#31 Dallas Start-Up Week AI
This week, I share what I learned from attending an AI panel discussion at the Dallas Start-up Week conference.
AI will change the world, but how will it change M&A? I want to focus on AI’s impact on M&A in this newsletter. I am not an expert on either M&A or AI, but I want to learn about both topics and how they intersect. I thought there might be others in my situation (or people who are experts in one field or the other) who would find information on M&A and AI helpful in their careers, so I created this newsletter to track and share what I learn.
Takeaways From Dallas Start-up Week
The DSW conference is a small and free conference for Dallas-based start-up founders, so I did not know what to expect. Upon reading the scheduled events, I saw that AI would be the topic of multiple discussions at the conference, so I thought I would attend.
The AI speakers and vendors were all placed in one small conference room. I am not sure why, as the room was packed! See the picture above.
The event organizers probably thought that the AI buzz was not as big as it was last year, where a person could not go anywhere without seeing something related to AI. Last week’s post backed up this assumption, as some professionals are now doubting AI’s capabilities. For this reason, they decided to throw the AI things in one small room, thinking that it would not be the main attraction for the conference.
The organizers were incorrect about the attractiveness of AI. The AI room overflowed for the entire time I was there. There were no empty seats and several dozen people standing in the room listening to the presentations.
I think there are some interesting conclusions about the state of AI that we can take from this.
What does this mean for AI in general? I think it is a clear indication that people are still interested in AI and AI is still a hot topic in certain circles (start-up founders, in the case of the DSW attendees).
One funny anecdote from the conference that I think explains the current state of AI: The presenter began his presentation by asking the audience to raise their hand if they had used ChatGPT. Of course, almost all of the hands in the room went up. Next, the presenter asked for a show of hands of who knew what “GPT” stands for. As you might have guessed, very few people in the room raised their hands. For what it’s worth, GPT stands for “Generative Pre-trained Transformer.”
What does this say about the state of AI? It says that people are familiar with AI and its capabilities, but have not quite caught up with the technical details. I think I’m in the same boat on this front. I have no technical AI training and have taken a "learn as I go” approach to AI. It also says how many people are curious about AI and have made at least some effort to see how it works.
AI Successes and Challenges Panel
The AI session that I attended focused on the successes and challenges that the panelists had with AI in their respective roles. I will give a summary of their responses below.
The panelists gave an idea of how their respective institutions use and are planning to use AI. The uses included the following:
Training an AI model on HR docs to allow users to interact with HR documents on a question-and-answer basis;
Coaching salespeople by having an AI model track sales interactions and provide performance tips; and
Lowering the cost of creating and training LLMs by using a teacher-student method where a small, expert model (the teacher) “checks” or “grades” the responses of a new medium-sized model (the student). This strategy gets the medium-sized model close to the accuracy and performance of an LLM without the high costs associated with creating an LLM.
It seemed like all of the panelists were still in the beginning stages of implementing
AI into their business or professions. Here are the challenges associated with their implementation of AI:
Identifying the problem and the solution: One of the most interesting challenges associated with implementing AI is identifying a problem and a solution that AI can solve. If an organization fails to identify a specific problem and a desired outcome, AI is unlikely to help. I think this is excellent advice. For AI to be effective and not just “cool”, organizations must be specific and intentional with how they use AI.
Types of problems that AI can handle: Certain problems are particularly well-suited for AI disruption. The panel mentioned that problems that are solvable by AI that occur with high frequency are ideal. Why? Because with repeated applications, AI’s outputs will improve over time. Low-frequency problems are bad for AI because there is not enough “practice” for the model to improve its outputs. So, the ideal problems are (1) solvable by generative AI and (2) occur at a high frequency. What challenges do you have that fit these two criteria? The panelists suggested finding a solution that will save 10-minutes a day.
Accuracy: Specifically regarding the application of AI to HR documents, what happens if the AI gives the wrong answer? Is that legally binding on the company? I do not know the answer to that question, but apparently a lawyer advised a panelist that any wrong answers from the HR AI bot could cause serious legal liabilities for the company. We have talked a lot about the accuracy of generative AI, but none of the methods offer 100% accuracy, and it’s unclear whether AI will ever achieve such a milestone.
Depreciation: The panelists pointed out that the longer a model is in use, the worse it becomes. Like a new car when driven off the lot, when AI models are put into use, they immediately begin to depreciate and lose value because the training data becomes old. This is particularly true when AI technology is progressing as fast as it is now. This made it clear to me that achieving one AI solution is not final—we must continually innovate to ensure our AI solution remains viable.
About me
My name is Parker Lawter, and I am a law student pursuing a career as an M&A lawyer. I am in my last semester of law school, and with some extra time on my hands, I decided to create this newsletter. I hope it is informative and helpful to anyone who reads it! I am not an expert at either M&A or AI, but I am actively pursuing knowledge in both areas, and this newsletter is a part of that pursuit. I hope you’ll join me!
Follow me on LinkedIn: www.linkedin.com/in/parker-w-lawter-58a6a41b
All views expressed are my own!