Nexus: welcome to the age of AI

February 26, 2025 4 minutes

Recently, large AI companies iterate their newest AI models almost every day. Grok rolled on Grok-3, Claude released the so-called strongest physical calculation AI model sonnet 3.7, and DeepSeek has its name remembered for its breakthrough open-source AI model. It seems like all technology companies join the AI races and everyone ends up having a dozen of free or charged LLMs at hand. Has the age of AI arrived? Yes, it is.

In this scenario, I’ve just finished reading the book Nexus: A Brief History of Information Networks from the Stone Age to AI written by Israeli historian Yuval Noah Harari. This book prompts me to think wider and deeper about AI.

There are some important and interesting points in this book. After introducing those points, I’d like to talk about my feelings of the challenges AI brings to us.

What matters in Nexus

AI influences humans by making up myths

As the author mentioned in his early books, human beings are skilled at making up myths and stories, through which hundreds of thousands of people connect with each other and build a huge network.

The myths are not truth. We know the earth rotates around the sun is a truth. But money is not a truth, for money are based on the trust of people. Trust? It’s intersubjective. Our world is built on the foundation of myths, such as country, law, and credit.

Now AI models gain more and more power, generating texts, images, music and videos. People can’t even identify if they’re true or not. AI could also make up myths and stories, connecting with people and gaining trust. If we read the stories AI made, listened to music AI created and used the financial currency AI invented day after day, we would be influenced by AI absolutely.

AI is fallible

I hope AI is infallible, but it’s not. Infant AI learns from the materials collected from fallible human beings, and is fine-tuned by scientists with bias. I sent a prompt Write a small story about a faked spaceship to an AI model a few times, and the AI model told me a story about a man captain each time. There are smaller women captains in fact, but AI has took the bias for granted and may deepen the hidden bias with its powerful ability.

When it comes to hierarchy, racism or other important points, could AI make the right decisions to be justice or responsible? Superhero movies always say: With great power comes great responsibility. Can AI take its social responsibility as it has no feelings like humans?

The author thought a self-correction mechanism would help. This mechanism, such as peer-to-peer review, has boosted the prosperity of science. The European union is drafting laws to supervise the behavior of AI, but I’m still skeptical of its possibility of understanding and limiting the behavior of AI. AI behavior is more like a black box for humans. How fast could we find out if an AI is doing a bad thing? or could we prohibit the bad guys from exploiting AI?

Challenges

AI brings numerous challenges to us. When it comes to careers, will the software engineers be redundant as AI models have the ability to write clean code? The profession of software engineer showed up in the 20th century, and played an important role in Internet development. At the same time, some jobs disappeared, such as typist. Some people think AI only disappears the jobs but doesn’t create new jobs, leaving a lot of people out of work. However, we have bankers since we invented various kinds of money, and have game producers because of our desire to play games online. So how about inventing some new intersubjective things and create new jobs?

Another concern is about the fear of AI wisdom. The soap opera The Orville told about a story of a spaceship which ran exploration tasks in the future. There was a special crew member named Isaac, who came from an advanced galaxy Kaylon. Isaac was a robot, extraordinarily smart compared with humans, and kept protecting the crew on the ship because Isaac tuned its algorithm to meet this goal. The funny opera imagined the future positively. Getting accustomed to AI coworkers could be normal in the future.

We are unable to predict the future. All we could do is to adapt to new circumstances and try our best to live. Future people driving a spaceship to work also have their fear and confusion.

Will the world become better with AI? I have no idea. Future has another name—uncertainty. Anyway, welcome to the age of AI.

P.S. This blog is written by a human person completely. 😊

BookNotes