Apple & AI
The AI race is at its peak, and every major tech company has a horse in the race in some capacity. Microsoft has a massive investment in OpenAI, Google has done a full pivot into Gemini, Amazon is catching up with an investment in Anthropic, and Meta is plowing ahead with a massive order of Nvidia H100’s to further train their Llama models.
Strangely quiet in all of this is Apple, who recently released their next big platform with the Vision Pro — which has been otherwise strangely silent around all the other AI chatter happening in the industry. What we do know about Apple is that they would rather be late and polished, than rush out a half-baked product which mostly let’s the public test for future improvements/direction (*cough* Google *cough*). We also know that Apple is always actively exploring how they can leverage their strengths and are unafraid to invest in these areas even if they never see the light of day — such as their recent cancellation of their self-driving and electric car team.
If we can connect the dots, there is very likely a clear strategy for Apple this year, which I’d like to take a moment to speculate on here:
- Gen AI Siri. Siri, along with its peers Alexa and Google Assistant, is a legacy voice assistant that is due for a major overhaul. Apple is likely currently fine tuning its own large language model in hope to replace the engine that is currently underneath Siri. This LLM will likely be two parts. One smaller parameter model that will run locally on devices, that will be able to hold simple conversations while still executing basic tasks like checking the weather, setting timers/alarms, and responding to messages on your behalf. A second, larger parameter model that runs in the cloud is also likely being developed that will power more complex commands such as photo, video, and productivity functions — aiming to be utilized by Vision Pro devices.
- Dedicated AI chip. To power the smaller parameter model, it is also likely that Apple is working on a customized chip similar to their A series that first started to power iPhones, and their M series that later powered their MacBook products. I wouldn’t be surprised if Apple is currently working on a “G” series of chips designed to specifically to run a ~7B parameter language model on devices to handle executing simple tasks and holding basic conversations with users that are more interactive and personal that the ones Siri can currently hold. The separate chip allows the AI to be consistently running in the background and improve its efficiency on a separate path from the general system architecture.
- AI-powered iPhone 16. It’s no secret that the iPhone 15 has been terribly underwhelming. Aside from a switch to USB-C, it seems not much change was introduced in this last iPhone and customers have mostly been holding onto their phones for longer — slowing sales and growth for the most important product within Apple’s portfolio. It’s very likely that this new G1 chip will be introduced in the iPhone 16 Pro line this year — released simultaneously with the announcement of Apple’s new, upgraded Siri powered by Gen AI. This would be the spark to light the fire of the iPhone and bring excitement to Apple’s big move in the AI space and likely be the peak of all this AI frenzy we are experiencing.
Of course, a lot of pieces will need to come together for all of the above to be successfully executed. However, given the fact that the beginnings of these initiatives are already taking place elsewhere (Google is transitioning Assistant to Gen AI, more companies are shifting to designing and producing AI chips, and Samsung is already integrating AI into their phones), it is very plausible that Apple is doing the same behind the scenes — all gearing up for a very exciting September/October season for all of us. And as someone looking to replace their iPhone 12 Pro Max, count me as one hopeful for all of this to be true!
