TJ Walker and Chris argue that generative AI like ChatGPT is overhyped and risks a crash similar to the dot-com bubble. Companies like Starbucks and Netflix have used other machine-learning tools for years. The real value comes from balancing generative AI with personalization, automated decision-making, and guardrails that know when to hand off to humans.
Chapters
They actually have predicted that there's probably going to be an AI crash similar to the dot com crash, because investors can't keep up with the demand and the hype of where AI is going. But again, if we look at AI broadly, Starbucks has been using AI for the benefit of the consumer for 10 years. Netflix invented the concept of user preferences and customization and personalization through AI in their algorithms.
You like this. Our AI says, you're probably going to like this.
And that's been around for 20 years. All of a sudden generative AI is in the news and OpenAI and ChatGPT, and everybody's thinking, well, I've got to use that. I hate to break it to people, but generative AI is not going to solve everything that you want it to solve. We have to figure out how to balance the generative piece of this and the other machine learning and AI elements that are actually beneficial to the consumer. Again, for example, how do I hand off to a real person versus automate this experience for the consumer based on their individual preferences? That's not generative AI. That is actually an automated learning process by which I understand your unique personality traits, your behavioral instincts, your aspirations. Translate that into a recommendation of, do I automate this or do I hand it off? Another great example of automation, when we think about AI, is those guardrails, what I was talking about earlier. The guardrails are. How many CFOs in this world do you think would accept an automatic, automatic extension for all missed payments? Almost none. Because they don't want to take that risk of saying, well, wait a second, I just authorized 100% of the people that are asking for extensions. I allowed an AI to make that decision and grant it. But yet, that's a great example of how do we automate? Look, I call in, I'm crying over the phone because I was in an accident. I can't make my payment. I've lost this, I've done this, and I. I need some sort of help. Can an AI sympathize, empathize with that caller, recognize the emotional levels? Yes, to an extent. And then grant that extension because they're asking for it with the guardrails in place to ensure that operationally and financially, the company is still in good position. Those types of complex questions have to be answered for this to be successful.
Chris, let me ask you about who I would assume is your favorite client in the world, who does everything you ask for your own company. You've been in the AI space for a quarter of a century. How do you use personally artificial intelligence to help you with your own business? You publish books, you do trainings, you have workshops, you have consulting. How are you using it in a way that you really think embodies best practices? Otherwise, why else would you do it?
I'm going to call out a company, Grammarly, and. And I'm sure a lot of people use Grammarly, especially if you're doing any type of public speaking and just want to make sure words are proper and especially writing. What's fascinating to me about Grammarly is that Grammarly has been an AI company for since their inception. That's the whole point.
