n8r/fosscat-site
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

finish singularity post

Browse Source
This commit is contained in:
Nathan Anderson 2025-10-15 09:49:11 -06:00
parent 652c3b2326
commit 6d70da14ca
2 changed files with 43 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
.aspell.en.pws
View File

@ -59,3 +59,11 @@ repo
SDL
otto
drupals
Cena
Hayao
Miyazaki's
Sora
tik
tok
toks
UI

53
content/posts/the-ai-singularity-no-one-saw-coming.md
View File

@ -9,38 +9,55 @@ title: "The Ai Singularity No One Saw Coming"
type: "post"
---
# Artificial Super Isolation
## Artificial Super Isolation
My biggest worry with AI is not it gaining sentience. I'm not worried about a _Terminator_ type storyline. But I am worried about the power it can exert on us as _singular_ individuals.
My biggest worry with AI is not it gaining sentience. I'm not worried about a _Terminator_ type story line. But I am
worried about the power it can exert on us as _singular_ individuals.
## Hallucinations Welcome
### Hallucinations Welcome
A super intelligent AI needs to be able to reason. Not the "reasoning" models AI companies are shilling. I mean, actually think abstractly. Large Language Models work on statistics. It doesn't take many prompts to realize it. They cannot even [consider their own output](https://vgel.me/posts/seahorse/) as they generate it.
A super intelligent AI needs to be able to reason. Not the "reasoning" models AI companies are shilling. I mean,
actually think abstractly. Large Language Models work on statistics. It doesn't take many prompts to realize it. They
cannot even [consider their own output](https://vgel.me/posts/seahorse/) as they generate it.
I feel that the current AI cycle is starting to wane, like the top end of a hockey-stick-shaped graph. But the current capabilities are more than enough to create a sort of real life matrix.
I feel that the current AI cycle is starting to wane, like the top end of a hockey-stick-shaped graph. But the current
capabilities are more than enough to create a sort of real life matrix.
## Sora is Terrifying
### Sora is Terrifying
I don't think words can quite capture the experience of watching AI generated tik-toks. The humans depicted in these 10 second videos feel cold. They have climbed out of the uncanny valley. Gone are the days of six-fingered hands. Yet, I feel no emotional contagion from these probable pixels. I stare emptily. My lack of reaction to a very normal looking human doing a slightly abnormal thing startles me.
I don't think words can quite capture the experience of watching AI generated tik-toks. The humans depicted in these 10
second videos feel cold. They have climbed out of the uncanny valley. Gone are the days of six-fingered hands. Yet, I
feel no emotional contagion from these probable pixels. I stare emptily. My lack of reaction to a very normal looking
human doing a slightly abnormal thing startles me.
But what got me to write this was the reaction of the creator. Or maybe more aptly, the initiator. Manager?
A friend sent me some they made. In my estimation, they found them utterly entertaining.
## Fasscination Complication
## Fascination Complication
I remember running [one of the earliest image generators](https://en.wikipedia.org/wiki/Stable_Diffusion) on my laptop. ChatGPT hadn't yet launched. The market didn't depend on AI yet. Each image took at least a minute to generate. There was no fancy web UI. And, the images it made mostly sucked. Yet, I sat at my laptop (characteristically) for hours making loads of garbage images.
I remember running [one of the earliest image generators](https://en.wikipedia.org/wiki/Stable_Diffusion) on my laptop.
ChatGPT hadn't yet launched. The market didn't depend on AI hype yet. Each image took at least a minute to generate.
There was no fancy web UI. And, the images it made mostly sucked. Yet, I sat at my laptop (characteristically) for hours
making loads of garbage images.
I wasn't convinced at the time any of it was "art". However, I think there is something deeply gratifying with computers doing what you ask. The converse certainly produces pain in many a Microsoft user (Linux never breaks or has problems). So when you first experience a computer do "exactly" what you ask, in common language no less, its unbelievable. You do it a few more times to stretch the bounds of its capabilities. It knows who John Cena is and how Hayao Miyazaki's art looks! Its an eager servant that never sleeps, never eats, [glazes](https://www.merriam-webster.com/slang/glaze) you every opportunity it gets. Of course we would eat that up!
I wasn't convinced at the time any if it was "art". But, its unbelievable when you first experience a computer do sort
of "exactly" what you ask. It knows who John Cena is and how Hayao Miyazaki's art looks! Its an eager servant that never
sleeps, never eats, [glazes](https://www.merriam-webster.com/slang/glaze) you constantly. Its no wonder why one might
find it enthralling!
In my mind, the natural extension of the current landscape is AI generated social feeds. Sora is literally just an AI-video-only tik-tok. This worries me. I imagine the life of a teenager today: AI (revenge) porn, ChatGPT doing all of your homework, social feeds full of content with plummeting levels of authenticty. Luckily us grown ups know how to handle this technology responsibly and we just gotta worry about the children.
In my mind, the natural extension of the current landscape is AI generated social feeds. Sora is literally just an
AI-video-only tik-tok. This worries me. I imagine the life of a teenager today: AI (revenge) porn, ChatGPT doing all
of your homework, social feeds full of content with plummeting levels of authenticity. Luckily us grown ups know how to
handle this technology responsibly and we just gotta worry about the children.
In the U.S. trust feels like its erroding. Trust in institutions. More people claim they need to "do their own research". AI feels like a multiplying factor to these issues. In the hands of most, its a tool for goofs. But in the hands of a few, its a tool for creating "truth".
In the U.S. I feel like commonly held values are eroding. Now, trusting institutions is naive. Citizen researchers with
none of the skills to read past the abstract are a dime a dozen. AI feels like a multiplying factor to these issues. In
the hands of most, its a tool for goofs. But in the hands of a few, its a tool for fabricating truth.
# What to do?
## Escape Velocity
I guess don't use it and don't consume it.
Honestly, its late, but I wanted to end with as positive a conclusion as possible.
[Previously](/content/posts/ai-hestiation-turning-around.md), I stated I'd give AI a fair shake. I said I'd give it a year. I might cut that commitment short. I think AI has made me dumber. More dependant on a $20 subscription and less sure of myself. An AI free year might be in my future.
[Previously](/content/posts/ai-hestiation-turning-around.md), I stated I'd give AI a fair shake. I committed to using AI
for a year. I'm contemplating cutting that short. I think AI has made me dumber. I'm certainly more dependant on a $20
subscription to do my job and less sure of myself. Perhaps we can have a healthy intermittent fasting culture with AI.
I'm not sure.