Anthropic’s Claude.ai is the new [AI] kid on the block. Yes, I did link to one of “those” videos. One of Claude’s unique features is that users can upload file(s) (.pdf, .csv, .txt, and the like) and then ask questions about them. You can’t do this with the free versions of ChatGPT, Bard, or Bing…
Category: Uncategorized
Make Microsoft Word “macros” your power hack for plain-language editing
Why it matters Readers know why they prefer writers who use short, simple, and familiar words and phrases: before > antecedent to | so > accordingly | matches > comports with. Unnecessary words should be struck (like pulling weeds from your garden and flower beds): enough > a sufficient number of | [omit] > as…
Lawyers need to “up” their pleading game when describing generative AI exchanges
A Gwinnett County, Georgia court will handle what’s thought to be the first defamation lawsuit about a ChatGPT response (h/t Bloomberg Law and Professor Eugene Volokh). The June 5, 2023 complaint is thin on important details about the non-party user’s ChatGPT exchanges (or any damages suffered by the named plaintiff who has a fairly common…
ChatGPT was never designed for legal research (and here is why it should NOT be used that way)
Even though ChatGPT has some pretty clear disclaimers about factual accuracy, two New York City attorneys and their law firm gained a lot of public attention for claiming that the bogus cases cited in their federal court filing were the result of (wrongly) using ChatGPT as a legal-research database. Federal District Court Judge P. Kevin…
When judges add descriptive footnotes (and an itemized appendix) as they “write” about “inserted images,” they preserve the judicial-decision tapestry.
Why? Because many third-party-online publishers leave images out from the original decision. Imagine footnotes like these being dropped into a judicial decision: [Proposed new footnote:] To help future readers who may read this decision in a third-party, text-only format that left out the image, Figure 1 is a black-and-white portrait photograph of Prince taken in…
Plain-language editing and other AI prompts (ChatGPT, Bard, and Bing) that work for me
A downloadable and printable handout is sometimes more useful than a text-based blog post. (At least for this “prompt engineer”.) Sharing what works for me, here is a downloadable three-page guide of the plain-language editing prompts that are helpful when I use ChatGPT, Bard, and Bing. Don’t be shy in sharing your experience and suggestions….
Here is how ChatGPT, Bard, and Bing measure sentiment, emotion, and tone
#AppellateTwitter wasted no time criticizing an experienced SCOTUS advocate for his recent brief’s word choice: But wait. Nearly one year (!) after the Court invited her views, the Solicitor General has delivered what can only be described as a hot mess of a brief. In it, the government claims to have unearthed a new obstacle…
I’ll pass on asking ChatGPT to reliably “summarize” court opinions
The Michigan Court of Appeals recently issued an interesting published opinion about two siblings charged with truancy, their right to counsel, and whether they waived that right in the trial court. Court of Appeals Chief Judge Elizabeth Gleicher wrote the lead opinion. She also penned a separate concurring opinion. Judge Thomas Cameron also wrote a…
“Chatting” with ChatGPT about plain-language editing
I had some Q&A banter with ChatGPT because I wanted to get curious about how ChatGPT works and later share a glimpse of how it can help during plain-language editing. Our exchanges are captured in this video. A transcript is reprinted below. (If the video text seems too small when the video is viewed within…