MC.40: Apple Intelligence is Privacy; AGI Postponed?

Apple Intelligence is Privacy

Apple just announced Apple Intelligence, which is basically a personal AI assistant. The main selling point is privacy for its AI products. By using a "challenger sale" strategy, Apple brands itself as the secure choice in a market with growing data privacy issues. This move might make people prefer Apple and force other companies to take privacy more seriously.

I’m siding with Apple; privacy should be a feature, not an afterthought. Microsoft learned the lesson and made Recall opt-in after outcry; the feature automatically records all user screen activities.

Elon Musk called Apple Intelligence and the integration of OpenAI into its OS "creepy spyware". This is ironic since Tesla has been recording data from its cars from the start. But he has a point: when your data goes to the cloud, you can't be 100% sure how it will be used.

This brings up an important question: would you share your health data if it could improve your health and help others? Google has developed a Personal Health Large Language Model that uses wearable data to provide insights and act as your personal health assistant. This raises the debate of privacy versus wellness.

Local AI ecosystems will thrive; AGI postponed?

Meanwhile, investors like General Catalyst are tripling down on Mistral AI with a €600M Series B, betting the AI market won't be winner-take-all due to local and national interests. They believe that strong local AI ecosystems will thrive alongside giants like OpenAI.

But the path to Artificial General Intelligence (AGI) remains unclear. The ARC Prize benchmark, meant to test general intelligence, shows AI progress is lagging, with top systems scoring only 34% while humans perform much better. Also, we might run out of human-made training data for large language models by 2026, which could limit further growth.

This serves as a counterpoint to a controversial 150-page manifesto claiming AGI is imminent. Despite this, OpenAI and others remain focused on the AGI goal. The strong use of AI by students and teachers shows that even small steps forward can be really useful. Yet, we need guardrails to prevent mass surveillance, as seen in a Shanghai district expanding facial recognition cameras to capture 25.9 million faces daily for data analysis.

Cheers,

Sam

PS: I’m trying this new format where I articulate news around the same topic under one story. Less links, more opinions. Share your feedback below.

Enjoyed this newsletter? Forward it to a friend and have them signup here.

Until next Thursday 🎉