Our brains don't have compartments for different types of information like apps on our devices. We do not have different brain areas that work only on messages, maps, or calendar items. The best way to define agentic AI is in two parts. First, pipes across various apps allow data between apps to mingle. Second, the intelligence part derives intelligence from the commingled data in a way that is more than what single app tools can derive from their data alone.
In some ways, Apple Intelligence heralds a new era of personalized AI, where all the data on your phone, from messages to calendars, is pooled into a personal data lake, then analyzed by an LLM to deliver superior utility. The same could be true for data lakes within corporations. This may spark the first major competition between giants, and not just giants, but all app developers. The transition from their current island-like world to Apple's ecosystem could pose significant challenges, raising the stakes of this competition.
Data is the gold of the digital age, and for app developers, there's little incentive to be submerged in Apple’s lake, especially if Apple reaps the most financial benefits.
Consider Google. A user like this author might have an iPhone but use Google for email, calendar, and contacts. Google has built its services over decades, and allowing Apple to leverage that data would be akin to handing over a pot of gold. Once app developers relinquish data sovereignty, it's a slippery slope.
Against their protectionist tendencies, there will be legal and user pressure to share data, and smaller developers (or those with a small number of apps like TikTok or even Meta/Amazon) will be particularly vulnerable. Their defense will rely on the growing power of the hardware provider.
For Apple, ecosystem control will tighten if it is able to create its dream data lake. This could potentially reshape the tech industry. It will certainly be able to afford to throw in the LLM on top free! Apple’s value proposition of better user experience along with better custody of user data will be difficult to refute, except with similar/better services by others and stiff protection of in-app data.
This raises questions about Apple's partnership with OpenAI. Could a tie-up with Google, despite potential antitrust concerns, have been a more strategic move? Either way, Google's response to Apple's announcement will be particularly intriguing.
