Hey, I’m Lindsay, the host of Make Sense w/ Lindsay T., Lady Engineer®, a weekly podcast where my guests and I simplify complex topics at the intersection of people & technology. We analyze whatever hot mess is in the news, evaluate trending innovations through a human-first lens, and take deep dives on a need-to-know basis.
Sign up to get weekly updates below, and I promise to keep it simple…in a way only a Lady Engineer® can!
If you’re a business consultant working with enterprise customers, you want to have an informed opinion on Enterprise AI. This week, my guest Joe Meersman and I help you develop a more nuanced understanding without drowning you in the intricacies of coding and complex large language models. Joe has spent over 25 years working in user experience and creative design and worked with IBM Watson on their AI initiatives in the early 2010s. There’s no one better to join me in helping you!
What is Enterprise IT?
Enterprise IT covers the use of information technology within large organizations – think Fortune 1000 companies. Enterprise IT systems include various software – Enterprise Resource Planning, Customer Relationship Management, Business Intelligence, Communication tools, etc. These are the tools that power the employee and customer experiences.
Enterprise IT is a Tough Problem
When 55%+ enterprise IT projects fail, I feel super apprehensive around ‘Enterprise AI’ discussions. Failures not only impact bottom-lines but also the livelihoods and mental health of all of the employees involved. Sixty-two percent of the US workforce works white-collar jobs. That’s a ton of people who have to use enterprise software every day.
That’s why I frequently ask, “Why does software at work suck?” I’d like us to stop asking how we can innovate in fields like Enterprise Resource Planning and Cybersecurity because we need something simpler: Software that reliably accomplishes its intended tasks, aiding workers in doing their jobs without unnecessary complexity.
This is my dream, yet I know the reality is that the enterprise pursuit of AI will not slow down. Executives, driven by fear of falling behind, will prioritize AI initiatives without fully understanding its implications or the problems it will solve or cause.
Yet, we can do better by maintaining a grounded perspective. Rather than succumbing to the allure of the technological buzzwords, organizations must first identify the problems to solve (a.k.a. Jobs to Be Done) and craft straightforward solutions. Only then should they consider using AI, recognizing it as a tool, not a panacea for all challenges.
So if you are invested in the success of such projects, remain informed and build your problem-solving chops. By approaching AI adoption with pragmatism and foresight, we can mitigate the risks of failures and ensure that our efforts yield tangible benefits.
In this week’s episode of Make Sense, we also chat:
➡️ Big companies’ initiatives to adopt AI including opportunities, challenges, and pitfalls to consider in enterprise adoption
➡️ The two schools of thought on AI development: Do nothing vs. isolated training
➡️ Training data and need for user interaction to boost the performance of behavior of AI tools
Reference the full show notes below.
Choice Quote
“An enterprise may say ‘You Can use GPT-4,’ which is great for those that are inclined to use the technology or who may have been using the technology anyway. What’s just as important is providing someone with a body of knowledge around how to utilize the tool.
– Joe Meersman, Creative Design Director & Business Consultant
Buck the media hype cycle. Calm the fear-mongering. Laugh at the inanity of Tech CEO “hero culture.” Be the smartest person in your peer’s LinkedIn feed:
Key Takeaways with Joe Meersman
For this episode with Joe, I pitched him the three predictions I made for this year. Here are Joe’s takes on them in the segment Crystal Ball: What does the future hold?
-
Many Generative AI Companies Will Fold: Yes
- Politicians will Blame AI for Election Results: Yes
-
One major copyright case against AI will change its trajectory: Sort-of. Listen to his nuanced take at minute marker: 11:35
Show Notes
00:00 Introduction
03:01 Crystal Ball: Predictions for AI
08:16 AI and Election Results
11:23 Copyright Cases and AI
18:35 Trough of Disillusionment
23:15 AI Still Needs a Human Chaperone
29:01 Debate on AI Awareness and Consciousness
30:15 Two Schools of Thought on AI Development
31:16 Control and Understanding of AI
32:13 AI’s Ability to Connect the Dots
33:40 Garbage In, Garbage Out
34:07 Teaching and Demonstrating with AI
35:27 Inputs, Transforms, and Outputs
36:21 Training Data and User Interaction
39:00 AI’s Prone to Hallucination
42:58 Problem Solving and Digital Transformation
Terminology Used in this Podcast
Moravec’s Paradox (~24 min mark): An AI concept that what is easy for humans (like perception skills) is difficult for machines, and what is easy for machines (like reasoning) is difficult for humans.
Turing Test (~29 min mark):A test of a machine’s ability to exhibit intelligence that is indistinguishable from that of a human.
Where to find Joe Meersman
Joe is a Partner at The Collective and an Adjunct Professor at The University of Texas at Austin where he teaches Visual Communication at the School of Design and Creative Technologies. He brings over 25 years of user experience and creative design expertise, and has worked at some big brands like Motorola, Razorfish and IBM
LinkedIn: https://www.linkedin.com/in/meersman/
Made by the Collective on Linkedin: https://www.linkedin.com/company/made-by-the-collective/
Adjunct Professor Page at The University of Texas at Austin: https://designcreativetech.utexas.edu/joe-meersman
Want to learn more about Enterprise AI and AI in general? Here are Joe’s recommendations:
- The AI Breakdown by Nathaniel Whittemore: https://theaibreakdown.beehiiv.com/
- Ethan Mollick and Azeem Azar.
Where to find Lindsay Tabas
LinkedIn: https://www.linkedin.com/in/lindsaytabas
YouTube: https://www.youtube.com/@LindsayTLadyEngineer
Instagram: https://www.instagram.com/lindsaytladyengineer
New to Make Sense?
Get Caught Up with the Top 10 Clips from Make Sense’s First Year
Make Sense is a video-first podcast dedicated to simplifying complex issues at the intersection of people and technology. There are many!
Access our Top 10 Clips from the first 30 episodes and get caught up with the latest tech predictions, headlines, and discussion points via accessible and easy-to-understand discussions with innovators across climate, HR, finance, connectivity, and healthcare.
Continue to be the smartest person in the room with this mobile-friendly list:
