Tech Titans Are Working on AI That Creates AI

Artificial intelligence’s reach into the technology industry is immeasurable right now. Every big tech company is hiring up AI, machine learning, and data science experts with large salary and benefits packages.

But even though the demand for AI knowledge is huge, the supply of experts and experienced workers is low. So of course, companies thought, “Why can’t we just build AI to build…more AI?” The main goal is to create AIs that can create and improve other AIs. And once it’s built, this type of AI can improve itself and finetune its skills as well.

Google announced the AutoML project earlier this year, which automates the design of machine learning and deep learning processes, going so far as to help you choose the right neural network architecture. It’s not automating everything humans do, but it’s helping experts cut out some of the basic, monotonous work. AutoML also learns how to build other machine-learning algorithms.

It’s not just big tech companies getting into AI; mobile app developers are incorporating AI into their own apps. Mobile advertising startups are utilizing AI to bring personalized ad experiences to users. And with Google’s addition of it’s machine learning platform, TensorFlow Lite, to Android Oreo, it’s easier than ever for Android app developers to pick up a few new marketable skills.

TensorFlow Lite is just one example of big tech companies shipping or selling software that allows smaller companies and individual developers to innovate with AI and machine learning. “We are following the same path that computer science has followed with every new type of technology,” says Joseph Sirosh, a vice president at Microsoft. “We are eliminating a lot of the heavy lifting.” Microsoft recently announced a tool to help build deep neural networks, the hottest concept in AI right now.

“Computers are going to invent the algorithms for us, essentially,” said University of California, Berkeley professor, Pieter Abbeel. “Algorithms invented by computers can solve many, many problems very quickly—at least that is the hope.” At Berkeley, researchers are working on ways for robots to learn new tasks while knowing what they’ve experienced in the past.

Even though we have some time until that’s achieved, it’s coming sooner rather than later. AI app development and machine learning will continue to evolve, grow, and innovate your mobile experiences. And though no one knows exactly what our digital future looks like, it’s definitely heartening to see large tech companies releasing technology for everyone to learn with.