I saw a headline Windows dashboard about a popular internet provider that filed for Chapter 11.  I was curious and decided to check it out and clicked the story.

The news item is from “TheStreet” and the initial page does not show the entire article, just the first part.  I think this is equivalent to “above the fold” in the newspaper vernacular.  I decided instead of scanning the text for the most recent Chapter 11 victim I would select the Copilot “Summarize” button nicely placed in my view, to the left of some adds monetizing the page.

I was surprised by the summary in that it mentions Wolfspeed (semiconductors) and Solid (fintech), neither of which would be classified as a popular internet provider.  

I then expanded the story and read further about Evergreen, which is a popular internet provider, filing for Chapter 11.  When I tell Copilot to Summarize again, after expanding to the full article, it does include Evergreen in the summary.

What happened?

I expected Copilot to summarize the entire article, not just the excerpt that was originally shown.  I think this is a logical expectation from a non-technical perspective – the button says summarize, but that is ambiguous and does not specify “Summarize only what you see right now as the article” versus “Summarize the entire article.”  I would guess from a technical perspective Copilot is summarizing what it has access to locally (i.e. in your local browser) and not making a second request to TheStreet for the entire article to then summarize.

 

Always clever Google

Tuckahoe! I wanted information on how RAG and Live Intenet Search work with LLMs.  I chose Gemini 2.5 Pro Reasoning, Math & Code for the task.  The final example included as a follow up to expand on real time search included my physical location, which is freely...

read more

Always clever Google

Tuckahoe! I wanted information on how RAG and Live Intenet Search work with LLMs.  I chose Gemini 2.5 Pro Reasoning, Math & Code for the task.  The final example included as a follow up to expand on real time search included my physical location, which is freely...

read more

Comparing LLM responses

LLM Responses compared I thought a nice exercise would be to take a relatively simple prompt and assess how the Closed and Open models currently available compare.  This is the prompt that I used: <PROMPT> I’m taking my daughter to an oral surgeon today to...

read more

Perplexed-ity?

Perplexed-ity? I came across this blog post from Cloudfare: Perplexity is using stealth, undeclared crawlers to evade website no-crawl directives I've read and heard a lot of positive things about Perplexity's Comet browser.  I want to like them and I want to cheer...

read more

AI Action plan, and stuff

AI Action plan Here ya go folks - this is the current administration's AI Action Plan:  https://www.ai.gov/action-plan Here are some words from the current administration about preventing "woke AI" in the federal government...

read more

Subscribed to One Useful Thing

One Useful Thing is the name of Ethan Mollick's substack newsletter.  Ethan is the Co-Director of the Wharton Generative AI Labs. Wharton Generative AI Labs has lots of good information including a prompt library: https://gail.wharton.upenn.edu/prompt-library/ Check...

read more

“Accumulation of Cognitive Debt”

There's an article published at MIT that studied "Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task https://arxiv.org/abs/2506.08872 This blog post was pitched as a rebuttal of sorts to the MIT study - definitely...

read more