|
| 1 | +--- |
| 2 | +layout: page |
| 3 | +title: AI-arrhea |
| 4 | +permalink: /aiarrhea/ |
| 5 | +--- |
| 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. |
| 8 | + |
| 9 | +# 100 Major Recent Developments in AI Engineering & Data Annotation |
| 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.* |
| 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. |
| 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. |
| 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. |
| 18 | + |
| 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 | + |
| 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 | + |
| 23 | +*Need any more proof?!!!* |
0 commit comments