Apple’s Subtle Pitch to AI Researchers; Index Backs Former YouTube AI Manager in Latest ‘Vector Embedding’ Deal

Before we get into this morning’s main funding news, here’s an interesting look into the strategy of Apple, which has caught flack from developers and industry pundits for falling behind Microsoft, Google and Amazon in releasing conversational AI. (That’s likely why it’s in talks with Google to license its generative AI tech for the iPhone, according to a Bloomberg report this morning.) In trying to catch up, the iPhone maker is doing something unusual: showing the public what its researchers are up to behind the scenes.

Late last week, the company quietly released an academic paper on how it's building advanced large language models that can process and generate text, image and other types of data. The paper goes into detail about the different types of data and model architectures Apple experimented with to build its best performing model. The disclosures are noteworthy at a time when big labs like OpenAI and Google are getting increasingly tight-lipped about their research. Even those companies’ “technical papers,” which they release alongside new models, don’t say as much as they used to.