Apple AI – Back of the pack

There is a little activity behind the curtain.

• It would be foolish to think that Apple is not trying to do something about AI but I suspect that it has done very little to date and it remains very far behind in a race that is not going to be quite as consequential as the cool-aid drinkers would have us believe.
• However, the one area where I think that Apple is on the money is its view that AI needs to run on the device as this aligns precisely with the findings of RFM’s research.
• The first problem that Apple has is that it is coming off a very weak base when it comes to AI although its ability to design processors remains second to none.
• Siri is easily one of the worst AI digital assistants available today and in RFM’s opinion ranks just behind Alexa and way behind Google Assistant in terms of quality of user experience.
• To date, this has not mattered much because after a wave of hype about how great these assistants would be in 2018, they failed to deliver and so most users became indifferent to their presence on devices and even forgot that they exist.
• However, the advent of generative AI has meant that having or being seen to have a generative AI service is becoming part of the digital ecosystem offering which means that Apple needs to address this somehow sooner rather than later.
• At its iPhone 15 launch in September 2023, there was barely a mention of generative AI despite the hype cycle being at fever pitch highlighting once again just how secretive Apple is.
• Historically in AI, this has been hugely problematic as up until recently, AI was an open and collaborative community but with commercial pressure to make money rising, the curtains have come down and everyone is much more tight-lipped about what they are doing.
• A good example of this is Open AIs GPT-4 vs GPT-3 where all the details of GPT-3 were disclosed in 2020 whereas one can only speculate about the architecture of GPT-4.
• This means that being secretive is no longer really a problem, but all of the signs are that Apple is only starting to address this now which puts it at least a couple of years behind everyone else.
• Many of Apple’s job openings require expertise in deep learning and I suspect that if cannot build its own foundation model, it is likely to acquire one.
• The time for acquisition is not yet ripe as capital is still pouring into the sector meaning that valuations remain sky-high and there are no bargains on offer.
• Once price competition between generative AI services begins in earnest, I think we will see a rapid reduction in pricing and when the money runs out, a fall in valuations as well.
• This is the time to make an acquisition, as good assets will be on sale and the bad assets will be quickly going out of business.
• It is going to take some time for generative AI to make it into battery-powered devices at scale and so Apple has time to get this right although it is not a good look to be seemingly doing nothing.
• Apple is by far the strongest brand, and I think that this will ensure that market share does not wither before Apple has a product ready for its ecosystem.
• That being said, growth is becoming hard to come by making the valuation of the shares look uninteresting from my point of view.
• I don’t own Apple and have no intention of doing so any time soon.