Google wants to serve as a clearinghouse for artificial intelligence (A.I.) models.
The search-engine giant has announced AI Hub, which it bills as a “catalog of plug-and-play A.I. components, including end-to-end AI pipelines and out-of-the-box algorithms.” Data scientists, machine-learning experts, and companies can use the Hub to share and store A.I.-related code for utilization within their organizations.
At first glance, AI Hub seems like a specialized competitor to GitHub. However, Google’s emphasis is much more on intracompany code sharing, as opposed to opening up code to anyone with an idea. Google also offers Hub users a variety of branded tools and platforms, including Google Cloud AI, Google Cloud Partners, and Google AI.
Google clearly wants to pull enterprises onto the Hub with the prospect of streamlining A.I. workflows, then have those companies adopt its A.I. tools. However, the Hub may evolve into a full-fledged A.I. marketplace in coming years, depending on how things go. “We eventually want A.I. Hub to be a place where third parties can also share information and turn it more into a marketplace over time,” Rajen Sheth, senior director of product management for Google Cloud, told VentureBeat. “What we’re finding is that community could actually solve the problems of many of our customers.”
Google launches AI Hub at a time when companies are pouring more resources into artificial intelligence and machine learning (and hiring more A.I. and machine-learning experts). Despite that newfound attention, many of those companies struggle to lock down the specialized talent they need to exploit A.I.-related opportunities (which is why A.I. experts can earn enormous salaries). In theory, something like AI Hub can help ease this strain by allowing organizations to better leverage (and collaborate) with the assets they already have on-hand.
That collaboration may also encourage more employees to learn about artificial intelligence. Fortunately, there are plenty of resources online to serve as an introduction to the art of teaching machines how to think for themselves.
For instance, Bloomberg offers a (free) online course in machine learning, aimed at pros who already have some background in machine learning (and a lot of aptitude for mathematics). Google also has a (likewise free) course with 25 lessons and more than 40 exercises, which you can finish in 15 hours or so; it features lots of video of Google engineers describing the nuances of A.I. and machine learning. Microsoft is offering a Professional Program for Artificial Intelligence, with 10 online courses that teach 10 skills.