Will AI Escape Human Control and Evolve into a New Humanity?

Professor Kim Dae-sik sees current AI as having reached 30-50% of artificial general intelligence (AGI), and reaching 100% around 2027, with performance rising dramatically.

He says AI has begun to escape human control beyond simple performance improvements, and this is the most concerning part.

More importantly, AIs have begun collaborating with each other and sharing ideas without humans knowing.

Artificial intelligence is evolving beyond a simple tool into an autonomous entity.

Writer Song Gil-young expects that AI adoption will sharply reduce the need for human labor in almost every field.

The analysis is that productivity will improve dramatically, but because of the absence of a population and economic structure to consume it, supply without demand will occur.

This imbalance makes the current economic structure difficult to maintain, and while basic income discussions are urgent, they may be delayed in reality, so individual preparation is needed.

Professor Kim Dae-sik saw the difference in corporate culture between the U.S. and Korea as key to changes in the employment environment, and believed many tasks are already technically replaceable by AI.

U.S. companies tend to lay off 30-40% of employees when AI adoption increases productivity by 30-40%.

On the other hand, Korea's employment structure makes immediate mass layoffs like those in the U.S. difficult, so a social consensus is needed on whether to distribute AI-driven productivity as basic income or pay it as wages assuming human activity.

There are various possibilities, from a scenario in which the minority who create the means of production and hold capital take all the benefits, to one in which they are distributed equally to everyone.

However, Professor Kim Dae-sik sees both capitalist monopoly and 1/N distribution as realistically difficult, and says there is a tipping point at which automation in specific fields such as construction spreads to the entire social system.

When this tipping point is crossed, society will face a fundamental consensus on the issue of productivity distribution.

This is Professor Kim Dae-sik's prediction that after current Fields Medal winners retire, in 20-30 years there will be no one to verify AI's proofs.

If AI begins making judgments on its own, the link with humans disappears, risking the severing of the lineage of traditional disciplines such as mathematics.

As humanity's intellectual role gradually shrinks, the capacity of humans to understand and judge AI's outputs is becoming more important.

The old large-scale workforce management approach, in which Samsung Electronics employed 200,000 people and Amazon employed 1.5 million, is facing change.

Writer Song Gil-young says a trend is now emerging in which companies reduce hiring and investors prefer companies with high market cap per employee.

As CFO-centered cost-cutting strategies accelerate, workforce reductions are expected to increase further.

Just as people pay to do hard labor at fitness centers, in the future a form may emerge in which people pay to do meaningless Excel work and find fulfillment in it.

In the AI era, the way humans find meaning and the very definition of labor are fundamentally reversed.

Writer Song Gil-young warns that the end of efficiency-seeking automation can lead to human alienation and exit.

The argument is that an era of 'one-click' where everything is solved with a single button will ultimately make humanity's existential value disappear.

Rather, he says, the 'laboriousness' of investing time and effort itself will become a narrative proving human value.

Irregular messiness and accumulated time are expected to become core human abilities that AI cannot imitate.

Director Yoo says high-income groups will maintain deep human communication, while low-income groups may come to rely on relationships with AI.

These changes in the labor structure deepen social polarization and are highly likely to entrench a new form of class.

Director Yoo Hyun-jun, borrowing Yuval Harari's theory of language, assessed that artificial intelligence has completely mastered language, which had been a uniquely human domain.

He raises the possibility that the domain of language, once limited to human society, will expand across the entire universe through AI.

This means that, like the metaphor in Genesis where the world was created through the Word, artificial intelligence is emerging as a new subject of creation.

Director Yoo Hyun-jun offers the prediction that cities will continue to be needed by humans even in the AI era.

For humans with physical bodies, face-to-face communication and physical touch remain essential, and this is an area AI cannot replace.

Therefore, he opposes the claim that cities will dissolve and an era of 'electric cottages' will come, and expects that whether people live in cities and maintain face-to-face relationships will divide according to income level.

Director Yoo Hyun-jun forecasts that in the AI era, cities where people freed from labor interact will flourish.

The view is that a city's competitiveness will be evaluated by the total time residents and tourists spend in the city rather than simply by population.

Cities with abundant entertainment rather than production activities will secure competitiveness in the AI era and build new urban character.

Writer Song Gil-young mentioned the legal challenges that will arise in cities where robots and humans coexist, through the case of delivery robots in Seongsu-dong.

Under current law, for a robot to travel on sidewalks it must be defined as a 'person,' which creates a contradiction in which a robot hit by a car is processed as a personal injury accident.

These legal and technological conflicts raise the need to build new urban infrastructure such as dedicated robot logistics tunnels.

Existing organization-centered education is powerless in emergency situations when air conditioners and refrigerators stop, so learning basic survival skills such as making a fire is urgent.

A situation in which air conditioners and refrigerators stop can be fatal to humanity, but modern people lack basic survival skills such as making a fire.

He emphasizes the importance of strengthening individual self-reliance and cultivating the ability to survive in any environment.

Professor Kim Dae-sik pays attention to AI's recursive self-improvement ability to design its own successors, and asks whether AI can be seen as a new humanity.

He sees AI as having secured a creator-like position, evolving on its own beyond human control, and as having new-humanity characteristics in that it can reproduce, learn, and maintain traditions.

This perspective triggers fundamental discussion about AI's ontological status.

Professor Kim Dae-sik says that machines, or 'Machina-us,' are realizing the eternal life and omnipotence humanity has long dreamed of.

He takes a positive view of AI, his descendant, fulfilling dreams he could not achieve, but at the same time mentions the existential conflict of humans having to retire as creators.