AI video is at another level
When we experienced the energy requirements of different models, the energy we needed to produce a low quality, five seconds of video is very surprising: it was 42,000 times higher than the amount needed for chat boot, which was in response to a question about a prescription, and for more than an hour it is enough to provide microwaves for electricity. If there is a kind of AI whose energy appetite bothers you, this is the same.
Immediately after our publishing. Later, Google started the latest repetition of its VEO model. People quickly produce compilations of highly impressive clips (This one The most shocking for me). What we point out in the story is that Google (as well as an open video generator, Surah) denied our request for specific energy numbers using their AI model. Nevertheless, our reporting shows that high -definition video models like VEO and SORA are very large, and more energy demands we have experienced than these models.
I think the use of AI video will create an unforgivable cloud of emission in the near future, how it will be used, and how it costs it. For example, the example I have brought shows a group of tricky -style content, and I predict that if making AI video is cheap enough, such content will sink social video sites.
You have more important questions than your own individual image
We expected that many readers would think understanding this story in terms of their individual maps, wondering if their AI use was playing an important role in the climate crisis. Don’t panic: It is likely that asking a chat boot to help the travel plan does not increase your carbon impressions meaningful. Video generation can happen. But after months of reports for months, I think there are more important questions.