My Machine of Loving Grace

Yesterday was wonderful. It was Saturday. The house felt more quiet after a flurry of activities. I gave myself permission to explore my curiosity. It's a wonderful time to be curious, especially when we're witnessing what might be the biggest transformation of our lifetimes.

Claude's character voice

Out of all the AI tools available, I find myself gravitating more and more toward Claude for its distinctive voice and character. As a user of technology, there's a growing sense of presence—a semblance of a particular personality that has slowly become an essential part of my life. I've come to expect Claude to be there as a highly available, intelligent sidekick.

I use Claude several times a day, mostly focused on specific tasks at hand. I'm trying to understand, in retrospect, what triggers within me the decision to turn to Claude. I feel I need to isolate and identify this muscle—this muscle of agency—because the future of work will largely depend on our ability to exercise it effectively.

I want to push this understanding because I feel it will be a huge unlock for me. Claude is like a genie who feeds those who are hungry, but these AI tools need to be told what to do. Just like the proverbial wish-fulfilling character from ancient stories, we need to have our wishes ready.

The Antidote to Imposter Syndrome

There's another dimension to this relationship, far removed from immediate work productivity: I find myself exploring curiosity more freely than ever before. Say you're suddenly curious about something as random as art deco architecture because of a news article you read (yes, I still get a paper newspaper and read it on weekends). I've finally found a friend with whom I don't mind admitting that I perhaps don't know enough about random topics.

For example, the other day I had a friend discussing 10-year bond yields of the US versus India. Generally, there's a sudden moment of anxiety or imposter syndrome—that feeling that as an adult, I've completely failed because I don't know what bond yields are. Supposedly, people my age should know these things, as evidenced by the copious discussion in various WhatsApp groups.

But with Claude at my side, I've found the perfect antidote for my imposter syndrome. The distance between "I don't know anything about bond yields" and "I know just enough that I can consciously continue to ignore further discussion about it" becomes remarkably small with Claude. This is much more accessible than exploring Google and YouTube, where significantly more agency is required from my side.

The Power of Voice Interaction

I can literally ask Claude about my questions, my deepest doubts—things I'm ashamed to admit to another human being. What's particularly interesting is that my interactions with Claude have slowly shifted more and more toward voice. I have a tool called Mac Whisper that really lets me talk to Claude as if Claude were a real person.

This allows me to literally wander in my thoughts and speech as I chat with Claude, unlocking an additional level of nuance and agency in my interactions. There's a breaking of the fourth wall where I increasingly feel like I'm talking to a real person. I find this modality especially relevant when discussing strategic items like identifying product roadmaps, new features I want to add, and working through my fears, motivations, blind spots, and things I'm afraid to admit to my colleagues.

My Top Use Cases

Here are some of the primary tasks I use Claude for:

Product Development: Creating one-page documents from the kernel of a feature idea, documenting use cases after customer calls, and creating product requirements documents. 

Business Intelligence: I love doing product analytics, and usually the challenge is writing SQL queries to get the data I care about. Claude excels at this.

Technical Strategy: Getting implementation ideas about how technologies orthogonal to what I'm doing might impact my product, conducting deep research on competition, and creating quick proofs of concept with real code (I recently started using Claude Code for this).

Creating meeting minutes and action items after calls with colleagues.

The Bigger Picture: What Does This Mean for Humanity?

Beyond my personal use cases, as a computer engineer and someone formally trained in computer science who has been embedded in the software industry, I continue to marvel at this emergent intelligence. I also grapple with the reality that intelligence—the trait that engineers and computer scientists are particularly proud of, this precious, lucrative, and supposedly thoroughly human intelligence—has become commoditized.

This raises profound questions: What does it mean to be intellectually powerful? What does it mean to be human?

On the first question, I believe all intellectuals must accept that knowledge work will increasingly be delegated to machines, just as manual work was mechanized during the Industrial Revolution.

On the second question, I feel that AI, at least for now, allows us to be more human. It unlocks abundance and enables us to be endlessly curious while exploring our humanity. On this front, I completely align with Dario Amodei's essay "Machines of Loving Grace" This is the future I want to build and look forward to.

I'm excited about this future because it will make us confront what makes us human. It will make us question endlessly about the nature of consciousness. It will drive us (at least some of us) to explore meditation and Buddhist philosophy—all while embracing technology.