My First Impressions of Siri AI on the Mac: A Mixed Bag
I've always been a bit skeptical of Siri, the virtual assistant on my iPhone. I've never seen the need for it on my phone, and I've long since disabled it on my Mac. But with the new Siri AI arriving on macOS 27 Golden Gate, I decided to give it a try. After just 24 hours of testing, my initial impressions are mixed.
The Good:
Indexing and Search: Siri AI is surprisingly effective at indexing and searching files and folders on my Mac. It can quickly find specific photos, documents, and even specific types of files, which is a huge improvement over the past. This is especially useful when I need to locate something quickly without having to manually search through my files.
Data Analysis: One of the most impressive features is its ability to analyze data. I tested it by asking Siri to calculate the average scores from my benchmarking screenshots. It successfully distinguished between single-core and multicore CPU scores, GPU scores, and arranged the results in easy-to-read tables. While it sometimes struggled with mixed test types, it's a significant step forward for Siri's data processing capabilities.
Visual Intelligence: Siri's Visual Intelligence feature is intriguing. I asked it to evaluate benchmark results on a spreadsheet, and while it had limitations, it can identify specific data points on the screen. This could be useful for quickly accessing information within applications.
The Not-So-Good:
Limited Automation: Siri AI still struggles with automation tasks. It can't launch apps or take actions within them, which is a significant limitation for my work. I tried to automate benchmarking, but the Shortcuts created by Apple Intelligence were incomplete, failing to run the actual tests.
Ecosystem Dependence: Siri's usefulness is heavily dependent on Apple's ecosystem. It excels within Apple's apps and files but falls short when it comes to non-Apple applications. For example, it couldn't access photos stored in Google Photos or Lightroom Classic, even though they are stored locally on my Mac. This makes it less versatile for users with diverse app preferences.
Accuracy and Context Awareness: Accuracy is a concern. Siri sometimes pulls the wrong data or misses important information. It also struggles with context, like when it recommended setting exposure to its current value when asked to match a specific photo style. This lack of precision can be frustrating.
Looking Ahead:
Siri AI is clearly a work in progress, and I expect significant improvements before its release later this year. The experience will likely be more refined on iPhones, where data is more centralized within Apple's ecosystem. However, for now, it's a baby step towards a more capable virtual assistant on the Mac.
As an early adopter, I'm intrigued by Siri's potential. I'm eager to see how it evolves and whether it can truly become a valuable tool for my everyday work. For now, it's a mixed bag, but I remain optimistic about its future.
