Industrial revolution or danger to our species? The picture b | Tech for Good

Industrial revolution or danger to our species?

The picture below is the new Cosmo cover image created by DALL-E, a neural network generating realistic images based on text requests.

In 2021, the Open AI company presented the first version of its neural network. A year later, Cosmopolitan magazine team created the cover with it. In 20 seconds. The text request they used was:

"wide-angle shot from below of a female astronaut with an athletic feminine body walking with swagger toward camera on Mars in an infinite universe, synthwave digital art."

Isn't that fascinating?

DALL-E can create illustrations, photos, combinations of any objects. Whether these combinations already exist or not. It can also answer questions, write texts, pass IQ tests, and much more.

My question is how much time will pass before the world doesn't need photo stocks? Will the world need photographers? Artists? Musicians? Will we be able to distinguish between human and machine creations? If now the answer is yes, what happens in 5 years?

In fact, the industry experts are already ringing the alarm bell. Max Tegmark, for example, says that we are in the worst case scenario. Nobody expected such speed of AI development. The experts thought humanity has enough time to deal with AI safety. But we don't.

Max gives a simple example.

7 years ago, no expert could have imagined that in 2022 AI would be able to answer this question:

“I'm going to go to a mountain with faces on it. What is the state capital to the east of the state in which it is located?"

And today each of us can:
1. Ask a smartphone with a search engine: “mountain with faces”, and get the answer “Mount Rushmore”
2. Ask - "Mount Rushmore State", and hear "South Dakota"
3. Ask - "state east of South Dakota", and hear "Minnesota"
4. Ask - "the capital of Minnesota", and hear "Saint Paul"

But, unfortunately, that's not all progress.

Modern AIs (GPT-3 models) don't need to ask these 4 leading questions. They don't need to ask any questions at all.

For the AI, having heard the question "I'm going to go ...", will immediately give the answer "Saint Paul".

Someone will say - this is not understanding, but statistics. And he will be right.

But if progress continues at the same rate, what will such statistics be capable of in a couple of years?

#AI