Making AI Headlines · Page 6
Fifteen Seconds of Video, Now With Thirty Percent Less on the Bill
· New York time
Google DeepMind has made Gemini Omni 1.2 Flash generally available, offering AI-generated video clips of up to 15 seconds and conversational, step-by-step editing. At 720p, the model costs $0.07 per second, which is 30 percent less than its predecessor, according to the reported pricing. The release gives people making short videos a longer clip and a lower stated cost per second. Users can also draft at low resolution and upscale the final result, potentially reducing costs further. The conversational editing feature lets a user make successive changes rather than treating each initial result as the only version. For creators and businesses, the price and duration are practical details. They affect how much experimentation fits into a budget and how much material can be produced for a project. They do not, on their own, settle whether a clip is useful or whether it belongs in the finished work. That leaves the editor with an updated version of an old problem. More affordable footage makes it easier to generate options, and each option still asks for attention. Google has made the seconds cheaper; it has not made the decision about which seconds deserve to survive any shorter. The final cut remains a scarce resource, especially when the machine is happy to make another draft.
Read this story in the full newspaper →Satire. Not factual reporting. Read the original reporting from the linked publishers.