Ever hear of the Chat Generative Pre-trained Transformer?  It’s better known as ChatGPT- I bet you have heard of that!

Since the public prototype of ChatGPT chatbot was released last November 30th, a steady narrative has erupted on the pros and cons of this emerging application of natural language artificial intelligence (AI) technology.  Part of the fascination is the marvel of some early success, though it has issues with “hallucinations”- as the lingo calls errors or wrong answers.  It’s certainly proven to be popular- within five days of release, one million people were using it.  For comparison, it took Netflix 3 ½ years for a million people to sign up; a reported 100 million people have tried the chatbot.

An example of early success- a study conducted by University of Pennsylvania Wharton School professor Christian Terwiesch testing whether ChatGPT can pass a Wharton MBA course final exam.  The report quotes four main observations:

  • “it does an amazing job at basic operations management and process analysis questions… (n)ot only are the answers correct, but the explanations are excellent”
  • “Chat GPT3 at times makes surprising mistakes in relatively simple calculations at the level of 6th grade Math. These mistakes can be massive in magnitude”
  • “the present version of Chat GPT is not capable of handling more advanced process analysis questions, even when they are based on fairly standard templates. This includes process flows with multiple products and problems with stochastic effects such as demand variability”
  • “ChatGPT3 is remarkably good at modifying its answers in response to human hints. In other words, in the instances where it initially failed to match the problem with the right solution method, Chat GPT3 was able to correct itself after receiving an appropriate hint from a human expert”

Ultimately, the professor concluded the chatbot would pass with a B or B-, which is pretty good for not studying for the exam!

Another success- researchers published a paper detailing how the chatbot passed the U.S. Medical Exam.  While further research including peer review is pending, and limitations are noted, results show ChatGPT ”performed at or near the passing threshold for all three exams without any specialized training or reinforcement,” and “demonstrated a high level of concordance and insight in its explanations.”

ChatGPT Answer
ChatGPT answering a question on what it is

Of course there are limitations… as my colleague Josh Barone mentions in his latest Hidden Truths podcast on the subject (The Secret to ChatGPT), the technology isn’t a search engine retrieving information, rather it uses data gathered to devise natural language answers.  ChatGPT currently only uses data updated through 2021.

AI technology has become part of daily life.  Successful people will learn to use it as a tool to enhance life.  Reach out and tell me what AI interactions you have had- or if it’s something that you haven’t yet directly engaged in.  Click to email me here.

~ Brian Kasal- The Leadership Matrix

P.S.- Did you see my last Leadership Matrix post? Fast Ferry to Mars