I got a Claude Pro Subscription

You read that correctly, I got a personal Claude Pro subscription.

Back in 2022, when ChatGPT was launched, subscriptions were cheap. Nonetheless, I held back from getting a personal subscription because most of my GenAI use was work-related, and work offered a secure subscription.

Until very recently, I had little use for GenAI, not other AI techniques. I write all my content, and use Grammarly to check my grammar and typos. I have a subscription for Grammarly Premium and intentionally toggled off the 'Train AI' option. Other than that, I was comfortable using the public AI tools because nothing I used them for was private or required extensive, complicated thinking. In fact, I can tell you I used it to provide intentional Google search keywords, or (in the name of honesty) to dump what I want to look for and find the right word or words for it. All that changed when a particular Python visualisation package decided to deprecate a much-loved studio and enforced anyone to move to an AI-centric solution. If you browse my website, most of my charts aren't rendering, and that means I need another solution. The mentioned Python package + studio were so easy to use, navigate and embed that, of course, finding an alternative was complicated.

BUT I did! I found an alternative and am, slowly but surely, recreating the charts. I also decided to rebrand the website and learn SEO, which is complicated.

It meant that I have a personal use case for GenAI.

Whilst none of the charts I published contained any private data, I now have an extensive use case for a personal GenAI subscription. I know the fundamentals, but do I need to learn new ways of generating the same thing?

I decided to use GenAI to bridge the gap, but I tell you this. Whilst I'm not doing a course, I still have to refer to documentation and learn. The GenAI output ain't always what you want, but with a bit of reading and glancing at the code, I could make the necessary changes for a more intentional output.

Why Claude?

Why not? But in all seriousness, my main use case for GenAI is coding, which, in my opinion, Claude has proven to be unparalleled so far. Moreover, I follow Anthropic's research, which aims to understand how LLMs calculate their outputs. I started the practice to move away from marketing language when talking about GenAI - it does not think, it performs logic.

Once that was decided, I did some research on Claude Pro privacy and their policy on using my input to train the LLM. Google's search said that the training is toggled on by default and that one needs to intentionally opt out. Maybe it's because I'm in the EU, but it was off by default, and hopefully it won't change without my consent.

The question some of you might ask is: "Will you use it to generate article content?" My answer remains:

"Absolutely not".

My mind is my own. I have a lot of ideas and opinions I want to share in my own voice. It may not be perfect, it may not be AI-friendly or anti-AI, but it is my own.