“The web is written by AI for AI”

From an opinion piece by Steven J. Vaughan-Nichols

Bots now account for roughly 57-58 percent of HTTP requests for HTML content, compared with about 42-43 percent from humans. Meanwhile, Imperva’s Bad Bot Report based on 2025 data put bots at about 53 percent of measured web traffic for the second year in a row, with humans at 47 percent. […] According to Pangram, an AI detection company, on websites such as LinkedIn, Medium, Twitter, and Reddit, “about one in four long-form items were fully AI-generated.”

The author uses Perplexity as his search engine not because it’s more accurate than other AI LLMs; “it’s that it shows me its sources. I can see if what it just turned up is the real thing or just BS. Guess what? It’s often crap.”

I’m not certain the other LLMs do not show sources. I think they tend to hedge on that score. But if half of the web is AI slop, I’m not sure it makes much difference.

“AI, please read my whole blog”

The work that Perplexity has done with my 25 years of blog posts makes me wonder if there have been similar efforts by other bloggers.

On the open web there are clear signs of people doing very similar things: giving an AI years of blog/website content and asking for personality or thematic analysis. Some examples include:

  • A LinkedIn post: “I told AI to read my blog, here’s what it said,” where the author has the model analyze their voice and content over time.
  • A Japanese blogger on note.com who “had an AI read my blog and perform a self‑analysis.”
  • Ed Harrison: “I Asked an AI to Read My Old Blog So You Don’t Have To,” describing handing 15 years of writing to an AI and reporting what came back.
  • A data‑oriented piece where someone scraped their site, fed it to an LLM via API, and did summarization/topic‑modelling of all posts.
  • Technical posts about summarizing old, personal blogs using local Llama models.

Those are all close cousins of what you’ve been doing with Claude, Perplexity, Gemini and your ~25‑year archive.

Your project’s particular flavor

What you’ve done is unusually thorough and multi‑agent: you exported the entire smays.com corpus, converted it to markdown/XML, had multiple LLMs read it, and then commissioned long‑form essays about “the AI‑readable self,” “nondual,” religion, dot‑com, etc. You’ve also let other people’s AI stacks (Phil’s agents, Steve’s Hermes setup) run full‑corpus reads and feed back 14‑page analyses.

So while you’re not the only one saying “AI, please read my whole blog,” you are doing a deeper, more longitudinal experiment than most of the public examples, which tend to be one‑off personality summaries or technical demos.

Religion in the Age of Stories

I prompted Perplexity to write an essay based on the seventy-five posts in the Religion category on my blog. (First one was in 2004.) Rather than have this written in my own voice with first-person interjections, I told Perplexity to present it more as a third-person commentary on “this blogger’s” long-running engagement with religion.

Across more than seventy posts in the “Religion” category, Steve Mays treats religion less as a pipeline to cosmic truth and more as an old, resilient operating system for human behavior and meaning. It sits alongside money, nations, ideologies, and corporations as one more way humans coordinate action through shared stories. In this framing, gods and doctrines are not unique intrusions from beyond the universe; they are examples of the broader human habit of inventing fictions powerful enough that people will kill and die for them. Continue reading

Looking back at the dot-com bubble

The dot-com bubble (and subsequent crash) happened over a two-and-a-half-year period between  2000 and 2002 (the year I started the blog). I was reminded of this period while watching a video comparing the current AI frenzy to Tulip Mania back in the 17th century.

I couldn’t remember if my early blog posts touched on the dot-com period so I prompted Perplexity to take a look and it came back with a “looking back at the dot‑com bubble” post (see below). This is just the latest example of how my resolve to not let AI bots write my blog posts is slowly eroding. After one read I knew I could not improve on this summary.


(Perplexity) If my archive is any indication, I didn’t spend the bubble years themselves writing hot takes about it. The more interesting stuff came later, in the early 2000s, when it was obvious the web had outlived its first speculative fever dream.

One of the clearest snapshots is a June 11, 2003 post I titled “Markets as conversations,” where I pulled a quote (via Denise Howell / David Weinberger) that still reads like a verdict on the whole episode: “The Bubble was never what the Internet was about. The Web is not primarily a commercial space, not even primarily an information space… The bubble went away, but the Web absolutely didn’t. The Web remains interesting and important.” I didn’t add much commentary because I didn’t need to; posting it was endorsement enough.

By April 2005, I was pointing readers to a BusinessWeek cover story, “Blogs Will Change Your Business,” and highlighting this contrast: “The dot‑com era was powered by companies — complete with programmers, marketing budgets, Aeron chairs, and burn rates. The masses of bloggers, by contrast, are normal folks with computers… and — that’s right — no bubble.” I liked that line so much I reproduced the whole setup, emphasizing that we’d moved from company‑driven “New Economy” hype to what they called “media of the masses.”

Around those posts, there’s a steady drip of “Early Net Culture” items that implicitly answer the question, “What actually survived the crash?” In June 2003 I noted Google News beating BBC and MSNBC for a Webby and mused about how AP’s real asset had been its “connectedness” — something the open web was now replicating for everyone. The question “Can we agree that has changed forever?” wasn’t framed as bubble commentary, but it’s about what didn’t revert to normal once the NASDAQ fell.

A couple of years later I was citing ad forecasts that put Google and Yahoo’s combined online ad revenue on par with the prime‑time haul of ABC, CBS and NBC, calling it a “watershed moment” in the evolution of the internet as an advertising medium. If the bubble was about speculative eyeballs, this was about actual budgets moving. Same network executives, different direction for the money.

So, reading my own trail backwards, my “dot‑com bubble” coverage is less about the boom and crash and more about the quiet, durable things that came after: the idea that the web was never just a stock chart; that normal people with cheap tools and “no burn rate” might be more important than the Aeron‑chair start‑ups; and that news, advertising, and culture were re‑routing themselves through a network that turned out not to be a fad.

The Fourth Turning Is Here

Recently watched an interview with Neil Howe, the author of The Fourth Turning Is Here: What the Seasons of History Tell Us about How and When This Crisis Will End. The interviewer begins by quoting Howe’s prediction that “In the coming year, the country will undergo a crises that will fundamentally reset society.”

I asked Perplexity for a synopsis of the book:

Neil Howe’s The Fourth Turning Is Here argues that the United States has already entered a long, cyclical “Crisis” phase that will upend existing institutions and values and culminate in a new civic order sometime in the early 2030s. It extends his earlier generational theory to explain how today’s Baby Boomers, Gen X, Millennials, and Gen Z each play distinct roles in driving and resolving this upheaval.

The book updates the Strauss–Howe generational theory, which sees history moving in 80–100 year cycles composed of four “turnings” (High, Awakening, Unraveling, Crisis), analogous to seasons. Howe contends that the current Fourth Turning (Crisis) began around the 2008 financial crash and will likely end in the early 2030s with a decisive reshaping of American political, economic, and social life.

And a counter view:

Many historians and social scientists see Howe’s framework as pattern‑spotting on noisy data rather than a testable theory, arguing that the “turnings” are cherry‑picked after the fact and stretched to fit events. Critics also note that the generational archetypes are so broad and fuzzy that they can explain almost anything, which makes the predictions more like narrative prophecy than falsifiable social science.

The interviewer asked Howe for if he was at all optimistic and he said the current crises is a good thing because the alternative is more of what we have now. I’m down with that.

What I Mean (and Don’t Mean) When I Say “Nondual”

Every word of this post —after this little grey box— was written by Perplexity. After finally getting around to uploading all 6,500 blog posts on smays.com (spanning almost 25 years) I was curious what an AI might come up with, being the only entity beside me to have read (analyzed? processed?) every post. We started with nonduality, a subject that has interested me for years. The essay (?) below drew on hundreds, maybe thousands, of posts touching on awareness, consciousness, free will, death, immortality, memory, reality, self, Tao, thoughts, time, and zen. I’ve included additional thoughts in the comments at the bottom of this post.

I’ve used the word “nondual” on the blog often enough that it probably deserves parole into plain English.

I don’t mean crystals, chakras, or a belief system you have to sign up for. I also don’t mean “nothing matters, it’s all an illusion, pass the Doritos.” If anything, nonduality has made ordinary things—espresso, a train whistle, the dog snoring—feel more vivid, not less.

What I’m trying to point at is simpler and slipperier. There’s awareness, there’s experience showing up in that awareness, and the boundary between “me in here” and “world out there” is not as solid as I was trained to believe. Most of the time, that’s all I mean.

The rest is stories and examples. Continue reading

Dr. John Middleton

It was just a few months after I got my Land Rover that I met John Middleton. Someone told me he had a collection of Rovers so I tracked him down and he invited me up to Columbia. In the ensuing years John was an invaluable resource whenever my Land Rover had a problem. Even joined the crew to help me take the top off and on.

Today I took the hoop kit (I used for just one summer) up to John’s farm and to see some of his recent acquisitions. During my brief visit I mentioned the MU vet school where John is a professor. He said something about his duties changing since we last talked so when I got home I asked Perplexity what it could find.


Continue reading

Perplexity is a better writer than I

A couple of years ago I pulled together some notes and files about Learfield’s early adventures with the Internet. (I’ve posted on this topic more than once.) Yesterday it came up in a conversation with Perplexity which asked I would like to see a “narrative chapter” of that history. I was pretty sure Perplexity would produce something more readable so I said “go ahead.” 

It boiled my fourteen page ramble down to three pages while creating a story that –to my mind– accurately captured the vibe of those frantic days. [Links above to PDFs]

I’ve resisted the temptation to let an AI write or rewrite these blog posts and will continue to do so. But I wonder if it’s much different than an author submitting a manuscript to an editor who cleans it up and generally improves it.

The more your AI knows about you…

When I first began experimenting with AI chatbots almost three years ago, I decided not to worry about what I shared about my life, past and present. I’d been blogging for more than 20 years so it was all “out there” already. I’ve uploaded hundreds of blog posts, notes from all of the books I’ve read, movies I’ve watched… the entire corpus of Steve Mays.

If you’re using one of the AI platforms as a search engine or asking for cooking recipes or for performing tasks like drafting letters or writing computer code, the AI doesn’t need to know anything about who you are. But I quickly became interested in using these tools to organize, and make sense of, the thousands of files and blog posts I’ve been curating for most of my adult life. To do that, the AI needs context. Who is this person? What does he think? What does he feel?

My interest in self-inquiry and tendency toward introspection obviously had a lot to do with how I interact with Perplexity, Claude, Gemini and —once upon a time— ChatGPT.

Today I added some work history to the “corpus.” Continue reading

The Ten Commandments for Al-Augmented Creativity

  1. Thou shalt not outsource thy soul. Use Al to enhance, not replace. If it writes for you, it thinks for you. The danger is the Al effect, one step beyond the Google effect, where your brain no longer expends the energy needed for cognition and creativity-and these skills atrophy.
  2. Thou shalt suffer for flow. The struggle of creation is signal. Flow requires challenge. Effort drives focus, and focus drives us into the zone. If we let the machines remove all the friction, we stop growing, stop flowing, and stop becoming.
  3. Thou shalt keep sacred the first draft. Let AI critique your work but never conjure the original spark. Creativity is associative: We draw connections from vast, messy networks of memory and intuition. If Al drives the process, you short-circuit your retrieval system and risk losing access to the deeper and weirder part of your imagination.
  4. Thou shalt use Al as a challenger, not a crutch. A good tool makes you think harder. Forcing the brain to work triggers the neurochemistry required to activate memory. If it’s too easy, you’re not learning. AI has to elevate your edge, not erase it.
  5. Thou shalt preserve the joy of creation. If Al removes the satisfaction of making, you’ve automated too much. That satisfaction feeds motivation, meaning, and purpose. If Al erases it, life feels empty.
  6. Thou shalt set boundaries for digital influence. Keep parts of your life untouched by algorithms so your thoughts remain your own.
  7. Thou shalt not mistake efficiency for depth. Faster is not always better. AI should deepen thought, not just speed up output.
  8. Thou shalt train thy mind alongside thy machine.The sharper the tool, the sharper the brain must become. Let AI steer you toward the right research papers, but if you don’t read them yourself, then you’ve stunted cognition and creativity, abandoned foundational motivators like curiosity and mastery, and limited your shot at flow. AI needs to stretch skills forward, not siphon off drive.
  9. Thou shalt honor serendipity and chaos. Don’t let Al’s predictive patterns strip your work of surprise, randomness, and the unexpected. Novelty is the seed kernel of creativity. We need the unexpected to drive the brain into new directions; it’s the basis of all innovation.
  10. Thou shalt remain the master, not the servant. AI is a choice. We can choose to unplug it whenever we want.

From We Are As Gods by Peter H. Diamandis and Steven Kotler.

I ran these by Perplexity and it suggested some ways to integrate these into our interactions. (PDF) As expected, Claude had a somewhat different take.