Extracting & Codifying Disease Outbreaks Using LLMs: Experiments With Bison, ChatGPT & Claude
Earlier this year we explored the use of LLMs for disease extraction and codification tasks from news coverage, testing how…
How Are Africans Talking About Climate Change And Who Is Doing The Talking?
This study identifies who is talking about climate change in Africa, both in the mainstream media and on Twitter, and…
Visually Describing TV News Broadcasts Using Generative AI Image Captioning: Experiments With Google's New Imagen
With the recent release into general availability of Imagen on Vertex AI, Google's new image-based generative AI foundational model for…
How Putin's Victory Day Parade Was Covered On Television News
Putin's Victory Day parade this past May was supposed to feature "10,000 people and 125 units of various types of…
Experiments In Meme Tracking: Extracting Titles, Affiliations & Relationships
Continuing our meme tracking series, we've seen that LLMs provide subpar entity extraction compared with traditional neural entity extractors. On…
Understanding Hallucination In LLMs: A Brief Introduction
There is a lot of confusion and misunderstanding about what "hallucination" is in large language models (LLMs), how it can…
The Growing Trend & Vast Dangers Of Outsourcing Decisions To LLMs
Not a day goes by the last few months were I don't receive at least a dozen pitches from various…
OpenAI's ChatGPT Holding Steady On Business TV News Channels But Little Mention On Others
The timeline below shows TV news coverage of OpenAI and ChatGPT, showing mainstream channels are mentioning ChatGPT little, even while…
Experiments In Meme Tracking: Cataloging Stories According To UN SDGs
Continuing our meme tracking series, let's look at how LLMs can be used to catalog stories according to their relevance…
Experiments In Meme Tracking: Western Values In A Globalized World, Context & The Problem Of Naive Guardrails
Continuing in our meme tracking series, let's explore using LLMs to catalog stories according to their expression of common forms…
Experiments In Meme Tracking: Cataloging & Classifying Memes By Conflict Enhancing Bias
Long before conflict becomes kinetic, narratives drive divisions within and between societies. Some narratives, especially those revolving around gender, racial,…
Inflation Coverage Remains Elevated On Business Channels & Has Faded On General Channels
Coverage of inflation remains down from its peak but steadily elevated across business channels, while on general news channels it…
Experiments In Meme Tracking: Summarization Stability & Plagiarization
Continuing our meme tracking series, let's take a closer look at the general task of summarization/distillation for English language online…
Experiments In Meme Tracking: Dividing Transcripts Into Stories – Part 4 & The Impact Of Capitalization, Punctuation & Hallucination
Continuing our meme tracking series, let's take a closer look at the general task of summarization/distillation. In particular, let's take…
Experiments In Meme Tracking: Dividing Transcripts Into Stories – Part 3
Following in our meme tracking series examining dividing television news broadcasts into their component stories, let's try yet another approach….
NYTimes: How Trump Benefits From An Indictment Effect
The New York Times examines media coverage of the GOP presidential field. Read The Full Article.
Experiments In Meme Tracking: Dividing Transcripts Into Stories – Part 2
Earlier today we explored a variety of approaches to dynamic semantic windowing of television news transcripts to break them into…
Experiments In Meme Tracking: Dividing Transcripts Into Stories – Part 1
Meme and story tracking involves identifying the core details of a specific story and tracking how it is covered over…
Debiasing Semantic & Generative Search Results: New Risks For Companies
For more than six decades digital search has been based on the humble keyword. A search of a document database…
Why Is Generative AI Red Teaming & Debiasing So Poor? Do We Need To Rethink LLM Red Teaming?
One of the more remarkable elements of our work across generative AI models is the degree to which models presented…
Just Who Is A CEO & How Do We Define Gender & Racial Bias In LLM Embedding Models?
Last week we explored just how devastatingly innate gender and racial bias is in both LLM generative models and LLM-based…
Ukraine's Counteroffensive Isn't Leading To A Surge In Media Coverage
The timeline below shows spoken mentions of Ukraine, including the ramping buildup and then how quickly the media pivoted away…
The Turkish Earthquake At Six Months
How much attention is the Turkish earthquake still receiving after six months? The timeline below shows total daily spoken mentions…
LLM Infinite Loops In LLM Entity Extraction: When Temperature & Basic Prompt Engineering Can't Fix Things
Earlier today we explored how LLMs can enter an infinite loop state with certain inputs when asked to perform extractive…