Sadly Not, Havoc Dinosaur: The personal website of David Anthony Colarusso
I'm a big fan of lists. If there's one productivity hack worth its salt, it's the list. Create a list of discrete tasks, and check them off as you go. It's simple and it works, and although people joke about their inboxes serving as to-do lists, let's face the fact that a lot of email authors don't always put their asks front and center. Given LLMs' facility with words, I think there's something we can do about that. Last week we introduced a template to help answer emails. Today, we'll ask our…
I built the LIT Prompts extension thinking it would be a nice sandbox for my students—a place for them to "play" with large language models (LLMs). It's turned into so much more. By making it possible to write functions that operate on text in the browser it's done something I haven't really seen anywhere else. Many of the prompts I've described over the last 10 weeks can be implemented in chat…
Simulation has been a large part of this series. See e.g., What if Members of Talking Professions Could Log Time in Simulators Like Pilots?, If a 3L Can Automate Law Students, I Can Automate Law Profs—Robo Socrates Lives!, Summon the Demon, and The Library of Unwritten Books. I like this use case for large language models (LLMs) because it plays to their strengths, or depending on how you look at…
Today if someone mentions "AI," it's a safe bet they're talking about large language models (LLMs). In 2014, however, the moniker most often referenced machine learning (ML). Below we will use today's AI (LLMs) to help make some good old-fashioned AI (ML). It's long been a joke to call statistical methods like linear and logistic regression machine learning/"AI." It's funny, however, because it's…
As many internet users of a certain age will tell you, Google Reader was peak Internet. I wanted to bring back some of that magic with a 21st century spin. So, I created My RSS Algo an open source client-side algorithmically-driven RSS reader. It lets users customize their algorithm and timeline behavior while keeping all that training data local, living on the user's computer. There's no login,…
The Library of Unwritten Books may be the coolest thing I've ever made, not the most important, but the coolest. It creates "novel novellas" on-demand. Unlike text-adventure games with fixed texts, these stories are an open-ended exercise in collaborative storytelling. You are a reader-author. Large language models (LLMs) mediate your collaboration, re-shaping and reflecting your words and those…
I teach a project-based class at Suffolk called Coding the Law. Part of the class involves teaching students to code, not because I want to turn law students into coders, but because if you can build a thing, chances are you have a good functional understanding of that thing. It's about preparing students for the world in which they will practice, giving then a lay of the land so they know what's…
More so than usual, today's template is a proof of concept. That means if you want to make it sing, you're going to have to put in some work beyond the base template. Basically, I took the Simple Training Sims template and added in a feeder template that populates the Sims prompt with a random selection of names and jobs for our client. You'll remember that this simulation is based on my…
If you've been with us since the beginning, you know I think most folks are looking through the wrong end of the telescope when it comes to using large language models (LLMs). That is, too many folks want to feed in 5 words and get out 500 (e.g., write an essay for me). LLMs are sentence completion machines, searching only to find the next most plausible set of words. The larger the prompt you…
Last Thursday on Pi Day, I found myself writing about a program I wrote to approximate pi, as one does, and it occurred to me that I hadn't talked at all about large language models (LLMs) and coding. I haven't written about LLMs and coding here for two main reasons: (1) I suspect as a law school instructor that my readers are more interested in the human language side of things; and (2) it's a…
I fear people hear "AI-augmented word processor" and imagine something that takes in 7 words and turns out 700 (e.g., write an essay on the French Revolution). I think that's a pretty unimaginative vision of AI-assisted writing. As I've said again, and again, that's looking through the wrong end of the telescope. More folks should be feeding these tools 500 words and asking for 5. Today I'll show…
It's been seven weeks since the first and last time we asked an LLM to help us write an email. That template was rather constrained in what it could say. In fact, we only trusted it to politely decline emails. See I'm Sorry, Dave Can't Do That. Today, however, we're interested in doing something a bit different. Today, we'll build upon our work creating dialogues to have an LLM read an email, ask…
Today's template builds on Tuesday and Wednesday's use of entity extraction. We'll be using an LLM to extract the citations from scholarly articles and save them to a csv file. After that we can add columns to the file to help make finding copies easier and to track if we've read the articles. Finding reading all the cites in a paper is often part of the larger process of conducting LIT review.
Today I'll show you how to add magic copy-and-paste to the LIT Prompts extension. What is magic copy-and-paste you ask? Well, its the name I'm giving to text-driven entity extraction. Which is a long-winded way of saying, you type what you want to copy from a page, and that content is "magiclly" copied into your clipboard. For example, if you want all the phone numbers on a page, just type "phone…
Do yourself a favor. Write down the books you've read. Otherwise, some day you'll find yourself wondering if that book looks familiar because you put it on a list to read once upon a time or because you read it once upon a time. Unfortunately, I've never had such a list, and now I'm starting to question things. So, today I created a set of LIT Prompts to help me engage in some digital archeology.…
More than a year ago, I started having my students provide weekly reflections written with the aid of AI. It worked muck like our text expanding Q&A template, stiching together a short journal entry based on the answers provided to a set of standard questions. I've used some variation of this tool every semester since.
We've played with summarization and simplification of texts. So, why not both in one prompt? Today's prompt will take a webpage and summarize it three times. The first summary is aimed at a 12 year old, the second at a general audience, and the third at an expert. Having long enjoyed Wired's 5 levels of difficulty videos, I really appreciate such leveled presentations for their ability to meet…
I hope folks are adding these daily templates to a growing list of LIT Prompts. I find myself reaching for various prompt templates throughout the day. That being said, today we cover a template that really should be in your collection. You may have thought to add it yourself, or it may have been so obvious that you overlooked it. I'm talking about a template for chatting with an LLM. That's…
Today's template will let you summon a rhetorical sparring partner, a devil's advocate intent on challenging your assumptions. As Damien Patrick Williams has noted, the discussion of AI finds itself infused with religious language. “Religious perspectives, myth, and magic are not merely evocative lenses by which to understand the work done by algorithms and 'AI' in the present day— though they are…
One of my favorite bits of writing advice is, "let your writing breathe." That is, don't write everything in one go. Let it sit so you can come back to it with fresh eyes. It's easy to reread something you just wrote and unconsciously fill in the blanks. If you want a different perspective, you need distance or another person. Well, maybe now there's a third option. ;) Like me, Quinten Steenhuis,…
My wife and I just finished editing our first book together—Ben's Frog. It's six pages long, what reading teachers call a decodable book. Think, "See Spot run." They help beginning readers practice decoding strategies for specific grapheme to phoneme patterns (e.g., how the letters "ay" make the sound /ā/). They provide readers with "books" that match what they're learning as they go, helping them…
Yes, this week's theme seems to be talking to robot writing assistants while drinking coffee at a cafe. Well, that and the idea that you should feed more words into a large language model (LLM) than you ask it to output. We do this to mitigate hallucinations along with other issues. We will continue this approach with today's template, The Elements of Interactive Style Guides. Here's how it works,…
Instead of feeding 5 words to a large language model (LLM) and expecting 500 (e.g., write me an essay discussing the lessons of the French Revolution), more folks should be feeding their AIs 500 words and asking them to generate 5. This funneling approach tends to mitigate hallucinations and the biases that creepin when LLM's free associate. If you're feeding in more than you expect to get out,…
Yesterday we used AI to shorten text. Today, however, we'll write a template we can use to expand text. That's right, we'll select one or two sentences and turn them into a paragraph. Here it's worth remembering some of what I said in that series' first post. I observed that I saw a lot of folks wanting to feed 5 words to AI and get out 500 (e.g., write me an essay discussing the lessons of the…
Confronted with a character limit, have you ever found your reply too long? If so, today's prompt template is for you. Compress your writing with the click of a button. No longer must we sign off with "I would have written a shorter email, but I did not have the time." Eat your heart out Pascal!
Recently, a company that is decidedly not DC Comics managed to acquire a trademark registration for "Harleen Quinzel" for use with various types of clothing goods. You may not recognize this name off the bat. However, do you perhaps know the name Harley Quinn? The famous DC Comics villain, whose origin story is closely tied to the Joker, has appeared in 3 major motion pictures in the last 8 years…
To the extent this week's templates have shared a theme, it has been analyze and structure. So, it was only a matter of time before we ran into a use case built around sentiment analysis. Today, we'll build a template that allows you to select the text of several social media posts and label each posts' sentiment while creating a JSON object to hold data about each post, including this new…
I'm not an AI doomer in the sense that I don't think much about p(doom)—the probability that AI will end humanity. I'm much more worried about megacorps exploiting AI for profit at the expense of the marginalized. Which is to say, I'm more concerned about the world of RoboCop than I am of the Terminator. I am, however, sympathetic to those who worry that AI-generated misinformation could…
Functionally, today's template, which creates a short summary of a webpage for sharing on social media, is a special-use adaptation of our very first prompt template. However, I would like to use this familiar task to explore two new ideas. First, the fact that you can create special-use versions of existing templates, and second, how one can use the CHAT behavior to do more than question content.
Over the course of this series I've come to embrace training simulations as a promising use case for large language models (LLMs). Such use recasts many of an LLM's traditional weaknesses as strengths. A tendency to make things up is after all an asset when performing improv. When I introduced simple simulations, I promised we'd eventually get to more advanced ones. See What if Members of Talking…
Yesterday, inspired by GPT-4 passing the bar exam, we tried getting an LLM to some reading questions about Hawkins v. McGee. The LLMs didn't do so hot, only getting 40-47% of the questions right. So, what was missing? Remember, way back in the first post where we said, "If you want an LLM, or the man on the street, to summarize a text, have them read it first." Well, it turns out, if you want an…
You may have heard that GPT-4 passed the bar exam, and though I think this fact tells us more about the bar exam than GPT-4, it's worth thinking about how it does this. As we've noted, large language models (LLMs) are sentence completion machines. They guess the next word based on what they've seen in their training data. They can answer questions because their training data contains questions and…
When counting to 50 even a super-short post which replaces "multiple choice" with "true or false" counts as a post, true or false? The answer is TRUE! This week started with a barn-burner in which I wrestled with how best educators should work to address a changing world, and yesterday I showed off how you can use LLMs as a time-saver when creating multiple choice questions. Today, we're taking on…
I began my professional career as a high school teacher, and I'm a big fan of formative assessment. So, when I returned to the classroom as a student for law school, I found its nearly exclusive reliance on summative assessments not only frustrating but perplexing. Such an inballance goes against nearly everything I learned during my Master's in Education. Now, I don't think a prompt template that…
Last week I couldn't escape the story of Bradley Neal, a 3L who has built a gen AI tool that creates case briefs and answers questions about cases. I love the story, and I love the tool. So much so I spent the weekend recreating most of its functionality. Like Mr. Neal's tool, mine can create case briefs, explain cases in plain language, and directly answer questions about a case. These are common…
As you know, we can seed a prompt with today's date to interesting effect, but whereas last week we used the date to produce poetry, this week we're looking to learn some fun "facts." I've placed "facts" in scare quotes because as we know, large language models (LLMs) are prone to hallucinations. Below I'll share a prompt that asks an LLM to share a significant historical event from this day in…
It's worth taking at least one day to explore the use of the LIT Prompts extension minus the use of an LLM. As it turns out, in making a browser extension that produces templatized reusable LLM prompts, I first had to create a tool capable of making standard fill-in-the-blank templates. We can use both predefined variables and user inputs, meaning these templates can get pretty sophisticated.…
The variety of human language has long been cast as an impediment to communication rather than a laboratory of descriptive experiments, each providing a unique vantage point on the human condition. In fairness, however, absent translation, the insights gained from such experiments would remain inaccessible to all but a fraction of humanity. The fear that insights from the speakers of one language…
A good deal of social science, and empirical legal work for that matter, relies on the parsing of large texts. Say you want to explore how often a court sides with the "government" over "industry," and more to the point, when and how much money is involved in such cases. Unless someone else has done the work, there isn't a spreadsheet you can consult. Someone will have to read through all of that…
As an experiential educator, I firmly believe experience is an excellent teacher. The question I have to answer is how best to introduce students to that experience. Some of my favorite experiences as a law student centered around simulations, practicing direct and cross-examination with actors, negotiating on the behalf of non-existent clients. When it's impractical, unethical, or dangerous to…
I'm sitting down to write this post the day before it will go live, but before I do so, I check the time. "In February's chill, at four twenty-two,/ Thursday's dusk whispers winter's adieu." This was generated by today's prompt template at 4:22pm on Thursday February 1st. It's a simple prompt, using the month, day, and time to write a short pome. I rather like it. As we've seen before, the LIT…
To be clear, the above title is a joke. It's also clickbait. ;) There isn't one trick to writing a great headline partly because folks can't agree on what makes a great headline. Is a great headline one that grabs attention at all cost (unsavory publishers) or one that properly sets expectations (mild-mannered readers)? Either way, I'll show you how to us an LLM to produce headlines based on a…
I'm a big fan of lists. If there's one productivity hack worth its salt, it's the list. Create a list of discrete tasks, and check them off as you go. It's simple and it works, and although people joke about their inboxes serving as to-do lists, let's face the fact that a lot of email authors don't always put their asks front and center. Given LLMs' facility with words, I think there's something…
It has been observed that the rise of search engine optimization (SEO) has lead online publishers to bury recipes and other useful nuggets of information below a pile of what looks to be lengthy essays. The idea being that such keyword-rich long-form articles will rises to the top of search results, driving traffic and ad revenue. The problem with this is that you and I dear reader are forced to…
If you want to start chatting with my papers, click here. I prepaid for some "AI" time. So, chat away while the getting is good. Otherwise, let's take a moment to talk about how we got here. Last week we kicked things off by summarizing and questioning webpages. We closed the week by showing you how to export our templates to create a free-standing web app/webpage. Today, we'll take these add in…
This week we've used LLMs to talk with texts, reply to emails, and figure out unfamiliar phraseologywith input provided by the browser. In future posts, we'll tackle more complex prompt patterns, but we'll also take advantage of LIT Prompt's virtual dice and export functionality to make some really interesting interactions you can share or save for later. Alas, before we run, we must walk, and…
If I had a motto, it might be "context matters." Except, I'm resistant to picking a singular guiding principle because, well... context matters. 😉 Earlier in the week, we wrote a template to provide definitions for words we highlighted in the browser, but we found when kicking the tires that sometimes this feel down when dealing with terms of art or initialisms claimed by competing collections of…
With apologies to Kubrick and Clarke, at least I can say I held out two full days before alluding to a killer AI. That's also how long I resisted sharing a use case in which we use an LLM to help write something. There's a great deal to be said about authorship in the age of ChatGPT, and perhaps we'll explore this more in future posts. For the moment, however, let me provide a framing I find…
I Google the definitions of words with surprising frequency, mostly as part of my writing process. As a dyslexic, it's mostly a backstop for spellcheck. After it has reassured me that I'm only using real words, I tend to use text-to-speech to read what I've written. If a word sounds off, I look it up to make sure it's really the word I intended. It's also good for double checking homophones (e.g.,…
I've started seeing variations on the following, "Sure, I played with ChatGPT when it came out, but I don't really get what the big deal is. You can't trust what it tells you, and it's a pretty mediocre writer." These are valid criticisms, but if you stop there, it's clear you've only experienced a narrow set of what these tools can do. When I speak of "these tools" I'm referring to a class of…
Every weekday for the next 10 weeks, I'll update this page with new prompt patterns and invite folks to play with them inside LIT Prompts. We'll be plugging into the technology that runs tools like ChatGPT, giving you a level of control that most people don't see. You won't have to do any "coding," but we will think like coders. I'm still suspicious of the term prompt engineering, esp. when used…