One person does 1,000 jobs, “shameless” has become a new way of making money

To play a little game .

Guess which of these 4 people is true and which is false?

The answer is: the two on the left are artificial people generated by AI, and the two on the right are real people.

Guess again, which of the following people are true and which are false?

Well, in fact they are all fake. But their faces are all from real people.

Start-up Hour One allows them to use their own face to do different jobs in different countries, different fields, and different companies around the world, and speak any language in any language.

One person can even do 1,000 jobs in 1,000 videos.

What people have to do is to "rent" their faces.

How to "sell face"?

We are no strangers to the "face changing" technology.

For example, let your face become older and smaller in the app; for example, a celebrity does not need to go to the scene to film a scene, and the face can be seamlessly synthesized in the video.

A while ago, Huang Renxun's "digital stand-in" exposed at the Nvidia press conference made people see AI with fake and real technology…

The technology of Hour One is similar to Deepfake, but it is a bit different.

Deepfakes is to superimpose the captured face on the existing lens, and Hour One allows people to take a video of the face, and then the original content is independently generated by AI.

This company, which was only established in 2019, wants to buy the license rights for your face, so that everyone can use their face to "work".

In the past, we were getting facial information all the time. Now, Hour One wants you to actively sell facial information.

The way to participate is simple.

First apply to become a role on the Hour One platform.

In order to ensure that there are as many role samples as possible in the database, Hour One will screen applicants and select people of different ages, genders, and races.

It is understood that they currently selected about 80% of the participants under the age of 50, 70% are women, 25% are white, and it is increasing every week.

After passing the screening, people only need to shoot a 10-minute video through their 4K camera in the Hour One studio.

During the shooting, people need to speak in front of the green screen and make various facial expressions at the same time to complete the task.

If you have equipment and scenes at home, you can even shoot remotely by yourself.

Next, it's Hour One's work.

They will put video data into AI software and output digital content through a creative process similar to Deepfake.

After that, they can make people's faces say anything in different body models as long as they enter text, and within a few minutes, hundreds of videos can be generated .

At present, Hour One has digitized 100 face models, and these "digital people" with different identities have entered the "work" of more than 40 companies.

At the same time, the person who rents out the "face" will also get corresponding remuneration. Every time a new customer uses their face, they will receive a small amount of money, just like the "copyright fee" for their own face.

Hour One did not disclose the specific amount , only that it is not enough for people to make a living, but the income is in U.S. dollars instead of cents.

If everything goes well, this will be a good way for people to earn part-time income.

What kind of work can your "face" do?

Liri from Israel only joined Hour One not long ago.

She is only 23 years old and has worked as a waiter and bartender in Tel Aviv, selling cars and sales in Germany, interviewing new employees in the human resources department, giving them induction training, and becoming a spokesperson for multiple brands.

Of course, it is Liri's "face" that really does so many jobs.

Now, Hour One has been involved in multiple industries , the core of which is marketing and education, as well as entertainment, games, e-commerce, real estate, digital health, and so on.

The good thing is that people have the right to choose the "face" they rent, and they can be allowed to do what they do or not.

Of course, not all jobs can be done with the rental face.

The "virtual digital person" created by Hour One at the beginning was to act as a receptionist in various places .

They cooperated with Alice Receptionist, displayed on the door screens of many companies, registered visitors, dealt with visitors’ questions, and had video chats with real employees.

And Hour One has made a sound in the industry, and it is a collaboration with Berlitz , an international language school .

At this time, the "virtual digital person" became an online teacher.

Video teaching is an important content of Berlitz , but the same teacher often has to record a large number of repetitive course content, and the cost of the filming and production team is also very high.

"Virtual digital humans" not only help them solve the problem of manpower, but also solve the problem of money.

Hour One has previously cooperated with the education industry to create 8 different teacher roles and quickly generated 13,000 instructional videos that can teach people in 3 languages.

Now Hour One is involved in more and more industries, and "virtual digital humans" can do more and more jobs, which inevitably arouses people's controversy over it.

The first thing to bear is that it will threaten people's employment.

In fact, the current technology of these "virtual digital humans" can only output fixed content and cannot conduct smarter conversations and services, but the wave of automation has already hit.

▲ Picture from: Stocksy

The "Future Jobs Report" estimates that by 2025, AI will eliminate 85 million jobs, but at the same time, it will also create 97 million new jobs.

Some repetitive, mechanical, and boring work will inevitably be gradually replaced, and people will also engage in creation, research, and composite work that can more invest their own value.

So what people should worry about "selling faces" is actually the security issues behind it.

Is it reliable to work with "face"?

Ten years ago, some Cambridge graduates "rented" their faces on the street to repay their loans , earning nearly RMB 500,000 in half a year.

Although they have turned their faces into living "advertising spots," they have no effect on their privacy.

In February of this year, the Japanese retailer "仮面屋おもて" started to pay for people's faces to promote mask culture.

People can sell a "face" for more than 2,000 yuan, and the boss sells the "face" as a 3D realistic mask, which can be sold for more than 6,000 yuan.

Hundreds of people soon came to "sell faces." The boss reproduced the facial features, freckles, moles, eyelashes, and beards of these facial masks one by one. Because they were too real, they caused a lot of controversy about facial crimes .

However, the face information is at least in the hands of the boss. He said that he will not disclose private data to the outside world. Buyers also need to contact the store to inform third-party information when transferring masks.

But now, face data has become more and more difficult to control, and the privacy issues behind it are frequently exposed.

From unlocking mobile phones, downloading applications, transferring payments, to checking in at work, entering high-speed rail stations, and going out to apply for a permit, face recognition has become a contemporary pass , whether online or offline.

All kinds of news about the leakage of facial information and the reselling of facial data in batches remind us all the time : "Being face" everywhere, don't "shame."

But every year, there will still be new privacy leaks. It is very likely that a bug will cause the privacy of countless people to disappear.

From the release of the video of the video conference application ZOOM last year, the release of a large number of celebrity health treasure photos at the end of the year, to this year's 315 exposed shopping malls that were not authorized to perform face recognition to collect customer data…

We are becoming more and more invisible right now.

Hour One has also been discussed by people about information security. They pointed out on the privacy ethics regulations page:

  • The video will clearly indicate that the characters are all generated by computers and respect people's right to know;
  • Unlike Deepfake, the use of character images in all ways will be approved by the parties and protected;
  • The use of character images is limited to "legal and safe content" and will not provide services for gambling, pornography, or politics that may have ethical issues.

But the specific technical operation process and usage tracking situation do not have more detailed content.

An editor of Ai Faner believes that "data privacy in this era cannot be maintained. It is better to sell it yourself if it is stolen by others." In his opinion, Hour One is like a "reverse thruster."

It is using magic to defeat magic, and using "infringement" of portrait rights to combat infringement of portrait rights.

Of course, at the moment when many technologies and rules are unclear, we do not encourage the rental of human faces.

But in the past, our private data was uploaded passively, with the purpose of creating better tools and improving services everywhere in our lives. The emergence of Hour One gave us a new option to actively sell our own data.

At least it seems that privacy can still be controlled by oneself at the moment.

#Welcome to follow Aifaner's official WeChat account: Aifaner (WeChat ID: ifanr), more exciting content will be provided to you as soon as possible.

Ai Faner | Original link · View comments · Sina Weibo