AI

How I Feel About Generative AI/LLM Tools

In short, conflicted but engaged. 

It bothers and frustrates me that many (if not most) models have been trained on materials in a manner that undermines consent, licensing, intent, attribution, values, and more. I’m not convinced anything meaningful will be done to address these issues, especially those rooted in the past.

I’d feel more at ease with these tools if companies were held accountable for how training data is sourced and if creators were compensated when their work contributes to a model’s capabilities or is used to generate commercial value.

At the moment this post feels very on point:

The underlying purpose of AI is to allow wealth to access skill while removing from skill the ability to access wealth. - Joseph Browning

That said, I believe that there are also incredible benefits to these tools. It excites me to see advances in many fields which do seem to be for the betterment of all.

How Do I Use AI?

It is also clear that these tools are akin to the industrial revolutions; failing to embrace the AI revolution means getting left behind. So I am constantly working to embrace AI tooling in ways that make the most sense to me and in ways that I feel most comfortable with.

I use a combination of ChatGPT and Claude.ai for research, the way I used to use search engines before they started to return mostly spam.

I use AI to help troubleshoot my code and to generate code snippets for me, and I have used it to help me better understand sample code, documentation, and code I've written a few years ago 😅

I avoid using AI for image generation to create "production" output. I have used image (and video) generation to help me convey an idea to the creatives with whom I am working.

I avoid using AI to write for me. I have used it to ask me questions about what I am writing, to provide me with alternative words and ways of saying something, and to create summaries of content I have written.

Inspired by /ai from Slash pages