AI in brief for the Mac Admin Purist

Or, how I learned to stop worrying and love my ability to act for myself

Genesis

Near the end of 2022 when ChatGPT was becoming all the rage, I purposely started using AI tools like Phind (at the recommendation of one of my sons) since it didn’t require a login and I was hesitant to let AI know too much about me.

Today, the following is one of my favorite initial AI prompts:

Who am I and what do you know about me?

I’ve trained my personal AI account on publicly available information about myself, while additionally training my work’s enterprise AI account on my team’s dynamics. I strongly recommend not inverting the former with the latter.

AI to AGI to ASI

Using the headline immediately above as an AI prompt, here’s Grok 3’s (current) response:

The progression from AI to AGI to ASI represents an evolution in artificial intelligence capabilities, each stage marking significant leaps in functionality and autonomy. Here’s a concise breakdown:

• AI (Artificial Intelligence): Narrow AI, designed for specific tasks. Examples include image recognition systems, recommendation algorithms, or voice assistants like Siri. These systems excel in defined domains but lack general reasoning or adaptability outside their programming. They operate on predefined rules or trained models, processing data to produce outputs without understanding context or generalizing to new tasks.
• AGI (Artificial General Intelligence): AGI is a hypothetical system capable of human-level intelligence across any intellectual task. It would match or surpass a human’s ability to reason, learn, plan, and adapt to novel situations without task-specific programming. AGI could, for instance, solve math problems, write poetry, or strategize in unfamiliar environments using general reasoning. While theoretical frameworks exist, AGI remains unrealized, with current systems like large language models being sophisticated narrow AI, not true AGI.
• ASI (Artificial Superintelligence): ASI surpasses human intelligence in all domains — cognitive, creative, and emotional. It would autonomously improve itself, solve problems beyond human comprehension, and potentially operate at scales or speeds unimaginable to us. Think of ASI as an entity that could redesign entire systems of knowledge or technology in ways humans couldn’t predict. This stage is speculative, raising ethical and existential questions about control and alignment with human values.

Or, as one of our internal AI Product Mangers observed:

AI likely won’t replace your job, but someone who knows AI better than you will..

Make The Putt

The following golf-related analogy sums up my current take on AI:

While AI may get you from the tee to the green, you still have to make the putt.

As I’ve embraced GitHub Copilot with Visual Studio Code, being a self-proclaimed purist, I’m as frequently impressed as I am frustrated (and I’ve certainly learned the location of the Escape key).

Even after training your AI of choice on the fact that you’re a macOS system engineer — and not using Zsh on Linux — you’ll still receive code suggestions which simply don’t work (and no amount of on-the-fly AI training seems to improve future responses).

Yips

Here’s a simple example of auto-complete where there’s purposely no nearby relevant context; notice how macOS 27 was twice suggested:

Hole-in-one

However, using /explain via the inline chat is invaluable for quickly understanding code you’re looking at for the first time.

Always remember:

The music is not in the violin.
— Kahlil Gibran

Resources

• Things as They Really Are 2.0
• AI in a Gospel Context
• Elon Musk’s New “GROK 4” AI System is a Massive Wake-Up Call