Toshinao Sasaki's thoughts on "Jobs and possibilities for people replaced by AI" Tech and Future Vol.5

AI and Turing test

A serial project in which we ask Toshinao Sasaki about the future of technology and society.

This time, the theme is the new world possibilities created by generative AI and the metaverse.

I would like to ask you about the possibility of generative AI and the metaverse coming together in the future to create a new world, but first, what are your thoughts on the potential of generative AI?

Toshinao Sasaki (hereinafter Sasaki): I think we need to have a certain vision of what generative AI itself will become.

Currently, generative AI is often introduced on TV talk shows as being used like an encyclopedia, but the problem of AI collecting incorrect information, such as hallucination, has not yet been solved, so I don't think we can expect much from it in that respect.

Of course, problems like hallucination may be solved with technological advances, but I think that using it for chores is one possibility.

For example, Microsoft has announced and released generative AI such as "Microsoft 365 Copilot," and the direction is to have generative AI do the daily chores that we currently do, such as creating Excel tables or generating PowerPoint presentation materials from Word files.

What this means is that the work done by humans will become more advanced. If there are 100 jobs done by humans, chores make up a very large portion of them.

For example, if you're a freelancer like me, I think about 30% of your time is spent on invoicing and other miscellaneous tasks. That 30% will likely decrease with the use of generative AI.

Routine work, not just for blue-collar or white-collar workers, will likely be carried out by AI.

Will the people who make up 30% of that work lose their jobs?

Sasaki: I agree. There is a problem that those 30% of people who did odd jobs will lose their jobs, which leads to the recent debate about whether AI will take over jobs.

This is important, so I'll mention it here: Given the current development stage of generative AI, I expect that routine work, not just blue-collar or white-collar, will be taken over by AI.

For example, even for white-collar workers, tasks such as creating materials and proposals will be done by AI, and the communication and creative aspects that AI cannot do, and even when writing manuscripts, writing from a novel perspective that no one would think of, rather than an average article on a certain product, will remain as human capabilities.

I think that such non-routine work will become the work of humans. This is not a matter of whether the work is advanced or not, but it is said that non-routine work will remain as human work, even for blue-collar workers.

I think that there is a possibility that routine odd jobs will decrease among white-collar workers, and the people who were engaged in those jobs will move to non-routine blue-collar work. In other words, it is a craftsman's job, such as a carpenter's.

What about other possibilities for AI?

Sasaki: Another possibility is that current human-to-human communication will remain in the hands of humans, but since we don't know how far AI will evolve, it's difficult to judge to what extent it will remain in the hands of humans.

For example, in today's society, where one in four people will be elderly in the future, it is expected that AI will be used as a conversation partner for elderly people living alone.

In fact, Amazon and others have announced that they are planning to release a smart speaker equipped with generative AI, which will allow you to continue talking to the smart speaker. I think there is a good chance that this will become a conversation partner for lonely elderly people.

However, the difficult part here is whether the other party will recognize whether it is an AI or not. In reality, the voice function has been rolled out to the ChatGPT smartphone app, which allows you to communicate with AI by voice, and using this can sometimes feel like you are talking to a human.

In the early days of computing, British mathematician Alan Turing devised a test to determine whether a machine was "human-like," known as the Turing test, and it can be said that current AI already passes the Turing test.

Eventually, when we are able to have normal conversations with AI, we probably won't think much about whether the other person is human or not.

In virtual space, everything will be personified,
and we will see a world like an expanded social networking site.

Will it be possible in the future to sample someone's voice and have a simulated conversation?

Sasaki: Yes, for example, there is already something called "AI Hiroyuki," and when I spoke with the president of the company that produced "AI Hiroyuki," he said that it would be possible to generate an AI for a person with about five minutes of recording, so the current level of technology has come to that level.

In the future, similar things will likely be possible with videos, and in Hollywood there is a business that uses AI to revive dead actors, and there is talk of American rock band KISS performing live with avatars in the future.

If we push the technology of generative AI, including deep fakes, it will become quite realistic to reproduce an exact replica of the real thing.

This leads us to the metaverse, and what will happen if such deep faked avatars appear in the metaverse? For example, it will be possible for deceased parents and siblings to exist in the metaverse.

These things exist in the metaverse as 3D beings, and when the currently low resolution of VR rises to 8K or 16K, the resolution of the human eye, I think it will become highly likely that it will become impossible to distinguish between reality and reality, and at least in the virtual space of the metaverse, it will become impossible to distinguish between humans and AI.

In the context of IoT, there is also the term social machine, and the concept is that home appliances are personified and turned into AI and connected to social media, but if this becomes commonplace, I think we may see a world where everything is personified.

I imagine a world like an extended SNS where real people and dead people exist, as well as fictional characters and non-human beings connected to the IoT.

Isn't it true that a world where everything is personified is a thing of the distant future?

Sasaki: I think it's possible even now if it's just text or voice. However, the metaverse requires the evolution of VR devices, so I think it's difficult with current devices.

I think it will be possible if it eventually becomes the size of glasses that can be used daily, but to do that, not only VR technology but also detailed technologies such as battery technology will need to evolve.

There is another meaning to personified AI. The entrance to computers has changed from mice and keyboards to touch screens, but it is said that in the future the entrance will change to generative AI. That means that personified robots and avatars will be the entrance.

For example, smartphones, which are currently the control towers of everything, are currently operated with touch screens, but it is said that in the future they will be operated with voice and gestures. In that case, the control tower will not necessarily have to be a smartphone, and a robot would be fine.

I think the image of the future is that in the metaverse, you can make requests to personified avatars, and those avatars will give instructions to various places.

Book Review

"Dismantling the 'myths' that are corroding this country: Citizen perspective, denial of technology, narrativization of terrorists, anti-authority"

"Power is always evil," "The common man's sense is always right," "The weak should be protected," "Artificial things are dangerous," "Natural things are best"... Outdated values ​​that continue to linger in Japanese society. This is the latest discourse that dismantles the "myths" that run counter to the evolution of cutting-edge technology and serves as a guidepost for thinking about the future.

Toshihisa Sasaki (author) Tokuma Shoten (September 28, 2023)

Profile

◉Toshinao Sasaki
Born in Hyogo Prefecture in 1961. Dropped out of Waseda University's School of Political Science and Economics. After working as a reporter for the Mainichi Shimbun and at the ASCII monthly editorial department, he is now a freelance journalist. He has written many books, including "The Age of Curation" (Chikuma Shinsho), "The Layered World" (NHK Publishing Shinsho), "Homemade Meals Are the Best Treats" (Magazine House), and "And So Life Becomes a Community" (Anonymous Studio).



Related articles

Sasaki Toshinao's thoughts on "The future of mass media" Tech and Future Vol.4

Sasaki Toshinao's thoughts on "Dialogue between AI and humans" Tech and Future Vol.3