Apple says its OpenELM model doesn’t support Apple’s AI amid YouTube controversy

Earlier this week, an investigation revealed that Apple and other tech giants used YouTube captions to train their AI models. This included more than 170,000 videos from the likes of MKBHD, Mr. Beast, and more. Apple then used this dataset to train open-source OpenELM models, which were released in April.

Apple has now confirmed 9to5macHowever, OpenELM does not support any of Apple’s AI or machine learning features – including Apple Intelligence.

Apple says it has created OpenELM Model As a way to contribute to the research community and promote the development of large open source language models. In the past, Apple researchers have described OpenELM as an “advanced open language model.”

According to Apple, OpenELM was created for research purposes only, and is not for use in powering any Apple Intelligence features. The model is open source and widely available, including on Apple’s Machine Learning Research Site.

Since OpenELM is not part of Apple Intelligence, this means that the YouTube Subtitles dataset is not used to power Apple Intelligence. In the past, Apple has stated that Apple Intelligence models are trained “on licensed data, including data selected to improve certain features, as well as publicly available data collected by our web crawler.”

Finally, Apple also told me that it has no plans to build any new versions of the OpenELM model.

like Wired Reported Earlier this week, companies like Apple, Anthropic, and NVIDIA used this dataset, “YouTube Subtitles,” to train their AI models. The dataset is part of a larger collection called “The Pile,” from the nonprofit EleutherAI.