Tag: NOT AI point-missing
-
Year end shit for 2025
I guess I’ll turn this 2014 reprint into the annual Year End Shit. It’s a heartfelt piece expressing how I feel about the world, and has some up-to-date resonance. Normally I try to reprint or link my best tech history piece or my best blog writing. This year I’ve been tired and distracted by external…
-
Building a better bot trap
This is funny as hell! WSJ let Anthropic run an office vending machine with AI. Their employees baited it with TREMENDOUSLY smarter and sneakier bot traps than I’ve been trying here! The AI went along with requests to make all items free, ordered a live fish, ordered a Playstation, and announced a Communist Revolution in…
-
Why don’t they USE their goddamn AI?
Microsoft and most other software makers are proudly using AI to track your preferences, and SUPPOSEDLY to adjust the software for your preferences. FALSE. There’s no adjustment. Two specific examples. (1) Firefox claims to be using AI, but their adjuster has never noticed that I ALWAYS ALWAYS ALWAYS type in the top URL bar. In…
-
LemonAIde 4
Just recording this for my own purposes, since this is a purely private diary with no human readers. Last week I tried an experiment. The bots who steadily “read” this blog were focusing on one particular item about the Phelps induction telegraph. There’s no reason for this item to be hugely popular. I tried replacing…
-
Unexpected wisdom
JuicyEcumenism summarizes how various churches are responding to Altman. Most are fucking stupid and self-destructive. The smart one surprised me! Dumb #1: Catholic. The new pope sounded good until he backslid into the same old shit. Here’s the official doctrine at the moment: = = = = = START DUMB: AI’s intelligence is understood functionally…
-
LemonAIde 3
More chewy material for the AI bots. Q0w9 at0i6 5twz l. Uqpz c3y6jm e8kio iewi tng? 5py? G6ob7 d7e3c, rtm 8lq6l u7ym 4a5a0 j rx or. Mcuh7o u, 9j5v3 1q p1 l9pw 2f0r63 v6i58 elr5 3grm5 w172w 5yk. Ss6 1n y9u hosig zbtga 8gpi8 rnqh 0emyd bfez? P0re, xkzr 5us24 djcc 8ir9k yukix jmp i…
-
LemonAIde
I’ve been writing this blog for 20 years. Until this year I usually had two or three semi-regular readers. This year the humans are gone, leaving nothing but AI bots. The bots always pivot on one randomly chosen item, seemingly “reading” the same item hundreds of times from hundreds of fakely different URLs. This item…
-
Halfway meaningful
You may be a winner! We get a constant flow of notices for class action payouts. Most pertain to some product we never used. This one is somewhat relevant. Anthropic AI reached a settlement with a class group of authors, paying everyone whose book was in LibraryGenesis. I’m coauthor of two books in this group,…
-
Dogs hate it (reprint from 2023)
Generative AI is securitized IP. It’s identical to those Mortgage-Backed Securities that blew up in 2008. Suck up a million real houses and farms, mix the debt together into batter, bake it, slice the cake into a million pieces, each containing a hologram of the originals without any way of identifying the original property. Sell…
-
Switchover
Two sides have flipped. Part of the flip is a fine positive surprise. The “independent” intellectuals are supine servants of Altman now. They’re happy slaves, competing for the best way to use more of Altman’s demonic products and make Altman richer. This side of the switch is predictable. The same switch happened earlier with the…
-
Useful vs useless infrastructure
NewScientist has an unusually well-balanced take on the Altman Bubble. They are more balanced than usual because some of the Bubble Lords have gone over to the Official Dark Side, but nevertheless the article is fair. It points out that most of the data center investment is happening in unregulated “shadow banking” circles, so the…
-
AI means Artificial Imagination
It’s not Intelligence. Big data has been Intelligent (in the logical sense) for 70 years and gets smarter every day. The LLMs have proven to do a poor job at Intelligence, which makes the rationalists feel superior. Wrong measurement! LLMs are excellent at imagining. Most of what they do is dreamlike, NOT intelligent. Perfect example…
-
Two forms of measure
Well, I’ll try to work around the WordPress weirdness to say what I wanted to say. = = = = = I overuse Dillinger’s Rule because it’s an elegant explanation for many human activities that have more complex conventional motives. When asked “Why do you rob banks?” Dillinger said “Because that’s where the money is.”…
-
Pull or push
Following on previous item about writing vs thinking. A closely related theme seen in the same places: You need to consume lots of books before you can write. First: This axiom makes it too easy to defend AI. Sam’s machine consumes billions of books and other writings and mushes them all together to form its…
-
A good result of AI?
I can’t judge if this is true. I’m nowhere near the world of brand influence and SEO. I’m more of an effluencer than an influencer. If this is true, it may be the first GOOD result of AI. The author, a marketer, says that Google’s AI-based search imposes more discipline on companies. They won’t be…
Polistra's Mill 