Apple is cooking up its own homebrewed AI chatbot, internally dubbed “Apple GPT” in hopes of taking on ChatGPT and other buzzy AI assistants, according to a new report.
Mark Gurman at Bloomberg News says Apple has been prioritizing its AI and machine learning projects over the past several months. Like ChatGPT and other large language models from Microsoft, Google, and others, Apple’s AI project is built on its own “Ajax” framework, which is based on Google’s Jax machine learning framework and runs on Google Cloud, no relation to the web development tool – and has been tested by some Apple engineers.
Gurman says Apple mulled inking a deal to license some of OpenAI technology, but ultimately decided against it. Apple’s AI chief John Giannandrea and software head honcho Craig Federighi are reportedly leading the charge on Apple’s hush-hush AI initiatives.
It’s a big deal because Apple has access to more intimate data about its users than any other company on the planet. An Apple chatbot would likely be trained on oodles of iMessage chats, search histories, health data, photos, emails, and more – letting it potentially converse in a scarily lifelike way. Despite being fashionably late to the AI party, Apple could still take the crown.
The report says Apple GPT may not see the light of day until 2024 or later. Apple hasn’t announced any plans to release a ChatGPT rival just yet. But with heavyweights like Google, Microsoft, and OpenAI now promoting their own AI chatbots and digital assistants, Apple will feel pressure to keep up with the Joneses. Siri could certainly use a boost too.
For now, Apple is keeping its AI plans tightly under wraps. But the company is making AI and machine learning a priority, so some kind of ChatGPT competitor seems inevitable. Apple GPT or whatever they end up calling it could be the next big leap forward for Siri and Apple’s voice assistant technology. We’ll have to stay tuned to see if and when Apple’s top-secret AI projects come to fruition.