Did you catch Google I/O this week? It’s All the time Attention-grabbing to see what the Android Inventors have been as much as in the course of the previous Annual Interval. 😂 This 17-minute recap will get you up to the mark for those who missed it, however it’s protected to say it had a theme. Cell options didn’t make the massive headlines this yr however had been nonetheless current with Android Studio getting AI-powered code help, a new on-device Gemini mannequin, and updates to Jetpack Compose.
However why am I speaking about that right here? Effectively, with Tim Prepare dinner saying they’re engaged on generative AI options that’ll debut “later this yr” and Greg Joswiak dropping big hints when saying the convention, it’s simple to see how WWDC could possibly be much like I/O this yr.
I assumed it may be enjoyable to assume by way of what that might imply, however to maintain this brief I’ll restrict myself to a few ideas:
First, Siri. It’s an apparent goal for an improve and generative AI is effectively suited to the duty. My solely fear is that it’s going to should be meaningfully forward of the competitors to beat the bias of adverse opinions folks have about present era Siri, and the competitors took an enormous step ahead earlier this week. 😬 I nonetheless hope this occurs, although.
Secondly, will we see a CoPilot-esque characteristic in Xcode? One factor I’d wager on is that if there’s any AI in Xcode, it is going to be strictly on-device. I discover it onerous to consider any massive firm permits its supply code to be transmitted to another person’s server with each keypress, however I discover it past perception that Apple would. If one thing like this ships in Xcode 16, I’d even be curious to know what coaching knowledge it makes use of. I nonetheless fear a bit of about GPL coaching knowledge in CoPilot, so I hope it’s not based mostly on OpenAI’s Codex as CoPilot is. I can also’t see Apple utilizing inside Swift code as coaching knowledge, are you able to? There are many tough questions right here, so I’ll be fascinated to get extra particulars if this characteristic ships.
My closing thought is a protracted shot. It’s not a prediction as a result of I don’t assume it’ll occur. It’s extra of a hope, and I’m together with it right here as a result of I feel Apple would do a fantastic job with it. I’d like to see a portray app that makes use of generative AI to work with your brush strokes to create artwork. It wouldn’t generate photographs from textual content however would use AI fashions that can assist you enhance your sketching and portray expertise. There are a lot of ways in which an app like this might work and differing ranges of help it might present so I received’t speculate on something extra particular, however that is my long-shot hope. That is by far the least more likely to occur, primarily as a result of if it existed it might have been introduced ultimately week’s iPad occasion alongside the brand new pencil. I don’t care, although. It’s on my listing!
Let’s see how fallacious I’m in a number of weeks! 😬