Ready for the weekend? Let’s get into some big ideas.
Wrapping up the Apple stuff this week, Interconnects looked at Apple intelligence and broke down the strategy. By focusing on personalization, performance, and size, Apple’s core models demonstrate strong performance compared to other language models. The company’s alignment strategies include fine-tuning techniques and the use of novel algorithms for reinforcement learning. Additionally, Apple has developed adapters and on-device strategies to optimize memory usage and latency speeds. This comprehensive approach positions Apple as a significant player in the AI race and emphasizes the transformative potential of AI for billions of users.
Ben Thompson wrote about how Apple Intelligence is Right on Time. He highlights the importance of smartphones as nearly perfect devices for the tasks we ask them to do and the app ecosystem as a natural state of affairs. He then explores the infrastructure concerns of AI and the competitive threats Apple faces, particularly from Google. He offers that Apple’s deliberate approach to AI aligns with its brand promise of privacy and security.
Information Week with “Should IT Treat Users like Clients?” Moving beyond my flippant “yes,” Mary Shacklett argues that IT departments should adopt a more client-oriented approach towards internal users to improve customer service and build long-lasting relationships. This includes demonstrating empathy, following up with users after projects, and anticipating their needs. While many IT departments have made progress in improving internal customer service, there is still work to be done in areas such as relationship building and interpersonal skills. CIOs are crucial in advocating for a client-focused IT and should lead by example.
And I’ll end with a great use case. The Wall Street Journal covers Google expanding its Green Light product. It uses data from new cars and navigation apps to optimize traffic light timing, creating “green waves” of consecutive green lights. The system has already shown a 30% reduction in stop-and-go traffic at intersections in cities where it is deployed. This approach could significantly reduce time spent idling at traffic lights and improve traffic flow. It also eliminates the need for expensive and time-consuming studies to change traffic signal timing.
Why do we care?
Big thoughts that I want to stand on their own, and you enjoy over the weekend.