Just one Sora, why the rush?

After OpenAI released the Sora model a few days ago, the entire AI community was in an uproar. Many bigwigs expressed their opinions, and even some media started to spread the story of making money. But the author's advice is: don't worry, don't rush.

Well, I’ve been bombarded with technology these past two days.

Did you see it?

Just like last year, Open AI made a splash during the Chinese New Year by launching a technology called "Text-to-Video Generation" and gave it a name, Sora . Since I'm not good at English, I'll just call it "Sora".

I believe you have witnessed its capabilities, and there is a lot of related content on the Internet. I was also quite curious, so I opened my ChatGPT account to check, but I didn’t have the experience permission yet.

Then, many experts, scholars, and founders in vertical fields in my circle of friends came out to talk about it. Some said: It’s amazing, you can generate what you want with just a spell. If you don’t know how to write it, let AI write it for you.

Since I can't experience the video, I'll just experience the magic of commands. I want to make a 15-second TVC for a brand, describing a creative advertisement of a woman holding a certain brand of milk.

I tried GPT, and the answer it gave was:

In the dim morning light, a lady in a light summer dress walks on the soft grass, holding a shiny bottle of branded milk in her hand. She walks gracefully among the blooming wildflowers, the breeze blowing, and the milk bottle sways gently with her steps, capturing every flicker of the sun.

The lady stopped, gently opened the milk, and poured the pure liquid into a transparent goblet. The milk met the morning light, as if blending into the freshness of nature.

The picture slowly moves forward, and she is seen taking a sip of milk, closing her eyes and enjoying this gift from nature. The picture freezes on her satisfied smile, and then the brand logo slowly appears, and the background fades into a warm slogan: "From nature to your table."

At this moment, everything seems so pure and peaceful, as if time has stopped at this moment, leaving only the harmonious resonance with nature.

I tried the GPT-4 spell with the domestic large model KIMIChat, and it worked. The answers given by Zhipu AI and other large models may not be as good.

As for what kind of creative videos the copywriting can generate, I don’t know.

When faced with a new technology or software, I follow a simple evaluation criterion: What is it? What does it do? How can I use it? Compared with macro-level discussions, I focus more on individual and small and medium-sized enterprise applications.

So, don't panic when something happens.

To sum up, everyone's evaluation mainly revolves around seven dimensions, namely: technological innovation, technological advantages, technical details, industry impact, ethical and safety concerns, market reaction, and user experience.

Can these things really bring about a revolution to small and medium-sized enterprises (individuals) quickly?

I don't think so. Why?

First, the issue of cycles.

Technologies have upstream and downstream relationships. If a new technology has been open sourced abroad, or its underlying principles have been made public for domestic reference, it will take some time for leading domestic companies to adapt.

After the leading companies develop the products and launch them on the market, whether small and medium-sized companies use them commercially or adopt a charging model similar to replicating GPT-4, there will be a cycle of half a year, a year, or even longer, which needs to be studied.

Imagine this:

JD.com, Taobao, and TikTok have developed a set of short video generation technologies to help businesses create more attractive advertising videos and improve customer acquisition efficiency.

It sounds great. In theory, this can make small and medium-sized companies' advertising attractiveness and sales ability soar. But to be honest, small and medium-sized companies may not pay much attention to it at first.

The first companies to try and widely apply this technology were B-side companies that provided software as a service (SaaS), especially those focusing on marketing technology (MarTech).

Because these companies can use this technology to provide customized advertising to customers, and in the marketing field, return on investment (ROI) is an indicator that everyone pays close attention to.

Therefore, only when the technology is proven to be able to produce more realistic virtual content, improve customer acquisition efficiency, and the simulation technology is powerful enough, small and medium-sized companies may truly begin to adopt it widely.

In other words, technology must go through a period of adjustment at first to prove that it can bring real benefits to the business before it can be accepted and used by more people. It can explode in an instant, but it is not quickly universal.

Second, technical practicality.

Vincent Video is currently still helping everyone create some things in the virtual world, just like the metaverse that was hyped up in the past few years.

People's food, clothing, housing and transportation will eventually return to the physical world. Now energy, food production, climate regulation, photosynthesis, biological rhythms, and environmental systems still rely on the sun to a large extent. Unless an artificial sun can be created one day, the lifestyle will not change much.

Fundamentally speaking, the ultimate goal of mankind is to survive and reproduce, which is engraved in their genes. Sora is currently unable to advance to important issues such as curing diseases, increasing food production, and boosting the economy.

Then, Wensheng Video has room for imagination. At present, it is still at the level of providing entertainment tools, increasing traffic and advertising clicks, which belongs to the redistribution of Internet attraction.

Furthermore, virtuality cannot be separated from reality.

Sora can create a fake world, achieve augmented reality, and even in the future, wearing Apple Vision Pro glasses will make it look more realistic. However, the fake world is always created according to our understanding of the real world, and it is different from the real world.

For example:

There are some scientific phenomena in the real world, including the laws of quantum mechanics, Newton's laws of motion, the laws of thermodynamics, the theory of evolution, the theory of plate tectonics... A fake world cannot be created.

Therefore, the fake world is a simplified version of the real world. At most, it can make us relax in terms of entertainment, but it is not very helpful for understanding the real world. Over time, it will make you live in fantasy.

I saw a joke yesterday. My friend said:

If 90% of the content viewed by TikTok users in the future is generated by AI, and they are unaware of it, what kind of world will this be? I also heard a joke: Now video creators use AI to outline the video, and after shooting the video, the audience uses AI to summarize the highlights of the video.

Well, very interesting and thought-provoking.

It has to be said that large and small models have indeed improved production efficiency and replaced a lot of simple work, but for the operation of a society and for reemployment at the age of 35, they do not seem to be so necessary.

Companies use AI to replace people to do work, which seems to improve efficiency. However, if they do not train their employees in new skills, it is equivalent to kicking people out of society . The money saved may eventually have to be spent on social problems.

Think about this:

There was an employee who took inventory of goods in the warehouse every day. It was a simple job and he felt very present. Suddenly one day, the boss said, I want to use technology to iterate you, create an inventory tool, and automate everything.

As a result, the employee lost his job. He tried to find a new job, but found that his skills were no longer up to date, and he gradually began to feel frustrated and powerless.

Finally, emotions not only affect oneself, but also colleagues and family, causing a chain reaction. Eventually, the company will find that the cost saved by improving efficiency is actually converted into social costs, adding more negative impacts.

Therefore: Sora is good, but it cannot replace people. It is a tool, and tools cannot be separated from people. Don't worry, there is still a long way to go...

Author: Wang Zhiyuan, public account: Wang Zhiyuan