Apple reportedly spending ‘millions of dollars a day’ on its own conversational AI

Competition in the Artificial Intelligence (AI) space has been fast heating up, with multiple tech bigwigs taking on each other, including the likes of Microsoft, Google, and of course, let’s not forget OpenAI.

Now, as per The Information, Apple has reportedly been playing catch-up, developing and training a Large Language Model (LLM) based conversational AI tool internally referred to as Ajax GPT. Access to the model is limited to very few even within Apple, and there has been no word on its rollout to the general public as of yet. The team responsible for the same is said to comprise of 16 members and is being led by John Giannandrea, a former Google employee.

However, Apple’s foray into the AI space is believed to not just be limited to conversational AI. According to the report, Apple also has an in-house team working on an AI-based image-generation model, while another is also believed to be working on a model that is capable of generating both – images and videos, as well as text.

Apple’s foray into AI

Apple is reportedly planning to integrate the conversational AI tool into Siri, to help it better compete with OpenAI’s ChatGPT, Microsoft Bing Chat, and Google Bard, amongst others. Ajax GPT has been trained on over 200 billion parameters, which are an indicator of the size and complexity of a machine-learning model. Hence, higher the number of parameters, the more complex the model is, and the more storage space as well as computing power it requires.

ALSO READ: OpenAI GPT-4 is here: What it means for ChatGPT and other AI chatbots

Now, while Ajax GPT is believed to be more powerful than OpenAI’s GPT-3.5, it remains to be seen whether Apple’s next iPhone or iPad will be able to handle such a large and complex model, both in terms of computing power and storage.

It is important to note here, that there has been no response from Apple regarding the report as of yet.