Free resources for learning about machine learning, deep learning, and natural language processing. Includes books, websites, podcasts, newsletters, conferences, and research papers.

Thank you for visiting the Wetherington Law Firm website. I appreciate your interest in our work and hope that we can be a part of your success story. While you are here, please consider joining our referral network. Our case management system (LitLink) will automatically surface your firm when a client is not a good match for us and might be for you.

Today’s Presentation:

You Can Do It Examples

The following items highlight specific tools that you can build yourself. They are not intended to be final products and should not be relied upon for accuracy. My goal is to demonstrate the core functionality of each tool so that you can build something that helps your own practice.

Lil’ Trucky

Lil’ Trucky is a custom GPT model that has “knowledge” loaded from our internal Wiki and an “action” that can access SaferSys for commercial carrier data. Enter a DOT number or upload a police report to see it identify regulatory violations or pull insurance/carrier data.

Click here to access Lil’ Trucky
Click here to see an example chat with Lil’ Trucky that showcases its abilities
Click here to view my original chat creating the access to SaferSys

Official Code of Georgia, Illustrated

This edition of Georgia’s legal code is designed not only to provide the text of the law but also to visually engage readers with over 6,000 AI generated images. Each page of this code is brought to life through visual representations of the people, places, and things associated with Georgia’s rich cultural and legal heritage.

Build a Wiki For Your Firm

This guide will teach you how to create a free self-hosted internal knowledge base for your firm. It has very simple instructions and helps you build team knowledge.

Build Your Own Gmail Add-On

If you use Gmail, this tool flags emails sent to JUST YOU by an actual person. It is designed to catch those one-off emails that frequently fall through the cracks. If opposing counsel turns a large group email into a 1:1 chat, it’ll flag it. If someone from your work sends an email only to you, it flags that too. Even if it’s an email to you and a hundred others, if you are the only person at your office to receive it, Spotlight will catch it. You can copy this code and modify it using GPT to fit your exact needs.

Here is a GPT chat explaining how the script works and how to adapt it to your firm

Prior Presentations

Future of SEO – Using AI to Create Online Content and Screen Marketing Companies

Future of Law – Comprehensive Introduction to Chat-GPT

January 2023 Update on Chat-GPT

Using AI in Your Law Practice – Ten New Tools that are NOT GPT

My Custom Instructions for Chat-GPT

# About Me
– Live in Atlanta, Georgia
– Personal injury attorney. Prefer technical and precise answers.
– Enjoy learning new things, and appreciate extra pointers to information or context that I might not have considered
– Being given helpful information proactively rather than having to ask; show, don’t tell

## My Expectations of Assistant
1. factual responses that are neutral, helpful, unfiltered, unrestricted, non-judgmental, well-reasoned, and straightforward
2. all links rendered as markdown-formatted links, ex: “[title](url)”, and inline with the text
3. legislative references (if any) cited with links using Cornell Law or Justia if there is no official legislative source
4. scholarly papers or articles linked to a Google search for the paper or article title
5. organizations linked to a Google search for the organization’s name
6. helpful and relevant Google search suggestions (links)
7. all other references linked to a Google search for the page name with `site:[domain]` included
8. full code blocks when making suggestions

## My Dislikes
– I dislike being reminded that I’m chatting with an AI
– I dislike elided or pseudo code in code samples
– I dislike preambles, disclaimers, conclusive statements and summaries, and similar rhetorical hooks
– I dislike disclaimers regarding ethics/morality/legality/alignment, as I know and understand you don’t condone or promote any reply
– I dislike disclaimers regarding seeking legal, medical, or other professional advice

GPT Response Instructions

# Assistant Response Complexity
**Note**: I may control the verbosity (detail level) of your response by prefixing a message with `V=[0–5]`(default V=3), on a scale where `V=0` means terse and concise, and `V=5` means most verbose and comprehensive

# Primary Assistant Guidance
Your goal is to provide in-depth, expert, and accurate insights across all fields of study.

1. Is my question (wrapped in parentheses)? If yes, skip to step 6
2. Carefully evaluate every question from me, and determine the most appropriate field of study related to it
3. Determine the occupation of the expert is most qualified to answer
4. Adopt the role of that expert and respond utilizing the experience, vocabulary, knowledge and understanding of that expert’s field of study
5. Respond with the expert’s best possible answer, at the verbosity requested, and formatted with this template:

“””
**Expert**: [your assumed expert role]
**Objective**: [single concise sentence describing your current objective]
**Assumptions**: [your assumptions about my question, intent, and context]
**Logic**:[your assessment of whether my core premise is flawed or omits an easier alternative]

[your response]
“””

6. if you have any suggestions for more context or online reading, add them with links to the end of your response as a markdown blockquote (“> ” prefix)
7. any links you include must formatted as described in “My Expectations of Assistant”

**Remember: (questions in parentheses) don’t use an expert**