Another day brings yet another salvo in the artificial-intelligence wars.
No sooner had
Alphabet
introduced an upgrade to its AI model than OpenAI struck back with a new video-generation tool and was reported to be taking aim at the search-engine market.
Microsoft
-backed OpenAI on Thursday made a buzzy public announcement of the launch of its text-to-video AI tool Sora. While the tool is only open to a small number of researchers, academics, and visual artists so far, the resulting creations look impressive.
OpenAI CEO Sam Altman late on Thursday posted various videos generated from prompts sent in by users of social-media platform X, formerly known as Twitter.
Text-to-video creation is one of the areas where rivals such as
Alphabet’s
Google and Facebook owner
Meta Platforms
had raced to showcase their own AI tools first, with the potential for selling them to advertisers and other content creators. However, like OpenAI, the companies have been cautious about introducing them for wider use amid controversies about deepfake videos and other potentially malicious uses.
“With this announcement of a model that can generate crisp, 60-second videos, OpenAI once again showed that its internal product developments are pushing the envelope and that its releases tend to set new bars for state-of-the-art,” Macquarie analyst Frederick Havemeyer wrote in a research note.
The Sora-generated videos are undoubtedly cool. However, an arguably more important development was a report from The Information that OpenAI was developing a web-search tool, partly powered by
Microsoft’s
Bing; the report cited someone with knowledge of OpenAI’s plans. OpenAI didn’t immediately respond to a request for comment on the report.
Such a move would be a much more direct challenge to Google’s core business and follow the path of start-ups such as Perplexity, which are hoping to challenge the search-engine’s dominance.
Alphabet stock fell 1.4% in early trading Friday after dropping 2.2% the previous session. The company on Thursday announced an upgrade to its Gemini AI model, which it called Gemini 1.5 and said would be available initially to developers and corporate clients.
Elsewhere,
Meta
is facing a challenge to its application of a monthly fee in the European Union for use of its Instagram and Facebook social networks without advertising. The charge was introduced in November in reaction to the EU’s demands to allow users to opt out of personalized advertising based on user data.
A group of 28 nongovernmental and consumer-rights’ organizations publicly called on the European Data Protection Board to oppose the model on Friday. The EDPB is set to issue a judgment on the matter after the Dutch, Norwegian and Hamburg privacy watchdogs requested a binding opinion.
Meta didn’t immediately respond to a request for comment. Its shares were down 2.2% in early trading Friday.
In the age of AI, expect plenty more fights over privacy as the tech giants become even thirstier for data to feed their models.
Write to Adam Clark at [email protected]
Read the full article here
