Contestants on the gameshow Who Wants to Be a Millionaire? get three different lifelines to help them navigate through the labyrinth of trivia questions: an "Ask the Audience" to poll the audience for their thoughts on a question, a "50:50" to get the game moderator to eliminate two of the four multiple choice options for a question, and a "Phone-a-Friend" where contestants can call up a buddy to get their thoughts on what the right answer to a question is.
ChatGPT is not the next Google. ChatGPT is the next "Phone-a-Friend."
As I wrote about a few days ago, I think ChatGPT is easily one of the coolest pieces of tech released in the last decade and finally gives a glimpse at what all these billions of dollars in VC funding for AI are going towards. That being said, I've seen people online predicting that ChatGPT could spell the end of anything from Google to the entire field of management consulting, and I think a lot of what is being said is BS.
For starters, using ChatGPT for five minutes and immediately predicting the fall of Google is fundamentally ignoring what Google is as a product. ChatGPT casts a small net. You ask it a very specific question, and it responds with an impressively specific answer. But it's not like people only go to Google for very specific questions. People go to Google to cast a wide net and get a broad pulse for what is being said about a topic, and that's a much different use case. ChatGPT is very informative, but Google is very exploratory. Now, I expect that the Google team is sweating pretty heavily right now using this ChatGPT demo, and they'd be idiots for not incorporating some of what's good about ChatGPT's AI into their search page, but for now, the reports of Google's death are greatly exaggerated.
Nor do I think the AI takedown of the consulting industry is going to pan out any time soon. Perhaps I'm mistaken, but I don't think consultants are selling knowledge; they're selling trust. And even if the answer ChatGPT would give to a complex business problem matches what a consultant would say almost identically, I don't think there's enough trust built out in AI yet for it to realistically compete with the consulting industry. We'll see if that changes in ten years.
I think ChatGPT is going to turn into an amazing resource for putting out feelers on a question and getting the first part of a brainstorm started. In other words, it's not a consultant where you immediately adopt whatever recommendation they give; it's more your buddy that you casually explain a situation to and ask for their thoughts just to hear another perspective that you trust. It's a "Phone-a-Friend," except you get an unlimited number of them. ChatGPT is going to fill that role for now. AI tools will still be viewed as novelties for at least the next few years, and we're not at a point yet where ChatGPT can handle all the nuances of a specific case study. But much like how I argued AI art generators will be a tool for creators rather than their replacement, I think ChatGPT will be an amazing tool for decision makers to use rather than the new decision makers. Having ChatGPT in your pocket at all times is like always having that "Phone-a-Friend" lifeline on you, and that friend you can call just happens to have the entire internet's knowledge tucked away in their brain. You're always able to ask a very specific question and instantly get a range of pretty good ideas to start your own decision-making with, and I'm excited to see ChatGPT continue to grow in this capacity.
Comments