Machine learning, the branch of artificial intelligence that helps computers learn without being specifically programmed, has quickly emerged as table stakes for companies to remain competitive in everything from media and advertising to enterprise computing.

Now, it’s apparent that Twitter Inc. aims to keep up in machine learning with giants such as Google, Facebook, IBM and Microsoft. Today, it announced it’s buying a U.K. startup called Magic Pony for a reported $150 million.

Twitter Chief Executive Officer Jack Dorsey (pictured) wrote in a blog post that the startup’s technology, in particular its deep learning neural networks that attempt to emulate how the brain works, is intended to boost its push in video and its renewed focus on live events–both of which are also activities that Facebook CEO Mark Zuckerberg is zeroing in on as well recently.

Magic Pony’s technology – based on research by the team to create algorithms that can understand the features of imagery – will be used to enhance our strength in live and video and opens up a whole lot of exciting creative possibilities for Twitter. The team includes 11 PhDs with expertise across computer vision, machine learning, high-performance computing, and computational neuroscience, who are alumni of some of the top labs in the world.

Indeed, it seems apparent that the deal, like many in the hot field of machine learning, is as much for the engineers and scientists as the technology, for which Magic Pony has filed 20 patents. Thanks to rampant hiring by Google and others, AI researchers are in high demand. Magic Pony’s team will join Twitter Cortex, its team of engineers, machine learning researchers and data scientists, for which it has a number of openings it’s trying to fill. According to one investor, Suranga Chandratillake, a partner at Balderton Capital, this was a quick exit.

Discussions between Twitter and Magic Pony, which was founded only in 2015, “very rapidly escalated from a casual conversation to a strategic conversation,” Chandratillake said in an interview from Copenhagen. He said Twitter was interested in both the technology and the team but acknowledged that “there’s often only a handful of experts in a particular AI field.”

Third AI acquisition

This is the third Twitter acquisition of a machine learning startup in the last couple of years. In 2014, it bought New York-based Madbits, which helped Twitter improve its image search. Last year, it also acquired and then shut down Whetlab, which helped automate machine learning experiments.

Along with those acquisitions, Magic Pony is intended to bolster Cortex’s attempt, as Dorsey put it, to build “a product in which people can easily find new experiences to share and participate in.” In other words, Twitter aims to apply machine learning to get more people to spend more time on the site.

More specifically, Magic Pony’s technology can improve an image by adding details, a feature that could be valuable on smartphone photos and videos. Twitter usage is almost all on mobile devices and it’s continuing to push its Periscope live video app and its Vine video sharing app. According to an MIT Technology Review article in April, the technology could be applied to improve video streaming or live streams of video game playing, even on mobile graphics processing units (GPUs).

The big advantage to Magic Pony of getting acquired by Twitter, said Chandratillake, is that “Twitter can offer them access to massive amounts of data they can’t get on their own.” That will allow the technology to advance faster, he said.

Taken together, the three acquisitions add to Twitter’s attempt to remain competitive in a key capability of the next age of computing. But given Twitter’s many struggles lately, especially its slowing growth, the company will find it tough to keep up with much bigger and more well-heeled rivals.

Image credit: jdlasica/Flickr/CC by 2.0