The Generative AI newsletter for legal pros everywhere.

🚀 Hello! We have reached 680+ subscribers in 21 days! I cannot thank you enough for subscribing!! 🎉🔥

NEW SUBSCRIBERS, to read previous posts, go here.

My goal is to reach 1000 subscribers. Do you think I can make it? With your help, I can, so please share this newsletter with as many people as you can if you are finding it useful.

This is day 22 of 100 (our goal - 100 consecutive days!) and today we will:

1. ask you about your ideas, tips, and input for issues 30-39

2. explain the power of priming ChatGPT

3. draft an SOW I will be using

4. walk thru Gamma.app - an impressive slide deck tool

5. talk news you can use and lose

👀 Looking ahead

As we approach the 30th to 39th consecutive editions, I want to bring you in more.
I plan to touch on things such as:

• Google Bard

• Bing Chat

• Slide Creation

• Image generation DALL-E

• Content products for client development

• Insights as a Service

• Creating tables & charts

• Real-Time daily hacks

• Using ChatGPT to Inform Decisions

• And much more...

What would you like to learn about these or other topics?

Have any examples, predictions, opportunities, key lessons, or links?
Reply to this email with them. 
📣 I will be giving shout-outs to everyone that helps.

Ok, let's kick this off, shall we?

TOPIC: Priming

PROBLEM: We have touched on priming before, but I want to hit it more deeply today. I hear from many beginners that their ChatGPT replies tend to look the same after a while or are generic/general or just not what they are looking for.

SOLUTION: Sometimes you need to prime the chat session so that ChatGPT has a better grasp of what you are talking about.

What is Priming? Think of priming as creating a shared experience between you and ChatGPT. Instead of coming in cold to a blank chat session, you can warm it up a bit by setting some context or priming it. ChatGPT will use this information as a reference point for the rest of the chat session.

Let’s show you what I mean.

USE CASE: SOWs

If you put together statements of work or engagement letters you likely know the ritual. You usually pull up one of your previous ones and start editing that based on the potential new engagement/matter/client.

I am sure some of you have your ‘standard’ template. That is great. If you are like me though, many client engagements are heavily nuanced and customized. This makes it a huge challenge to be efficient and effective while not spending literally hours putting it together.

And I have noticed that the longer I am in business the more prone to habit I become. Meaning I might not care all that much about writing a comprehensive SOW... I mean who reads these things anyway? Most clients want to know how much you are charging. Beyond that, they are basing their decision on your prior interactions, your reputation, and what they need that you say you can deliver on.

So we have a new client that we are in SOW drafting and negotiation with now. I figured why not take this opportunity to refresh my SOW skills and drafting?

And you know what? It was so darn helpful. Shocker! Maybe I am just not smart or lazy - or both - but going through this process was a great refresher and actually made drafting the SOW . . . fun? Yes, it was FUN!

Prompt: 

So I will show you an unprimed response first.

To get started I just brain dumped a bunch of key points from our initial high-level proposal and subsequent talks. I had basically a page and a half in a Word doc. Nothing formatted or numbered. Just literally sentence fragments, random thoughts, and then copy-n-pasted text from our PowerPoint brochure.

▶︎▶︎PROMPT

Can you draft an SOW based on the following partial information? If there are sections of a typical SOW that this information does not address, note them and leave them blank.

[pasted in the text from my Word Doc]

Here is a link to the full prompt and response.

It was less than desirable. It didn’t even use a recognizable SOW framework. I hit regenerate and hoped it might change a bit. It didn’t.

This is because I did not prime it.

Ok, so how did I prime it? Simple. I asked it about how to structure an SOW first.

▶︎▶︎PROMPT

What are the elements and structure of a professional services SOW?

I got back a nice complete framework. You will see that shortly.

I then reran the same exact prompt as I did the first time when I didn’t prime ChatGPT.

▶︎▶︎PROMPT

Can you draft an SOW based on the following partial information? If there are sections of a typical SOW that this information does not address, note them and leave them blank.

[pasted in the text from my Word Doc]

Here is the link to the prime response and then the full SOW response.

🤓 Do you see how the same exact prompt generated two completely different responses? 

That is because I primed the second one.

How you should think about priming ChatGPT?

Before you ask it a question, think of the general area of knowledge your question is set in. Are you asking about geography, organizational design, culinary delights, area of law, type of contract, etc.? Just ask it to explain the basics or something particular in that domain that your answer will likely draw from.

Then after it replies, review it to make sure it looks like it knows what you are going to talk about.

If not, reprime it.

If so, proceed with your question/request/prompt.

Today’s tool is Gamma.app - An AI slide deck creation assistant

Its tagline is “A new medium for presenting ideas. Powered by AI”

What it does? You start with a prompt window (similar to ChatGPT) and you put in what you want your slide deck to be on. It then literally takes that prompt and creates an entire deck - complete with relevant text, sections, layouts, and pictures.

I have seen many of these and this is one of the better ones which is why I am sharing it with you all.

Here is my <5min video of me walking you through it.

News you can use: Law student turns AI entrepreneur!
This is not one of my students but it is exactly the type of moxy, ambition, and hustle I preach to my students.

They are worth a follow and I will be having them as a guest on an upcoming podcast. Stay tuned.

News you can lose : ALL CHATGPT DEVELOPMENT MUST STOP!!

In a letter that has been circulating in the press, 1000s of tech leaders, scientists, and AI experts are calling for a ChatGPT TIME OUT. While I can appreciate the motives and concerns, aren’t the genies out of the bottles, horses out of the barn, and toothpaste out of the tube . .. well you get it.

Wall of Feedback

Landscaping? Who knew? Love this. 🤟👊

With your feedback, we can improve the letter. Click on a link to vote.
If you click on Meh or Getting bored - please leave me a note so I can improve things for you:

• 👍 This helped me. Thanks

• 😐 Meh - was ok.

• 👎 Getting bored.

Build your own? 👉 FeedLetter.co

That's a wrap for today. Stay thirsty & see ya next time! If you want more, be sure to follow me on Twitter and LinkedIn.

DISCLAIMER: None of this is legal advice. This newsletter is strictly educational and is not legal advice or a solicitation to buy or sell any assets or to make any legal decisions. Please be careful and do your own research.