Google's Teachable Machine Screen Google

Google Unveils 'Teachable Machine' to Showcase AI to Users

The machine learning simulator requires no coding and lets users explore how AI works by using images and other simple tools.

Google has unveiled an AI experiment that's designed to expand and promote the understanding of artificial intelligence and machine learning by allowing anyone to try their own online AI experiences.

Called "Teachable Machine," the project is a simple experiment that lets online participants teach a machine using their computer's camera right within a web browser, with no laborious hand coding required, wrote Barron Webster, a designer and machine learning enthusiast in Google's Creative Lab, in a recent post on The Google Official Blog.

"From helping you find your favorite dog photos, to helping farmers in Japan sort cucumbers, machine learning is changing the way people use code to solve problems," wrote Webster. "But how does machine learning actually work? We wanted to make it easier for people who are curious about this technology to learn more about it. So, we created Teachable Machine."

Teachable Machine "is the most direct example showing how the learning process itself works, but there are lots of other fun ones demoing what can be done with machine learning," a Google spokesman told ITPro in an email reply.

Using the experiment, participants can explore how machine learning works, in a fun, hands-on way by "teaching" a machine to use your computer camera, live in the browser, according to Google. Users train the neural network locally on their device, without sending any images to a server, which allows it to respond quickly to inputs.

One participant used the Teachable Machine to make their hand say "moo," while others have performed other demonstrations. Visitors can use the record button to share their contributions on social media with the hashtag #teachablemachine.

There is no correct way to use the experiment. "Keep playing around," Google advises users. "Seeing what works and what doesn't is one way to explore how machine learning works. Keep in mind that your machine doesn't have an understanding of higher level concepts, like faces or objects. It's learning through the examples you give it. So if it's not working the way you want, you might want to click the x to reset your classes and try out different approaches."

The experiment is built with a new hardware accelerated machine intelligence coding library called deeplearn.js, which makes it easier for any web developer to get into machine learning by training and running neural nets right inside a web browser. Neural nets are interconnected groups of compute nodes, which together simulate the interconnected nodes of a human brain. Google makes the code available as open source to inspire others to build additional new experiments. Deeplearn.js was originally developed by the Google Brain PAIR team to build powerful interactive machine learning tools for the browser. The library can be used for everything from education, to model understanding, to art projects, according to Google.

The Google PAIR team is devoted to advancing the research and design of people-centric AI systems, using the full spectrum of human interaction with machine intelligence, from supporting engineers to understanding everyday experiences with AI.

Teachable Machine is not an official Google product, but is instead a collaborative effort by Støj, Use All Five, Google's Creative Lab and Google's People+AI Research Initiative (PAIR) team.

Google's AI Experiments website is a showcase for simple experiments that make it easier for anyone to start exploring machine learning, through pictures, drawings, language, music, and more. Since 2009, coders have created thousands of experiments using Chrome, Android, AI, WebVR, AR and more. 

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish