Apple AI – From Zero

Almost anything will be an improvement.

• The scene is set for WWDC to be the launchpad for Apple’s ambitions in AI and given its history in this area, almost anything will represent a significant improvement.
• To be fair to Apple, its products already use AI but it’s the sort of AI that runs in the background which makes the user experience better, but no one is really aware that it is there.
• Good examples of this would be the excellent algorithm that optimises battery utilisation based on usage patterns and the AI that is used to improve its photography functionality.
• However, it is here where the confusion begins because, once again, everyone is getting mixed up when it comes to what actually is AI.
• In 2016 everyone thought that AI meant deep learning and in 2024 everyone thinks that AI means generative AI but the reality is very different.
• Artificial Intelligence is a broad church in that it encompasses everything from simple statistical analysis, software, deep learning and now generative AI.
• Apple has been using deep learning in its products for some time and in some specific areas, to great effect.
• However, where it has come up short is in AI-powered systems that interact directly with the user.
• These are agents like Siri, Alexa Google Assistant and now ChatGPT, Claude 3 and here Apple appears to be way behind.
• This is supported by the fact that Siri has been 3^rd rate for years and Apple has done very little to improve it since most people lost interest in digital personal assistants over 5 years ago.
• However, all that changed in 2022 when Open AI launched a chat system that could converse extremely well in natural language even if it does have a tendency to make things up.
• Since then, hype has gone to fever pitch and now all of the big ecosystems are under pressure to offer generative AI to their users or face having their users go elsewhere.
• Whether this is a real reflection of the true state of affairs has yet to be seen.
• This has put Apple in a difficult position as it clearly was unprepared for the rise of generative AI and its secrecy means that it never usually speaks about anything before it launches.
• I am not convinced that it has anything ready to go and so the keynote at WWDC may contain commentary aimed at placating those screaming for generative AI on iOS without actually launching a generative AI chatbot.
• I think that the most likely outcome will be a promise to upgrade to Siri such that it is powered by generative AI which is precisely what Amazon has done with Alexa.
• The only issue here is that Siri is currently implemented on the device whereas Alexa is completely in the cloud and given that generative AI at the edge is in its infancy, I suspect that Siri would have to return to the cloud.
• I suspect that Apple will also focus on generative AI in computational photography which is something that Google through Samsung has done quite successfully.
• This would mean adding functions like moving objects within a picture, making them bigger, filling in missing areas and so on.
• This would seem to be the most sensible approach and would involve talking about long-term Siri evolution and then adding some useful and cool features that users can actually get some utility from now.
• Generative AI is all the rage, but there is still a lot of uncertainty in terms of how a generative AI chatbot will make one’s digital life on a smartphone better.
• Consequently, I think Apple will talk a good game on generative AI at WWDC, but the features launched are likely to be focused on computational photography which has an immediate use case and which its competition is already featuring.
• I continue to think that Apple’s weakness in AI generally is not going to hurt demand for its products yet which gives it some time to get its offering in line.
• Should the AI bubble burst and interest flag, then this would be a net positive for Apple giving it more time to get its offering right and it would have to pay less for talent and equipment.
• Apple remains expensive for a company facing the multiple challenges of no growth, China and AI weaknesses and so I remain unexcited about taking a position.