Artificial Intelligence Can Now Independently Design New People and Worlds

Christopher Benek

American technology company Nvidia, known by gamers throughout the world for their computer graphics cards, has made a major advancement in Artificial Intelligence. Now AI can create completely artificial images including human and animals that look 100% real. Additionally, Nvidia’s AI can independently design complete virtual worlds on its own.

In a paper titled "A Style-Based Generator Architechture for Generative Adversarial Networks," Tero Karras, Samuli Laine, and Timo Aila of Nvida detail the specifics of the development. The abstract of the report indicates that the new technology “leads to an automatically learned, unsupervised separation of high-level attributes (e.g., pose and identity when trained on human faces) and stochastic variation in the generated images (e.g., freckles, hair), and it enables intuitive, scale-specific control of the synthesis. The new generator improves the state-of-the-art in terms of traditional distribution quality metrics, leads to demonstrably better interpolation properties, and also better disentangles the latent factors of variation. To quantify interpolation quality and disentanglement, we propose two new, automated methods that are applicable to any generator architecture. Finally, we introduce a new, highly varied and high-quality dataset of human faces.”

Included in the paper are detailed AI generated human faces, cars, bedrooms and cats.

This development comes on the heels of Nvidia’s release of a video that shows their AI creating new virtual worlds. Using real life imagery, the AI is able to construct wholly new virtual environments for exploration. The implications of this breakthrough for the continued progression and development of Virtual Reality is tremendous. You can see the full video below.

Of course, while the development of such new technology now opens up the doors for further breakthroughs in AI learning and VR advancement, it also opens up the potential for risk as well. Founder of Wired Magazine Kevin Kelly recently tweeted an image of the AI generated faces. He wrote: “None of these faces are real. All made up by AIs. The end of photographic evidence.”

Issues of identity, privacy, and espionage – just to name a few - come into play as well. There is also potential for further creation of AI-generated fake news. China for instance, has recently shown its ability to utilize AI in its recent creation of a fully AI news anchor. In spite of these dangers though, the breakthrough also offers new potential for creation in the world that has never existed before.

Reality Changing Observations:

Q1. What do you think are the ethical implications of AI being able to replicate your physical image exactly and at high speeds?

Q2. What might the impact be on human existence if AI is able to create VR worlds via human verbal commands?

Q3. How might these advancements in AI and VR help humans to better live into our capacity as creative beings?