OpenAI Taps Google in Unprecedented Cloud Deal Despite AI Rivalry

OpenAI plans to add Alphabet's Google cloud service to meet its growing needs for computing capacity, Reuters reported Tuesday, marking a surprising collaboration between two prominent competitors in the AI race. From the report: The deal, which has been under discussion for a few months, was finalized in May, one of the sources added. It underscores how massive computing demands to train and deploy AI models are reshaping the competitive dynamics in AI, and marks OpenAI's latest move to diversify its compute sources beyond its major supporter Microsoft, including its high-profile Stargate data center project.

It is a win for Google's cloud unit, which will supply additional computing capacity to OpenAI's existing infrastructure for training and running its AI models, sources said, who requested anonymity to discuss private matters. The move also comes as OpenAI's ChatGPT poses the biggest threat to Google's dominant search business in years, with Google executives recently saying that the AI race may not be winner-take-all.

Bad news for NVDA

[Pinky's Brain]

The apparent width of the CUDA moat just got reduced.

Sure lazy academics will stick to CUDA, but apparently for the big boys jumping the moat isn't all that hard.

Re:

[luvirini]

There never was a proper CUDA "moat".

Sure it is a convinient way to do some things with a lot less work than other approaches, but most sane people doing anything on large scale will eventually want to optimize the actual running of things. As long as the development was fast enough in algoritms and hardware that any incremental gains from optimiation were less important such was obvioisl low priority. But once the number of users for a service grows huge, even say a 5% imprevement is suddenly huge numers i

Building the slop cloud

[xack]

The "traditional" internet has already been fully scraped, what they are building now is an internet full of slop. Computer science has already been solved, WWDC was just rehashing themes from 20 years ago. Schools and universities are going to rapidly go back to 1950s style teaching to escape the slopnami.

Re:

[nightflameauto]

The "traditional" internet has already been fully scraped, what they are building now is an internet full of slop. Computer science has already been solved, WWDC was just rehashing themes from 20 years ago. Schools and universities are going to rapidly go back to 1950s style teaching to escape the slopnami.

Sadly, it seems right now that schools and universities are turning to the slopiverse en masse. Since most schooling in the west is geared toward creating worker bees, they need to indoctrinate folks to the "new normal" of the slopiverse being the dictator of terms for reality. Business will run on the slopiverse. Governments will rely on the slopiverse. Oligarchs will decide fates based on the slopiverse. And all of us will be subjects of the slopiverse.

If schools turn away from it, it will have to be driv

exhaustive listing of meteorological front types

[Pseudonymous Powers]

When yet a sixth LLM story drops in a single day, I think people like me need an official thing that we can talk about instead.

Re:

[know-nothing cunt]

This [ucar.edu] will get us started.