Artificial intelligence is difficult to visualize, which is why attempts often fail. Google Deepmind offers a free remedy, but it also wants to paint the picture of its core technology as positively and aesthetically as possible.
If you type “artificial intelligence” into Google’s image search, you’ll see the same images over and over: “streams of code, glowing blue brains or white robots with men in suits,” as Google Deepmind puts it. After VR stock imagery, AI imagery is perhaps the most cliched, boring, absurd, and repetitive of all.
These “misleading visualizations” are a thorn in Google DeepMind’s side, as they would negatively shape the public’s perception of AI and create unrealistic expectations.
“Imagery of AI often excludes global perspectives and this lack of diversity can amplify inequalities in society.”
Google Deepmind offers high-quality AI visualizations for free
That’s why the company has launched the “VisualisingAI project”, in which it’s working with artists from around the world “to give a more multi-dimensional representation of this society-changing technology.”
Google DeepMind explicitly states that the images were created by humans. Inspired by conversations with scientists, engineers, and ethicists at Google DeepMind, the visualized topics range from large-scale language models to AGI and robotics to sustainability and generative AI.
According to Google Deepmind, the project has produced 100 images that have been viewed more than 100 million times and downloaded 800,000 times. The images have been used by media, research, and society organizations.