All over the world there is a buzz about how data centers are constantly popping up more and more.
But is there really demand for all the computing power being built into data centers?
Well, it seems that there is. At least based on the fact that Financial Times reports (paywall) that Google has significantly restricted one of its biggest paying customers from using AI, because Google simply didn't have enough AI computing capacity available.
The big customer here is Meta, i.e., the parent company of Facebook and Instagram.
Meta uses Google's Gemini AI models specifically in internal product development, and not so much for the operation of user-facing elements. In addition to product development, Meta is reported to use Gemini, among other things, as an aid for Facebook moderation and as a tool for detecting social media scam attempts.
Meta also has its own AI models, under the llama brand, but Google's Gemini has proven to be a more functional tool in many of Meta's AI use cases.
Due to the usage restrictions set by Google, several of Meta's projects are already reported to have been delayed from their planned schedules.
Well, it seems that there is. At least based on the fact that Financial Times reports (paywall) that Google has significantly restricted one of its biggest paying customers from using AI, because Google simply didn't have enough AI computing capacity available.
The big customer here is Meta, i.e., the parent company of Facebook and Instagram.
Meta uses Google's Gemini AI models specifically in internal product development, and not so much for the operation of user-facing elements. In addition to product development, Meta is reported to use Gemini, among other things, as an aid for Facebook moderation and as a tool for detecting social media scam attempts.
Meta also has its own AI models, under the llama brand, but Google's Gemini has proven to be a more functional tool in many of Meta's AI use cases.
Due to the usage restrictions set by Google, several of Meta's projects are already reported to have been delayed from their planned schedules.
