Apple intelligence and AI maximalism


This post is by Benedict Evans from Essays – Benedict Evans


No-one outside Apple has really used any Apple Intelligence features yet. It won’t launch until the autumn, in stages, and even then it won’t work on 80 or 90% of the iPhone installed base, since it needs on-device processing that’s only on the iPhone 15 Pro. Nothing is ever as good as the demos (except the original iPhone), Apple might mess up the developer incentives again, and it all might just not be as good as it looks. We’ll find out.

But meanwhile, if you step back from the demos and screenshots and look at what Apple is really trying to do, Apple is pointing to most of the key questions and points of leverage in generative AI, and proposing a thesis for how this is going to work that looks very different to all the hype and evangelism.

The ‘AI Maximalist’ view is that general-purpose chatbots, with multimodal and ‘multi-agentic’ capabilities, will be able to take over broad classes of complex multi-stage tasks and problems that today need lots of different specially-written software, and will also be able to automate whole new class of complex tasks that could never be done in software before, again in one single general-purpose interface. The chatbot might replace all software with a prompt – ‘software is dead’. I’m skeptical about this, as I’ve written here, but Apple is proposing the opposite: that generative AI is a technology, not a product.

Apple is, I think, signalling a view that generative AI, and ChatGPT (Read more…)