Recent Comments

"I don't watch CNN..." well said, LOL! All the colors in my peaceful Islamic pie chart are zero, too - good one. Whether Sen. Kennedy said that or not, it's still true/funny! The Kurt Russel meme is spot on. Oh heck, I could go on commenting but don't have the time. Great collection, Dobrien, thanks for posting!

Great addition, OUC! Thanks. Now, I'm going to be rolling Pink Floyd through my brain all day - not a bad thing, LOL! "Learning to Fly" is one of my favorites. Ha, the last frame of Spring Break calls up whack-a-tollah to mind, LOL.

Nicely done, OUC!

I believe it was 2012 for almost 2 months - 12 hour days, 6 days a week. I worked in the Richmond facility north of Oakland - place was huge almost 3000 acres if I recall. My hotel was in Oakland. Some time before that I contracted for software work on a water purification plant for Oakland where methane was produced from the waste and used to run a pair of power generators. More generators were being built because all that Oakland poop made lots of gas and excess was burned off in the atmosphere. That's how the folks at Chevron got my name and called me for the job.

"Fruits and Nuts" LOL. I'm from western NY, but knew an acquaintance 40 miles outside Oakland and asked if we could meet for a dinner. He agreed and drove in to say howdy and chat. Some of his first words were a statement that he very rarely goes to the communist parts of the state. I didn't like Oakland at all.

Thank you. I enjoyed Part I and am looking forward to Part II. I am grateful authors like yourself are willing to share some of your work here in the Gulch.

My favorite is 'Where Does It Say..'
Seen it before, 20 times, but it still hits.
I've moved around over the years,
seen all kinds, it is not an ethnicity thing.
One family, arrived from Iran in the sixties- solid patriots.
Some locally born who should be sent off- well somewhere else.

What Years? I worked at Chevron (IT) in 1992... For 93 days. But in San Ramon, near Oakland!

I left because I realized CA was full of Fruits and Nuts (people, not food), and they elected FLAKES to run the place.

Ya got me right off the bat with "1 in a million actually turn out to be human being" . . . .

The AI Singularity is HALF about Believing it. Right now it is the Gold Rush Days. The Manufacturing Days, or the Nuclear Days. We are throwing money at it because THEY believe that whoever gets there first, gets such a major advantage, like we had with Nuclear Bombs.

You are right about the LLMs not "really" being intelligence, per se. We have an information 'encoding' problem. Currently all of the best data is basically "information" in the form of language. It's all we have, and we have GOBS of it. And it is how we communicate. It is how we express questions and answers. And 99% of the problems we need to solve are OKAY to be solved with language and a DECENT answer.

The best comparison is that it's a 5 year that has memorized a full encyclopedia. It can sound smart and deep, until it needs to create something.
At which point it will dream up something that "sounds good" to it's word processor algorithm.

Right now, they are BETTER search engines than google ever was. As for coding. I love it. because of over 50 years of coding... I can get pretty direct/clean/detailed answers or code... Because I know how to write a spec.

A few years back, I started hacking on PostgreSQL source (PG). And the hardest part was the 2 days it took me to figure out how to install all of the compilers, and everything else, WHILE I was learning Linux. Before I Could even begin to start "searching" through the code. Hours and Hours searching through FILES to get to the ONE FILE, to get to the 3 areas I had to change.

This year, I downloaded Cursor from Cursor.com. I gave it access to a sandboxed (clean Linux Box). I told it to download the GitHub code for PG, and to compile and test it. To install all of the appropriate tools to do that. In 15-20 minutes it was done with a few prompts, which I checked the box to stop prompting me, and just do it. And then I told it the change I wanted to make. In about 10-20 minutes it found all of the code, made the changes wrote a test script, and validated that it worked. then I jumped in and tested it. It created the patch for me to submit to the PG hackers to review and accept. (I have NOT sent that, because I really need to review the code, the comments, etc. Much deeper, or risk my credibility with the group).

Anyways. That is a LONG way down the line from what you are saying the limits currently are.
Now, the "Singularity" is not here. The code is not improving geometrically. But it's usefulness is growing. it's capability is growing.
The goldrush is on for that Singularity.

These things save me DAYS of work every day I use them. it should NOT replace ALL of your coders. That's like a bank with ONLY ATM Tellers (Ally, LOL).

You are 100% right about POWER. Gates, et al have dropped Climate Change like a hot potato. Why? Just the hunt for RED AI?
or have people realized, you take advantage of the FEAR and GREED, and make sure OUR country will have PLENTY of Electricity (Even Nuclear), and PLENTY of Chip CREATION ability (3 new FABS, here in the US).

I think they are using the AI Singularity to create what we need anyways. To clear the roadblocks to Nuclear Energy, and power plants being built. (The ugly secret is that the chips keep improving so fast, that 36month old chips have to be replaced, because they are so SLOW by comparison).

May we live in interesting tims.
PS: I am a naysayer on business adoption of most AI, because all of the projects I watch are failing. Especially the HR replacements. (Nothing like BAD Advice from an Automated HR Department...)

I’ll see you on the dark side of the moon. I bought that album when it first came out. Fantastic , first song to last. Btw “Don’t memetion it” just posted

. . . that's hilarious! More of the same. Decades of progressive propaganda served to fear and detest the Militia.

Best to attend local Pew Pew shows to find one.

. . . I doubt it. It's very expensive and known as the idiot model because it's self-adjusting.

The lunatic is on the grass remembering daisy chains and laughs hahahaha HAHAHA . . . .

Hello mccannon01, and thanks for taking a moment to comment. I Part II I bring the influence up to the present, including the "reaction" to postmodernism.

The Rocket thrust and refracted light crosses the blood brain barrier, oh, I mean, crosses into the outer border. You can't do this with a photo trick, it has to be done, one color perfectly matched at a time. The refracted light was easier compared to the translucent multi color rocket thrust.
Seeing that this was outer space, why could they not continue forever.

Just needed to change it up, create more beauty and learn a new skill at the same time.

Nice work brother The lunatic is in my head….

Unless it's putting several cells into one single cell that you've been doing for a while now, I'm not sure what the new Visual technique is. Please elaborate.

Created a new Visual technique, (the hard way) let me know what you think.

Also, the AP-NORC poll conducted March 19-23, 2026 is worth examining in detail because it is among the most methodologically sound surveys being fielded, and because its own data, read carefully, undermines the way its findings have been reported. I just didn't have room in the main post.

Start with what AP-NORC does well. They use the AmeriSpeak panel, a probability-based panel recruited from a frame covering 97 percent of U.S. households. This is categorically different from opt-in online panels or phone surveys with single-digit response rates. The 2024 vote composition of the sample, 29 percent Harris, 30 percent Trump, 41 percent didn't vote, closely matches actual election results and turnout, suggesting the partisan balance is reasonable. If you are going to poll Americans about a war, this is a defensible way to find them.

Finding them, however, is not the same as getting them to answer. The weighted cumulative response rate is 6.8 percent, meaning roughly 93 out of every 100 people in the original probability sample never completed this survey. The weighting adjustments correct for observable dimensions of non-response (age, gender, race, education, 2024 vote) but they cannot correct for the unobservable: the correlation between willingness to answer a survey about Iran and the strength or direction of one's opinion on the topic.

The respondents who do answer then encounter a question sequence that shapes their responses before they reach the war questions. The topline questionnaire reveals that respondents were first asked about Trump's overall job approval, then about his approval on the economy, trade, foreign policy, and Iran specifically, all before reaching "Has the U.S. military action against Iran gone too far, not far enough, or been about right?" By the time a respondent encounters that question, they have already activated their partisan disposition through a sequence of Trump evaluation prompts. The "gone too far" answer is downstream of a priming cascade the survey itself induced.

Finally, the subgroup sample sizes are too small to support the analysis most commonly built on them. The margin of sampling error for independents is +/- 8.8 points. When coverage reports that independents oppose the war 64-28, the real confidence interval for the opposition figure runs roughly 55 to 73 percent, the difference between "independents are mildly skeptical" and "independents are overwhelmingly opposed." The data cannot distinguish between these readings, but every headline treats the point estimate as settled fact.

ugh, one missing or accidental "*" and the whole thing gets funky formatting. Apologies for missing that, but I'm not paying to edit posts when there is no preview.

Thanks for your support . . . .

"We are looking at the AI Singularity."

Not even remotely true. I work with so-called "AI" professionally and even develop tooling for it. But it is not actual AI. What is being called "AI" is actually just Larger Language Models, and there is no reasoning, no thinking.

The LLMs are "trained" on human text (mostly) and input, and as such are basically a mirror of what they are trained on. They aren't remotely as good as the doomsayers in media claim, and not even advancing in that direction.

Consider this: who are the loudest voices claiming the impending all mighty AI? Those who are financially invested in people buying their product. Those companies are losing money hand over fist. They need investors, and that means scaring people by proclaiming how dangerous they are because that is basically fear driven hype.

Yet looking below the surface you can see the cracks and failures immediately.

Lets start with "Job replacement" claims. One of the areas they can work decently - if they are managed by an expert engineer - is writing code. Code has never been the goal of software engineering, it has always been in fact the lowest value component (we used to call it SMOC - Simple Matter of Coding). Yet, these companies who have CEOs or investors out decrying how there is going to be massive job loss because of their product are not shedding coders - they are hiring more and more. Why?

Because engineering requires intelligence, reasoning, and even creativity. And LLMs have, and provide, none of that. Look at YouTube and what are the bulk of the "AI" videos about? Hyping basic automation of low-value, low-skill content creation, and selling their courses on how to do it. The software ones in particular are all basically how to make marketing websites.

But if you learn to fear this all powerful "AI" is going to eat your lunch as a business, you just might buy into it with your limited cash flow. That is why the fear-hype game is in full play; the "AI software" companies are hemorrhaging cash with no real plan to even break even. They are drunk on their own propoganda.

The hardware side, however, has seen the underlying ugliness and found an alternate path: electrical power. If you pay attention you see that these companies are quietly becoming electricity producers to power these massive dedicated datacenters. It is an easy to see play: build the ability to power your DC, then when the bubble pops, you can sell that electricity to the grid. Because that is where their real value lies.

Those who should be concerned are those producing crap easy to mimic fluff content, specifically low-information, low-attention videos and articles - which is one of the major "learn this one trick with AI" content - conveniently made via LLMs - genres. And yet these LLMs can't even accurately do that without significant guidance and management by humans. This is where the term "AI Slop" comes from. You know them, you've seen them. And with no small dose of irony, LLMs can even recognize them.

This is exposing the sheer bulk of that type of content being written by humans. Those are lazy people who will not learn how to use LLMs as a tool to actually get better, but instead will fall for the allure of more output with less effort, thus degrading their already low-value content further. Ultimately it won't be "AI" that puts them out of a job, they'll do that to themselves due to their lack of real interest, work ethic and, for many, lack of ability to actually think, analyze, and face the reality that what they do is simple and simpleminded; they'll not lose it because these mythical AIs are so powerful.

I suspect it will be pretty close to being back to normal operations by maybe mid-late September. The conflict in the Strait is really one of logistics and regardless of political opinion on it, Iran is and will continue to lose that at an accelerating pace.

From a military perspective they are small-squad fighters trying to play set piece battles than anything else. They've chosen a battleground they cannot maintain control over, and to play against a force that has one of the best logistics systems in the world. Meanwhile their logistics is not built to handle the type of battle they chose, and attacking others in their region isn't going to do them any favors. China won't do anything because they can't; for all their bluster they are landlocked and lack the capacity or will. Same with Russia, slightly less landlocked but also lack the capacity or will.

It is possible Iran may try to reach out to European states with their attacks, and if they do it will end even sooner. I don't expect WTI to be at a premium by October, November at the latest. If it is is won't be by much.

Aye, refineries are complex operations that can yet appear almost primitive in many ways.

To us folks who do not have such knowledge, your article is a light going on in a dark room. Thank you for opening a well of understanding.