter Cahill, an authority on A.I., who has spent the last 15 years studying speech technology and neural networks from Dublin, Ireland. “It’s easier for companies to see the clear benefits from it because technology has largely caught up.”
Maybe they’re right this time, or maybe the Farmers Business Networks of the world will eventually join the startup graveyard, alongside Fab.com, Homejoy and so many others. But as Mittal points out, this almost blinding sense of optimism — that any startup with a good idea can succeed — is what makes Silicon Valley unique and, dare I say it, innovative. Because for every 100 startups, 10 of them may become successful, and perhaps one has the potential to become the next transformative company like Facebook (FB).
“It’s a large part of the reason I’m still here,” Mittal confesses, with a sheepish grin.
No doubt many in Silicon Valley would agree.
JP Mangalindan is a senior correspondent covering the intersection of business and technology.
More from JP: