The Exigent Duality
Integration Points - 17:12 CST, 9/01/23 (Sniper)
I'm sitting at the Toyota dealership getting my first free oil change and tire rotation for the 86. It's amazing how well Apple's products all integrate: I plugged the iPhone 14 Pro Max into the car, was following CarPlay's "satnav" the way here, while listening to some "Galneryus" via Apple Music. I got here, opened up the M2 MacBook Air, and it asked me if I wanted to connect to my phone's 5G connection-- within two seconds of that, I was online.

I'm not enamored with Apple as a corporation-- they are woke and evil like the rest of them. But between them, Google, and Microsoft, they are the "cleanest dirty shirt" by some margin. And I'm ok with living in a pseudo-walled garden, provided things work. I never run into bugs with Apple's software, it all "just works", it stays out of my way and lets me work, and their hardware is absolutely top notch.

About the drive here itself: the best route to get from my home to this dealership takes me through some absolutely zany country roads. In track mode especially, the 86 is a sensational car on those kinds of commutes, where the relative lack of power isn't as relevant, because you can keep the momentum up between the turns. In "The Need for Speed" fashion, I've concluded that with the stock tires, 110 mph is at the "95% of limits" threshold for the typical arcing country bend: the car is very communicative, and I could tell that even another 5 mph would break the rear loose.

Artificial "Intelligence"

This video was kind of annoying to watch. It's not that I fault people for not understanding how these "Large Language Models", or "LLMs", work-- not everyone is a computer scientist, and as a software developer even I don't know all of the intricacies. But all the same, the aforelinked video is generating a lot of alarmism, and I think I can assuage at least some of those fears.

Imagine this: you show the computer the text "Roses are red, violets are blue." Then you prompt it: "Roses are red...", to which it replies, "violets are blue." It's not that the computer program has a brain, or is "thinking". It has no "intelligence". Applying the term "AI" to these LLMs is a total misnomer; all the program is doing is taking text you feed it, and extrapolating.

Companies like "OpenAI"-- another misnomer, as nothing their company does is "open"-- don't just feed that one, simple poem into the program: they shovel all of the text on the whole internet into it, and then some. But the principle is the same: if you ask the program to create a short horror story, all it's doing is replying with the equivalent of "...violets are blue". "Which words are associated with horror stories", and the program generates paragraphs with the associated words and notions.

It's not "smart", it's not "creative", it's not "thinking"-- in fact, it's very basic and dumb. It's just taking text inputs and replying with text outputs. Input, output.

People have asked, "Why does ChatGPT 'make things up'". The answer is that it doesn't! A co-worker asked it for a waffle recipe, and it said you should add a waffle iron as one of the ingredients. ChatGPT and its ilk don't "know" what a recipe is, they don't "know" what an ingredient is-- they just see that the text "waffle iron" is frequently associated with the word "waffle", or the words "making waffles", so they throw "waffle iron" into the response.

People have also asked, "Why can't ChatGPT do math very well?" Again, these LLMs are dumb: they just produce text. They have no idea what a "number" is, or what "math" is. They don't have brains; they have the IQ of a goldfish-- less than that, even. So when you feed them a math problem, they're just looking at what words are associated with the number words you feed them. Sometimes by chance that will produce a sentence with the correct math answer-- most of the times it doesn't, just like one would expect given how the software works.

"AI" developers are currently mitigating limitations like this by adding "plugins", such as calculators-- so when the LLM gets what looks like a math problem, it can feed the equation into the math sub-routine-- another dumb program-- and proxy back the output. But that's not "intelligence" either! You wouldn't call Excel "AI" because you can put a formula into the cell, hit enter, and get the answer.

In the hacked "Matrix Awakening" demo video linked above, when the guy asks one of the NPCs "Are you a demon?", ChatGPT is scouring its terabytes of text, and is merely echoing back answers it's found in fiction works and elsewhere, where that question has been asked. When the person poses to the NPC, "What would you do if I could get you out of the simulation?"-- once again-- ChatGPT is scouring databases of science fiction text and everything else, and is extrapolating response text in the form of a sentence.

What it isn't doing is "thinking"-- like the Tin Man, it doesn't have a brain! It has no concept of anything! It's just a dumb computer program which generates natural language. That's it.

Eventually, maybe humans will be able to make a fully-"neural net" computer program like the character "Data" from "Star Trek: The Next Generation"-- who knows. That concept, by the way, is called "General Intelligence". But ChatGPT and other similar systems do not function that way: they are as far from "General Intelligence" as your kitchen toaster. When you ask it if it can be "possessed by a demon", it's not "answering" you-- it's just generating text!

Late-Stage "Capitalism"

As I've pointed out before, in my job I raise the share price of a multi-billion dollar corporation by a tiny, hardly even measurable fraction of a fraction of one percent-- but for that infinitesimal gain, they are willing to pay me and my coworkers multiples of the median income in the country. We're so deep into the "squeezing the financialization lemon"-- "late-stage Capitalism", if you will-- that it's beyond absurd.

Sticking to the "AI" subject then, take a look at this article. Any pretense of these corporations or power brokers caring about their fellow man is right out the window by now: they are willing to replace the livlihoods of actual human beings with these pieces of software, so that instead of keeping on Sniper, who raises the share price by .00001 percent, they can replace him with an "AI" which will raise the share price by .000011 percent! The lemon is so dry now, that it can only be further squeezed by tossing millions of people out of their jobs.

I also noticed that Sony has drastically increased the cost of all three tiers of their "PS Plus" service. This seems really tone deaf, as drug epidemics rip through the world, and people are increasingly unable to buy food. Maybe the electric bill on their server farm went up, and they need to pass on the costs-- who knows. But I wouldn't be surprised if their customers said, "Screw it, I'll just play games on my PC where the 'online' bit is free."


Here are a few assorted crappy things going on in the world today. First, there is this video which confirms that, yes indeed, the government is spraying junk into the atmosphere, and has been for decades. "Junk" which, incidentally, causes cancer in humans.

Then you've got this situation brewing. First the Cultural Marxists took over the elementary and high schools-- now they've captured the law schools. Can you imagine in a few years, once these people are fully-barred lawyers, trying to get a fair trial in America? These people are taught that you absolutely must take the race and sex of the perpetrator and victim into account. So much for Common Law, and equality under the law! And I'm not being theatrical about the subject: this will have a very real, disastrous impact on real human beings you know and love, if not even yourself!

Also in the United States, the cities are turning into London, with cameras everywhere.

And speaking of cameras, VAR is so great in the world of football, that stadia aren't allowed to show replays of "controversial" match events! They say transparency is the best disinfectant-- what does it tell you about the effectiveness of VAR, when they have to hide the incidents from the spectators? In the case of this specific match, Ciro was shoved down and Zaccagni had his ankle demolished, both incidents having taken place in the area-- no penalty either time. I thought VAR, in exchange for slowing the game down and turning every incident into a circus, was supposed to "lead to the right decisions"? What exactly did we gain from this trade-off we've made?

Another crappy thing going on is that the EU is trying to censor the entire internet. Post the WuFlu IFR rate straight from the CDC web site? Why, that's medical "misinformation" you insidious jerk! You trying to get grandma killed?? Say that women don't have penises-- straight to the penal colony with you, bigot! Andrew Torba has been getting a lot of these European "you must take this posted content down" requests, and he tells them to shove their requests up their arses. But not a single one of these other tech companies have Torba's backbone holding them up.

The final point for now in this litany of bad news: if the mandated kill switch wasn't enough to get you to buy a "new car of forever" prior to 2026, how about this nonsense? Look, I get that this woman is grief stricken, and I'm sorry for her loss. I would be devastated if this happened to either of my kids. But this is "my child got struck by lightning and died, therefore we need to ban all human beings from going outside" territory. Bad things happen in the world: that doesn't give you the moral latitude to then foist your insecurities on the rest of us.