ChatGPT: Rise of the Machines
The colossal impact of the recent developments in large language models - the major tech of ChatGPT, Midjourney, etc. is still not known yet. It is not a cute toy anymore; it will change the working process.
Let's look into the economic perspective in two sections:
1. Price drops are much heavier in digital services and computers
We have seen many price drops throughout the last decades by optimizing machines and cheaper operations (fair/unfair) —fast fashion, fake meat, and cars. The list is long. But the decrease is between 10-50%. We wouldn't expect that the clothes will cost 100 times cheaper now.
The digital world is different, including the computer/electronics industry. If you can remember the 90's, an 8 MB RAM was ~250$. And now an 8 GB RAM costs $25. The price per megabyte dropped 1024X in 25 years! If we apply that scale of price decrease to a pair of shoes that cost $50 in the 90s, they would cost 5 cents now. Sounds ridiculous, right? AI is going to make this ridiculous change, and this is going to happen, mostly in services. To understand it better, let's look into the second aspect.
2. A system solution and not a point solution
A point solution is that you change a process in your entire value chain and get improvement in your production or services. For example, a manufacturer replaces a manual process with a robot. Or a retailer enhances their customer support with an intelligent AI-powered chatbot, saving many hours of staff time. The operational workflow of the company doesn't change, but some points are changing.
But a system solution is changing the value chain. It happens rarely, but when it happens, it almost closes one age and starts a new one. Let's look into some system solutions:
18th century: The steam engine (cheaper and faster production)
'20s: Electric motor (cheaper energy, machines can be distributed)
领英推荐
'90s: Computers were in most industries a point solution except in financial and academic institutions, but at the end of the 1990s and beginning 2000s, computers changed many industries, like ads, entertainment, insurance, architecture, and healthcare. A GPS not only eradicated the taxi training programs but, with AI prediction, created an entirely new category of transportation — ride-hailing (Uber, Lyft, Bolt, etc.).
2023: AI with LLM (ChatGPT-4, Mdjourney 5) is capable of taking over some services humans have operated for decades. A machine can write, calculate, predict, decide, and build many things in the digital world cheaper and faster than humans.
To stay up to date, and don't miss the train, here are some great handpicked AI resources for you:
Ethan Mollick 's newsletter: https://oneusefulthing.substack.com
Book recommendation: Power and Prediction: The Disruptive Economics of Artificial Intelligence by Avi Goldfarb
People to follow:
AI experts and my fellows from Maven instructors,