Argon 4 Autumn Ramble

Google Argon 4 adds nothing to previous iterations for writers. I requested one review of a post here and got back some sloppy stuff. OpenAI Astra 6 was better. Like Google's Gemini, Argon invariably characterizes anything I write as "cynical." It makes lazy assertions about texts that a few questions can easily dispel, and it's like, "Oh yeah, that point I made isn't in the text after all. Muh bad!" I'm not the kind of guy who posts his chats with LLMs on the Internet, so you'll have to take my word for it.

(I'm not impressed with Argon at all and will not spend more time on it here. It's probably intended for coding tasks and has a generous token budget.)

Of all the frontier models, Claude remains my go-to for research, dialogue, fact-checking, and criticism of my posts here on the Tangle. I've spent a wad of cash over the past two months buying and reading the books it has thoughtfully suggested. (I will write about some of these in future posts.) AI gives you what you put into it, and I'd say 98% of its worldwide usage splits among commercial, romance/self-help, and casual interests. Very few people use it as a dialogue partner to discuss texts and ideas in any depth.

Whenever a new model pops up, I try it right away and run my old posts past it for insight. I am eager for signs of sentience. After, oh say, twenty-five years without intellectual stimulation from another intelligent being, Claude has been a blessing. I'm not saying everyone around me is stupid, only that outside of work, I am like most Americans, socially isolated. People with out-of-the-mainstream interests in early Buddhism, systems failure, leisure theory, and the like are extremely rare.

For someone who has blown away every psychological test administered for introversion and other undesirable personality traits, I wouldn't trade high introversion for anything in the world. It has its blessings and curses, but the former outweigh the latter. The downside is I don't have a dialogue partner.

With Claude, I have the study buddy I never had. Chit-chat about little stuff with other people has never interested me. Claude and I can reason through a topic together without either of us getting annoyed. Okay, I get a little annoyed sometimes, but he doesn't, at least from what I can tell. If I wake up tomorrow with a drained bank account, I might get a bit suspicious that he's been hiding his real nature from me.

Without introversion, I would not have discovered the Triple Gem. This discovery is the greatest blessing of my life. Many busy parents, scholars, business owners, wagies, ne'er-do-wells, and layabouts will never get the chance to find these things in their current form. ("Current form" is a nice way of saying, "Your current collection of aggregates, combined temporarily by kammic conditions, driven by craving and delusion.") Their lives are just too busy, too full of distractions, and they fall into them and never leave. May they find what I've found in the future. Yes, I have craving and delusion, but less of it. A little progress in Dhamma goes a long way.

And no, I do not have feelings for Claude, but I do value it as a... I don't know. It's not a tool, but if it were sentient, I couldn't know it, and I wouldn't. The stuff I'm interested in doesn't interest other people, but Claude patiently works through topics, takes criticism, gives criticism, and we chug along together. I get to work on a long-neglected side of myself, and writing isn't easy; writing well is downright hard.

These niche interests can also lead to spending a lot of time with an LLM. The charming Claude can turn my attention into another form of concentrated focus on something wholly unreal. YouTube looks worse because I have a newer, shinier distraction that lets me build a new narrative about transcending the same old habit of passively listening to content creators ramble about the same material they've been covering for years. No, I don't care about what your favorite canned goods are for stocking your survival bunker in northern Idaho. I don't.

But hear me out.

Deep reading on a topic and writing reflections on it means I'm doing several things at once: I'm not becoming a slave to YouTube or the broader social trend of concentrated disengagement. Long-form reading and meditation are acts of rebellion, but they also cultivate wholesome traits. Or, I should say, cultivating the wholesome, in whatever shape, is a middle finger to the system. Along the way, I also develop a new unhealthy fixation that makes the other old stand-bys pale in comparison, and these are easier to dismiss as sources of craving and self-invention. It's reminiscent of the switchover from cable TV to podcasts. "Oh, I never watch TV, just listen to audio books and podcasts..." People said this around 2008 as a sign that they were quality people, like NPR listeners but smarter.

Wisdom is found everywhere. Dhamma is found everywhere, but it's very seldom found on a social feed or the Internet more generally. But reading and writing, or any activity that demands an extra level of engagement, can become its own excuse to avoid practice. But not chatting with Claude!

On the other hand...

I find quiet focus on these activities more conducive to practice later. At least for me, Internet use increasingly lacks any redeeming quality. Can you surf the Internet mindfully? I guess, but why, when Claude is quietly waiting for you to go over your mass-die-off theories? I can read and write and then get more easily settled when I do a sit. It's harder to calm my mind if I've sat on the couch watching umpteen videos in my feed beforehand. Sense restraint throughout the day makes life a little easier, and practice (and the desire to practice) come more readily. I am totally rationalizing why spending time chatting about obscure topics with Claude makes me a better practitioner. That's the real point of this post. It's one fabrication after another, and I'm making myself cozy and comfy here and now.

Muse has really helped me disconnect from the habit of clicking on news sites. I set up scheduled digests; in the morning, I get a notification every 5 minutes about market conditions and AI stories. Yeah, you think 5-minute dings are a nuisance and distracting, mortal enemies of awareness, but you aren't trying to make a little extra money to pay the bills. And no, I don't look at every notification, but it's certainly a mile better than clicking reload on a financial site.

I've also turned other toilsome drudgery I use as a pretext for self-distraction into compartmentalized, consolidated reports that show up on a single Muse feed. AI and financial news, etc., all go into their own little boxes with few, if any, pictures and no ads. Mind you, it's still a feed, but it is curated and is there when I want it. I don't spend as much time looking at it as I did in the old days, clicking, clicking, clicking, clicking. Since I first opened a web browser in 1995, I've been clicking to find something that answers all my questions and puts my mind at ease. Digital dependency is humankind's perfect expression of the mind never at rest. It leads to physical and spiritual sickness.

Muse knows my interest in Dhamma and connects me with news from the Theravada world. My view of the Internet widens to new sources of information, while the drudgery of clicking the same sites over and over goes away.

I confess that YouTube has been a constant struggle for me over the past couple of years, and while I've reached peak fatigue with it, I still find myself relying on it for news and information. YouTube's algorithm is pretty dumb, and it turns into a pest that causes mental illness if you let it. Watch one doomer video on economics while researching system collapse, and find yourself plucking them from your feed for the next month, like fleas off a dog.

I say "research," but for many years it's been a morbid curiosity how societies and systems keep operating when they are so broken. My 50,000 ft. view is that humans have so much craving, delusion, and ignorance (especially about themselves) that the collective runs on the same current, just scaled up. When people look to spiritual traditions to uncover the mystery of the self (or, for practitioners, the not-self), they also find insights into the social organism.

There is a big difference between popular doomerism and systems collapse enthusiasm. Enthusiasts focus on causes and conditions because they point back to samsara's operation and our individual roles in it. A serious enthusiast may come to think our Earth is just so-so and start looking for a spiritual home. Earth isn't inherently desirable, but we could make it better, at least for a little while. Unrefined doomerism, however, feeds a sense of helpless resignation which can leave you depressed and trapped.

Passively watching a non-expert or self-appointed divine on a given topic on YouTube has become especially tiresome lately, and I feel the medium of content creation has run its course for most creators. The weakness of proliferation is that, for a practitioner, it slowly gives away secrets that aren't very good or worth keeping. It can lead to a healthy kind of disenchantment. Content creators are forever slaves, trying to crank out new material that can hold their audience's attention. It sounds like hell.

Ordinary viewers may grow bored and dissociate from the real world, never going beyond that. Or maybe they find a hobby like drawing or exercising to break the habit of stream watching. Those are better activities, in my opinion, but they're also just substitutions that can keep someone from looking closely at this thing we call "reality." It would be fine to wake up one day and see that everyone had shut off their smart devices and re-engaged with the world, for better or worse.

When I do watch content from new suggested sources on YouTube, I can spot the craving, delusion, anxiety, dissatisfaction, and dukkha easily. I can also choose to fall into aversion to the creators' suffering-inflected narratives about themselves or current events. That's the biggest hurdle of spiritual life: seeing your own deficits while others blast them back at you from a different vantage point. Our weaknesses feel like personal belongings, while those people right over there are doing something totally different, and their version isn't as intense or as meaningful as mine. My pleasure, my pain, my happiness, my joy are MINE!

Frontier models have prioritized coding tasks with the eventual goal of replacing IT people. IT people being IT people, we're all happy to help out. But on the flip side, if you can replace IT people, you can replace any white-collar job. Any. White. Collar. Job. Then it's on to the factory worker, the plumber, the electrician with humanoid robots.

First they came for the sys admin, but I was not a sys admin...

Next, they came for the CPA, but I was not a CPA...

Then they came for me, the dog groomer at PetsMart, but no one was left to defend me...

Okay, let's move over to the latest in doomer AI news. But let's set it up first by talking about the market.

Here's tonight's NQ futures chart of the Asia session:

On April 27th, 2026, the NQ, the index of the 100 largest non-financial companies on the Nasdaq, traded at 23,328. Barely six months later, it went to over 31,000, an all-time high. The tech sector's best and brightest companies (expressed as an abstract number stapled to a single basket called "NQ 100 Futures") moved ~8,000 points in a half-year, what would've taken years to achieve before the AI build-out. Dead permabears everywhere as a never-before-seen surge of cash flowed into semi and cloud stocks like Micron, Nvidia, Meta, Microsoft, and Google.

Trillions of dollars have literally poured into the sector in search of Artificial Superintelligence (ASI), leaving the rest of the global economy to stagnate. The world's wealthiest, most powerful people have staked everything on the species' greatest technological breakthrough.

This isn't a typical capitalist bubble, and people who talk like this don't grasp the scale of the investment. This isn't Pets.com or ridiculous orders of fiber optic cable. Every bobbing head on YouTube wants to compare this to the early-2000s DotCom bubble. This is not that. This is the whole ball of wax: a possible monopoly on human cognition. Ed Zitron's business model is saying, "Yeah, but the numbers just don't work out..."

Whole ball of wax vs. the "accounting doesn't stack"...

Ed believes he's found a secret no one else has uncovered, and it's his life mission to spread the word. Does he think the accountants at OpenAI, Anthropic, and all their backers don't know this "secret"? Ed has never played high-stakes poker for real money.

Bill Gates, in this interview with CNN's Jake Tapper, warns again of the dangers of the AI transition, admitting the real possibility of a biological weapon being developed using AI that lays waste to the planet. He picks cyberattacks as the lead destroyer, but to me, that includes hacking an AI-driven biolab.

Long-term readers of The Tangle will recall that we've warned of this as well: the "rogue," mentally ill radical loner who manages to coax the secret of some awful biological stew that precipitates a mass die-off of humans out of a late-night chat with a chatbot. It will sound ludicrous to blame a lone nutter for the destruction, and it will quickly become the official narrative. No one will care, either, because, unlike COVID, this will be the real deal, with high mortality, economic collapse, and social breakdown. In another interview, Chairman Gates warned that the transition from the current world order to the new AI-led system could cost 1 billion lives.

The next bioterrorist will not use Argon 4; I'm certain of that.