This year is the first year when deals in artificial intelligence are really taking off. AI startups have already attracted well over a billion dollars in exits in 2016--and the year isn't quite done. AI's fast progress has attracted other types of attention too, like the recent White House Report on Artificial Intelligence. Or Elon Musk's comments about his fear of artificial intelligence.
Let's take a look at some of the key artificial intelligence deals so far this year. What trends do you spot?
1. Samsung acquires Viv.
Viv was on my list of 15 most disruptive technologies last year. Looks like Samsung thought so, too. Viv was billed as an AI "super-chatbot" that raised $30 million but otherwise was under wraps. However, Samsung wasn't talking about chatbots when it spoke with Tech Crunch: " . . . one way to think about this is that you've got this massive app ecosystem out there. What you'll find in this effort is that we're going to be slowly moving over to this post-app world."
Viv is supposed to dynamically generate programs that help you get what you want, even if that access wasn't explicit before. (Think, "Hey Viv, register me for that event.") The founders of Viv previously created Siri, which was bought by Apple. All indications are that Samsung, like Apple, Amazon, Google and Facebook, are pursuing AI the way they pursued mobility--as a new platform and new frontier.
2. Apple buys Emotient.
Emotient, based in San Diego, helped computers recognize emotions through facial expressions. It helped doctors rate pain for patients that couldn't speak, for example. It had raised $8 million from Intel and others. Apple hasn't confirmed how it wants to use its new skill at reading your expression, but some speculate one potential application is an emotionally-aware Siri or making artificial intelligence personality.
3. Google gets API.ai.
The California-based startup had just raised money from Intel and others in August of last year, when it was bought this September by Google. The speech/text language recognition software helped developers make mobile phone and tablet applications with voice interfaces.
4. Intel is acquiring Dublin-based Movidius for $400 million.
Movidius makes visual processing "AI on a chip." Unlike cloud-based organisms such as Viv or Siri, with Movidius chips you can run visual learning software on-site that doesn't require the cloud for processing. Co-founders Sean Mitchel and David Maloney drew the original system block diagram on a menu card of a cozy Dublin pub in 2005. Investors had $86.5 million invested. Its biggest application yet are drones.
5. Intel acquires San Diego-based Nervana for $350 million.
Nervana developed the open-source developer platform Neon. Neon is famous for simplifying running AI computations. This acquisition makes the list in part for size of acquisition relative to the investment--investors had $24.4 million invested--and in part because it shows the breadth of the AI universe. It's early days in AI. Developer frameworks are still settling. Intel has already hooked the Nervana community to product development and is launching a learning platform chip next year. In other words, the bet is AI is the new "Intel Inside."
6. Microsoft acquires London-based Swiftkey for $250 million.
Swiftkey is one of my favorite examples of machine learning solving a time sucking pain. Over half the world uses more than one language. If you type in more than one language, it's really clunky to "switch keyboards" to continue in the other language. Plus, you probably don't cleanly use one language, but do some mixing --which totally doesn't work with the "keyboard switch" approach.
Swiftkey lets you use multiple languages naturally and learns how to type ahead for you. No more "oops I'm in the wrong keyboard" moments. The team figures it's saved its users over 100,000 years of woulda-been wasted typing time. Investors had $21.49 million in at exit.
7. Twitter acquires London-based Magic Pony for $150 million.
This deal hinges on the 20 patents Magic Pony had and also the team Twitter acquired by buying the business. CEO Jack Dorsey commented, "Magic Pony's machine learning technology will help us build strength into our deep learning teams with world-class talent, so Twitter can continue to be the best place to see what's happening and why it matters, first."
8. In its first acquisition since going public, Etsy acquired Blackbird.
Blackbird is a search recommendation tool that helps shoppers find things similar to what they already like. What's interesting about this, and why it made the cut for this list, is it shows a relatively small company in terms of market cap--Etsy, at $1.4 billion--making a strategic acquisition based on AI. Blackbird had 10 employees and customers like NastyGal, which got booted when the platform become proprietary to Etsy. Etsy's stock went up significantly right after the announcement.
9. Oracle snags Tel Aviv's Crosswise for $50 million.
The application helps determine accurate identity. "A lot of deterministic data is actually garbage," says Steve Glanz, Crosswise founder. "If my kid logs in to a game on my phone with their email address, suddenly my phone is matched to my kid's laptop. It's 'true' and there's a connection, but those two devices don't belong to the same person." Crosswise within Oracle will help create more accurate maps of the overlapping billions of devices and billions of users. Investors had $5 million invested at the time of acquisition.
10. Salesforce buys Metamind for a reported $32.7 million.
Salesforce CEO Mark Benioff joined forces with Khosla to put $8 million in to launch Metamind in 2014. It was acquired by Salesforce soon after building a small customer base. "With MetaMind and Salesforce coming together, we'll be able to offer customers real AI solutions with breakthrough capabilities that further automate and personalize customer support, marketing automation, and many other business processes," MetaMind co-founder Richard Socher wrote on a blog post.
What's ahead for AI acquisitions
The AI acquisition cycle is still young. First movers are mostly large technology companies like Google, Facebook, Intel and Samsung who can scale the value of these resources (or limit the damage to their business by taking them out).
AI is getting easier to develop as the developer platforms and hardware mature. It's also not forbiddingly expensive--just look at the ratio of dollars invested to acquisition price in the examples above. Just my speculation, but in 2017, I expect many more acquisitions from corporations that are not global technology players. Sector and industry specific AI, particularly in manufacturing, customer service, calendaring, cyber defense, fintech, and sales, are all ripe for significant opportunities for automation that further solves real customer pains.
Artificial intelligence acquisitions just beginning
When you consider how fast machine learning and AI is progressing, it's mind blowing. However, not being able to relate to the speed, didn't stop us from breaking the sound barrier. Not being able to relate to the distance, didn't stop us from going to the Moon. On a personal note, I'm confident we'll work toward a useful, beneficial AI and these deals show some of the promise of why it matters.
(These are just 10 deals--if you'd like to see the growing list of 2016 AI deals, it's posted on Valor's blog, The Real Deal.)