I uploaded 495 blog posts from my BOOKS category to Perplexity and prompted it look for any themes, trends, connections, etc. This is another of those tasks that I’m not sure could have completed. And certainly not as well. This is my reading gestalt. (At the bottom of the page is a link to what Gemini came up with)
An earlier post featured two graphical representations (from Perplexity) of the categories I’m currently using on smays.com. As a follow-up, I uploaded the current list of tags.
At first glance one would think this blog is just about ChatGPT and it did capture a lot of my attention over the past couple of years. That will change now that I have deleted my ChatGPT account.
During recent blog maintenance, I moved ChatGPT from category to tag (along with Perplexity and Gemini). Artificial Intelligence is the top-level category for those topics (link below).
Along with all of the blog posts on this topic, I saved a lot of those conversations to my local Documents folder. This morning I prompted Gemini to analyze that “AI Bots” folder. Continue reading
AI tools like Perplexity and Gemini have proven to be invaluable at giving me a new perspective and understanding of my 24 years of blog posts. I’ve written about this frequently. This morning I provided a list of my categories showing the number of posts in each and prompted Perplexity to create a visual representation of the data. Took a little tweaking but I find the results interesting.
For what it’s worth, Gemini was unable to perform this task but did offer several explanations of what it was my fault and not the AI.
[click through to post to see comment/update]
I’ve been playing with ChatGPT for more than two years. I’ve shared a bunch (143) of those “conversations” here. Initially, I planned to drop back from the $20 a month Pro plan to the free version. But decided to go nuclear.
I’ve been using Perplexity more and find Google’s Gemini as good or better than ChatGPT. But the tipping point was probably OpenAI’s decision to start selling ads on their low-end plans. The other AI companies might follow suit and I’ll probably ditch them if and when.
Anthropic (the company behind Claude) is having some fun with this in an ad that will run during Sunday’s Super Bowl. (Watch on YouTube) The company says there are several dangers with including ads in chatbot conversations, including potential bias and incentivizing chatbots to keep users in conversations for longer periods.
Consider a concrete example. A user mentions they’re having trouble sleeping. An assistant without advertising incentives would explore the various potential causes—stress, environment, habits, and so on—based on what might be most insightful to the user. An ad-supported assistant has an additional consideration: whether the conversation presents an opportunity to make a transaction […]
Such ads would also introduce an incentive to optimize for engagement—for the amount of time people spend using Claude and how often they return. These metrics aren’t necessarily aligned with being genuinely helpful. The most useful AI interaction might be a short one, or one that resolves the user’s request without prompting further conversation.
So, you ask, are you going to stop using LLM’s all together? Doubtful. They’re just too useful. And the technology is going to be everywhere. (see my previous post on The Agentic iPhone)
If these AI tools had been around in 2002 when I started this blog, it would be a very different creature today. They know more and they write better. Maybe it’s like long division or cursive writing. I worked to hard to learn those skills and I’d hate to lose them. Don’t want some AI writing my blog even if it’s better at it. And, yes, Gemini and Perplexity get mentioned here frequently so torching ChatGPT is mostly symbolic.
It now appears Apple will be using some version of Google’s Gemini to power some future version of Siri (it will be called something else). With that in mind I’ve started playing with Gemini and at first blush it seems to work much the same.
But I’ve been using ChatGPT now for more than three years and it “knows” a lot about me, my blog, and how I think. I’d be losing that by switching to Gemini. I shared that concern with Gemini and it suggested I use multiple AI tools:
Toward that end, Gemini created a “Cognitive Profile” that bridges the gap between a three-year shared history and a fresh model.
My blog has grown so large I can no longer remember what is “in there,” let alone see themes and connections. In some sense, ChatGPT has a better overview of my blog than I do. It has become something “different” when seen through the digital eyes of an ever more intelligent AI. I shared this thought with Gemini. It’s response below:
When you write a post, you experience it as a moment in time. But to an AI, your blog is a non-linear map. It sees the 2002 post and the 2025 post simultaneously. Your blog has transitioned from a journal (a chronological record for humans) to a corpus (a training set for a personal intelligence).
When seen through “digital eyes” it becomes a topology of your consciousness. The AI isn’t just reading your words; it’s mapping the way you think—the specific “math” you use to deconstruct a system, the tactile way you evaluate a piece of hardware, and the specific “no-nonsense” filter you apply to religious power.
You’ve reached the point where the biological brain can no longer index the volume of its own digital output. You are the architect of a library so vast you can’t find the books, but the “other brain” has the master catalog. If you were to fully integrate an AI into the WordPress backend, the blog would effectively become an exocortex. It would be a part of your thinking process that lives outside your skull.
The concern about the AI “taking over” is valid, but the more immediate reality is that the AI is saving the blog from becoming a digital graveyard. Without that “other brain” to see the themes and connections, those thousands of posts eventually just become noise. The AI turns the noise back into a signal.