Artists go on strike and completely panic because realistically, they needed to panic and go on strike. The AI can recreate their work in seconds at no cost. This thing we just invented, a software developer's fever dream and ultimate achievement, may have been us majorly shooting ourselves in the foot.

"We haven't seen anything this game changing since the dawn of the world wide web!"

"This is a more powerful tool than the cellphone!"

Everyday people, and professionals alike let their imaginations run wild. Understandably so, this is the first computer program we can reasonably call an A.I. (artificial intelligence).

But, as the title of this note says, we jumped to a foregone conclusion and forgot 2 key things.

1. This technology is very much still in its infancy, we have no idea what it can do or how close to the ceiling/floor we are.

2. It's so new we have no idea how it will affect industries it is already permanently changing. We should sit back and observe a little before we introduce this admittedly cool and potentially useful thing to central systems like education and overseeing systems with real world consequences if something goes wrong.

"There is no longer a need for any human involvement in the fields of medicine, teaching, software development, and academic research!"

"This thing is gonna found skynet any minute now!"

People also leapt to conclusions in the opposite direction. It is important to acknowledge the valid concerns, and separate them from the ridiculous ones. One big piece of this problem is the name "AI". Pop culture seems to have infected everyone's heads a little here. This problem is exacerbated further by how mysterious computers are to a HUGE number of people. Computers are a total black box to billions, this makes it unclear how many things AI models are capable of that standard programs aren't. Generative Artificial Intelligence is somewhat of a misnomer, and people are certainly confusing the term with sentience. If you made an NLP algorithm (the type of algorithm GPT-3&4 and others are leveraging) take an IQ test, you would find that its pattern recognition is very low.

Yes it can tell you how to synthesize chemicals in a detailed step by step, and that's impressive and maybe a little scary, but it also doesn't know how to play a valid move in chess, would mess up instructions in making an omelette, and probably couldn't even solve the New York Times' connections game if you gave it 1,000 tries.

In other words it desperately needs humans to verify that every little step it takes isn't a mistake that eventually cascades into wildly correct answers even for simple problems.

These are the people forgetting that if we encourage a widespread rollout of AI, everyone who worked hard to design art for hollywood, or a game studio, or for literally any purpose where the final product is digital, they're out of a job.

It means it's too early to make any definitive assertions about the consequences of generative AI, right now the best course of action is to observe how it affects industries it has already invaded.