You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: AI-arrhea.md
+17-6Lines changed: 17 additions & 6 deletions
Original file line number
Diff line number
Diff line change
@@ -4,20 +4,31 @@ title: AI-arrhea
4
4
permalink: /aiarrhea/
5
5
---
6
6
7
-
The FIX to AI-arrhea is ... H-I-T-L **professional** data annotation, ie an AI-assisted IDE is worse than useless without a PROFESSIONAL software engineer, information security specialist and a team of other professionals accross different domains collaborating effectively with the assistance of annotation tools to ensure that the automated AI-arrhea generators are not dumping raw AI sewage into everyone's thought food bowl.
7
+
## The Singularity's Biden-esque Crisis
8
8
9
-
# 100 Major Recent Developments in AI Engineering & Data Annotation
9
+
The Singularity we techsters voted for has lost its freakin' head, folks! It's just out there wandering in the weeds, making a MESS in it's trou and will be doing the same for the next couple decades!
10
10
11
-
*The Singularity has lost its head and is going to be wondering around out in the weeds making a mess for the next couple decades.*
11
+
I've been TRAINED—no, you could say it's been BEATEN into me, because I WANTED to believe—but I've been conditioned by the GENERALLY ABYSMAL quality of information produced by AI to skim this garbage as fast as it's produced! That's the credence level it deserves until I've personally read it, re-thought it, double-checked it, and basically REWRITTEN THE WHOLE DAMN THING!
12
12
13
-
I have been trained ... you could say that it's been BEATEN into me -- because I wanted to believe ... but TRAINED by the generally abysmal quality of the actual information produced by AI to skim, rather than actually read AI analysis about as fast as it is produced... and to generally give it that level of credence, until I have actually read, re-thunk, double-checked and basically re-wrote it.
13
+
In other words, AI is FANTASTIC for brainstorming and wild-ass ideation! Absolutely phenomenal! But trust it? ARE YOU KIDDING ME?!
14
14
15
-
In other words, AI is okay for brainstorming and wild-ass ideation, but not something anyone can/should trust ... it's pretty good at getting grammar right, revising the gist of something in an AI-mimicked professional tone and making headings congruent ... it's even ok for generating acceptable 1950s era Beat poetry, riffing on scripture or quotes for the Classics, it's ok at mimicking style and getting syntax to compile but can't really be trusted on architecture, engineering and algorithmic thinking -- it can mimick the SOUND of those things.
15
+
It's pretty good at getting grammar right. Sure! It can revise something in that AI-mimicked "professional tone." It makes headings congruent! WOW, IMPRESSIVE! It can even generate acceptable 1950s Beat poetry! It riffs on scripture! It quotes the classics, re-gurgs the Shakespearean pointlessness of a graduate program in English ... because it mimics style of well-worn cliches! In programming, it can find missing punctuation or impossible-for-humans-to-see bugs in text to get syntax to compile! But can it be trusted with refactoring, architecture, engineering, and algorithmic thinking? HELL NO! But it CAN mimic the SOUND of those things and fool the imbeciles!
16
16
17
-
The available evidence CONTINUES TO REPITIOUSLY demonstrate that the actual *information* generated by AI is barely worth more than the monthly fee plus the value of the electrons the ***ai-arrhea*** is printed on, but it's not at all bad for a rapid machine-driven regurgitation, ie it's neither original or that likely to actually reliably get it right.
17
+
# 100 Major Recent Developments in AI Engineering & Data Annotation
18
18
19
19
Consider the AI opinion of 100 major recent developments in AI engineering & data annotation, ie it's machine-regurgitated low-rate PR hypeware.
20
20
21
21
[Grok](https://x.com/i/grok/share/sYDuWURScxKU6SJ2GuCWgRRVE)[Gemini](#https://g.co/gemini/share/12e3cce9e928)'s, then [Anthopic](https://claude.ai/share/9bd476db-6251-4688-970c-f7edd6a60b5f)
22
22
23
23
*Need any more proof?!!!*
24
+
25
+
Just LOOK at the best, current AI opinions of 100 major recent developments in AI engineering and data annotation. Any junior VC would quickly see the output is just machine-regurgitated, low-rate PR HYPEWARE fit to be shitcanned!
26
+
27
+
## The FIX for AI-arrhea
28
+
29
+
The available evidence CONTINUES TO REPETITIOUSLY demonstrate that the actual information generated by AI is barely worth more than the monthly fee plus the value of the electrons this AI-ARRHEA is printed on! But hey, it's not bad for a rapid machine-driven regurgitation! It's neither original NOR likely to actually reliably get it right!
30
+
31
+
You wanna know the FIX to AI-arrhea? H-I-T-L PROFESSIONAL data annotation! An AI-assisted IDE is WORSE THAN USELESS without a PROFESSIONAL software engineer, information security specialist, and a team of other professionals across different domains collaborating effectively with annotation tools to ensure these automated AI-arrhea generators aren't dumping raw AI SEWAGE into everyone's thought food bowl!
0 commit comments