At a recent presentation in San Francisco, CB Insights CEO Anand Sanwal said half-jokingly that if startups want attention from investors, they should put phrases like "artificial intelligence" and "machine learning" in their pitch deck.

While there's an argument to be made that AI is over-hyped as a technology, there's data to back up Sanwal's tongue-in-cheek advice: Mergers and acquisitions of AI startups increased by a factor of seven between 2011 and 2015, from five to more than 35 deals, according to the research firm.

The increase runs against the grain of what the Silicon Valley Business Journal reported as an overall decline in the number of exits and deals of startups across fields starting in the second quarter of 2014.

The poster child for the recent surge in interest in AI: Twitter's acquisition of machine learning startup Magic Pony Technology, announced earlier this month. If you look at the deal primarily as an "acquihire," Twitter is reportedly paying $13 million per machine-learning PhD.

To date, there have been 24 exit deals with AI startups this year, compared to a total of 38 exits that included two IPOs in 2015, according to CB Insights. (When it comes to exits of venture-backed startups, M&As dwarf IPOs as a rule. Data from Reuters provided by accelerator 500 Startups partner Emily Chiu shows there were 372 M&A exits and 77 IPOs of venture-backed startups last year.)

CB Insights points out a couple interesting trends in AI startup M&As:

  • Most AI startups that were acquired in early stages--within four years of initial funding
  • Google is leading acquisitions, with around 10 acquisitions of AI or machine learning startups since 2011

Here's a list of the companies most active in this wave of M&As since 2011, according to CB Insights:

  • Google
  • Twitter
  • Apple Inc.
  • Intel
  • Salesforce
  • AOL
  • IBM
  • Yahoo