I prefer more open systems, so of course I lean toward supporting the EU’s DMA. Third-party marketplaces on iOS represent real progress that wouldn’t have happened without regulation. But what about Siri AI? It is not coming to the EU because Apple doesn’t trust other developers with access to user data via third-party App Intents.

It struck me today what a contrast this is with the rest of the AI world, increasingly connected via MCP. Imagine if Gmail was only accessible to AI on Android phones. If Teams was only accessible from a Windows Surface laptop. That is the kind of environment Apple’s policies create, with data locked into silos and monopolies free to leverage their existing strengths into dominance of new technologies too.

No thanks. Just as I should be able to sideload an app onto my own phone, I should be able to choose whether ChatGPT can access my iMessage history. The world is not going to come crashing down by giving users control of their own data.

Ironically, Apple’s approach to AI — driven by a well-meaning but almost fanatical devotion to privacy — is also going to cripple their ability to build a truly universal Siri that works across even their own platforms. That leaves a small opening for competition.

Apple is betting that Siri AI will be perfect for most iPhone users, and they’re probably right. But it’s a fairly limited vision that, while safe for users, forecloses most of the potential upside of truly powerful AI to work on users’ behalf in the future.

As we get closer (next year) to the 10th anniversary of Micro.blog, I’m thinking about whether our strategy with indie blogging is still relevant for the next decade. What is true now that will continue to be true years from now? Some thoughts…

• While there will be a lot of robots generating content on the web, people will still want to write and own their work. If a large percentage of the web is fake, that will mostly be felt in new curation costs for platforms. It won’t diminish people’s desire to write.
• We’re unlikely to see new social protocols beyond the ones we have now: ActivityPub, AT Protocol, Nostr, and the IndieWeb. It will be hard to justify building something new that isn’t based on one of these. Some company founders will still try to create silos and those will fail.
• Reducing friction in blogging will manifest in new ways. Not just de-emphasizing the title field and other controls, but also using voice more: audio narration for posts, and tools to dictate blog posts. Wherever possible, this can be aligned with podcasting.
• AI will continue to be divisive. The best products will find a way to balance features that appeal to users who hate AI and those who love it. I blogged in January about how Micro.blog is going to navigate through this.
• People will be exhausted by the noise and online drama. They will feel overwhelmed by the abundance of content. Some people will step away from the social web or be drawn to platforms that make the web a more peaceful, less distracting experience.

I’m feeling good about everything we’re working on and how our priorities align with where the web is headed. Inkwell in particular will have a big role, eventually becoming easier to use as a standalone product without Micro.blog.

Thanks as always for everyone’s support on this journey to make the web a little better.