Lizette Chapman (Bloomberg) -- Artificial intelligence was a no-brainer bet for humans in 2018. Venture capital funding of AI companies soared 72 percent last year, hitting a record $9.3 billion, according to a new report from PwC and CB Insights.
The surge comes after three years of steadily increasing investment, with the average annual increase from 2015 to 2017 at 28 percent. The recent boost partly reflects the frothy funding environment overall: venture investing in the U.S. hit $99.5 billion in 2018, the highest level since 2000.
It also indicates larger excitement around the industry. College students last year enrolled in introductory AI and machine learning classes in record numbers, the number of academic papers on the topic shot up and officials mentioned the technology in more than 70 meetings of the U.S. Congress, according to Stanford University analysis of transcripts.
AI technology has matured in recent years, as more companies have started using predictive algorithms and other automated techniques across myriad disciplines. In the fourth quarter, the largest AI deal in the U.S. was a $400 million funding round raised by synthetic biology startup Zymergen Inc. Zymergen uses AI-powered robots to genetically engineer microbes to create new materials like flexible glass and improve existing ones like paint that resists radar detection.
The largest deal in AI in the U.S. for all of 2018 was self-driving car startup Zoox Inc.’s $500 million funding round. And even that was dwarfed overseas by the $600 million investors plowed into Beijing-based SenseTime Group Ltd. SenseTime sells software that recognizes people and objects, and with a $4.5 billion valuation, is the world’s most valuable AI startup.
But the growth hasn’t been distributed evenly. For the first time since at least 2013, the number of deals in AI decreased, even as total money invested surged -- an indication that investors are beginning to concentrate their bets on the companies they see as the winners. The trend toward selectivity was especially pronounced in seed-stage startups, with seed-stage deals falling from 39 percent of total deals in 2017 to just 30 percent in 2018.
"There’s big overcrowding at the seed stage," said CB Insights intelligence analyst Mike Wholey. "It’s gotten really easy with open-source products to start an AI company."