Watch On:
Summary
There are some firms, who have taken a more relaxed approach to safety in the name of getting their product to market, and then there’s Waymo. It’s done this with extensive computer simulation as well as real-world testing, and its Waymo Driver program is not only operating in several cities in the US, it’s doing so with an incredible safety record. In these simulations, the Waymo Driver AI was able to avoid all of the collisions. In contrast, the benchmarked competitor, the fictional NIEON (non-impaired, with eyes always on the conflict) human driver, could not.
Show Notes
Although the effect is rather crude, the system offers an early glimpse of what’s coming next for generative artificial intelligence, and it is the next obvious step from the text-to-image AI systems that have caused huge excitement this year. In the last month alone, AI lab OpenAI has made its latest text-to-image AI system DALL-E available to everyone, and AI startup Stability.AI launched Stable Diffusion, an open-source text-to-image system. They’re also trickier to train, because there aren’t large-scale data sets of high-quality videos paired with text. To work around this, Meta combined data from three open-source image and video data sets to train its model. Standard text-image data sets of labeled still images helped the AI learn what objects are called and what they look like.
Source
https://www.technologyreview.com/2022/09/29/1060472/meta-text-to-video-ai/
