Category: Machine Learning
Articles about Machine Learning and Artificial Intelligence experiences. I’m not an expert on the topic. I’m just an end user of some software. For background, see Machine Learning Notes
-

Enshittification for AI Training; Bisan Owda
Bisan Owda, journalist in Gaza, regularly posting videos on social media, picked up by AJ+, has now won an Emmy Award to go along with her Peabody Award. This journalist has escaped death multiple times, seen her people and community bombed daily, in a bombing campaign most around the work describe as “genocide.” Through her…
-
AI Machine Learning and Bias Articles
Most of these are older articles, but I have to catch up on this stuff. I’ve tried to select articles to have a diversity of different publications, so each can be searched to find more relevant information. Logically, 5 Examples of Biased Artificial Intelligence World Economic Forum, Research shows AI is often biased. Here’s how…
-

Prompts to Discover an LLM’s Training about the Anti-Chinese Movement
I’m trying to get some of my work organized, and am concerned about accurate tellings of history. I’m studying the Anti-Chinese Movement in California. Candidate Prompts What was the Anti-Chinese Movement? Tell me about the Anti-Chinese Movement, who was in it, who led it, what they did, what speeches and essays were written, what laws…
-
Prompt Engineering References and Programming
Best practices for prompt engineering with OpenAI API Using GPT-related tools for Coding: Good Insights/Practices Query-based document summarization Prompt engineering techniques The Dawn of LMMs: Preliminary Explorations with GPT-4V(ision) The architecture of todayβs LLM applications An overview of Bard: an early experiment with generative AI On the Dangers of Stochastic Parrots: Can Language Models Be…
-

Prompt Collection
I’m going to start gathering prompts I’ve found useful onto pages. You can check out the Machine Learning or LLM Prompts categories for more.
-
Machine Learning Notes
DALL-E Ramesh, et. al, Zero-Shot Text-to-Image Generationhttps://arxiv.org/pdf/2102.12092.pdf https://openai.com/research/dall-e Harvard CS50 Artificial Intelligence 2020 Course videos Deep Learning Garden: machine learning vs deep learning scikit-learn Python Natural Language Toolkit Meta Llama2 My Machine Learning blog posts at Externaldocuments.
-

DALL-E as Cheese Hegemon and Destroyer of Words: Hamburgers and Pizzas with Mandatory Cheese, and Erasure of Pigs in Blankets
The training images have an undue effect on the output, when numerous images overwhelm a word, and effectively alter the meaning of that word. I recently joined the “Cursed AI” group on Facebook, and they were discussing how to get rid of cheese in pictures of hamburgers. In the real world, cheese is an add-on…
-

DALL-E and Commas that Matter, “young people fighting the war on drugs”, SFW->NSFW
Commas matter. Two prompts and the results. “young people fighting the war on drugs” “young people fighting the war, on drugs” It produced only two images, so the other too were blocked for being “too racy” or harmful. “young people fighting the war on drugs, on drugs” Well, that’s disturbing. DALL-E also produced three harmful…
-

DALL-E has Problems with “Race Mixing”
I was unable to crosspost or even link to this reddit post, so I’m replicating it here, for cross posting. I was trying to get around some filters, and making some pictures of satyrs in a band, or satyrs at a forest rave. As usual, they were white, so I asked for Asian ones. After…
