AGI Shockwave, AI Self-Replication Loop, Entry-Level Jobs Obliterated, US-China Chip War Escalates

● AGI Shockwave, AI Self-Build Loop, Junior Jobs Wiped Out, US-China Chip War

It might be best to prepare yourself a little before reading this. What we’ll cover today isn’t just a simple technology forecast, but a story about ‘the day after AGI (Artificial General Intelligence) arrives,’ directly revealed by Anthropic’s Dario Amodei and Google DeepMind’s Demis Hassabis. In particular, we’ve included sharp perspectives on the moment the ‘loop where AI creates AI’ closes—a topic not often covered in other news—and how junior-level jobs will disappear because of it, as well as the chip war between the US and China. From now on, I will explain the core point of the future seen by the world’s top AI leaders very simply.

The Shocking AGI Timeline Predicted by the World’s Two Great AI Giants

Since their meeting in Paris last year, DeepMind’s Demis and Anthropic’s Dario have reunited. At this gathering, which felt like The Beatles meeting The Rolling Stones, the hottest topic was undoubtedly ‘when AGI is coming.’

The ‘Loop’ Where AI Creates AI Itself Is Closing

The most notable point is Dario’s remarks. He maintained his existing stance that a model with capabilities at the level of a human Nobel Prize winner will emerge by 2026 or 2027. However, what’s important here is the driving force of ‘speed’.

Dario revealed that coding automation has progressed to the point where internal engineers at Anthropic say they “no longer do the coding directly, but tell the model to do it and then just make corrections.” He predicted that within the next 6 to 12 months, AI models will perform most of the tasks of a software engineer.

The key takeaway here is the ‘completion of the Loop’. The moment the cycle of self-improvement closes—where AI directly performs coding and AI research to create the next generation model, and that model creates an even smarter AI—technological development speed will be far faster than we imagine. Demis also agreed that it is possible for this loop to close in verifiable domains like coding or math.

Exploding Corporate Value and the Acceleration of Technological Competition

The competitive landscape and growth trends of the two companies are also terrifying. Dario mentioned that Anthropic’s revenue would grow 10 times every year, from $100 million in 2023 to $1 billion in 2024, and to $10 billion (approx. 13 trillion KRW) in 2025. This is at a level that surpasses the growth curve of a simple startup.

Demis also showed confidence, stating that Google DeepMind is reclaiming the lead through ‘Gemini 3’ and others. In particular, he emphasized that after the ‘Code Red’ alert within Google, the organization has rearmed itself with a startup mindset and is rapidly turning research results into products. In other words, the Generative AI market has now entered a phase of all-out war between giant corporations where massive capital and research power are concentrated.

The Future of Jobs: ‘Quiet Layoffs’ That Will Strike White-Collar Professions

The topic that office workers will find most sensitive is jobs. Dario maintained his stance that “half of entry-level white-collar jobs could disappear within the next 1 to 5 years.”

There Is No Place for New Hires and Interns

Signs of this are already appearing in the software field. Demis predicted that while productivity will increase in the short term thanks to new AI tools, on a 5-year timeline, hiring for entry-level jobs such as junior levels and internships will slow down.

This is because simply handling AI tools skillfully can produce results that surpass existing training processes. The problem arises when the speed of technological development overtakes the speed at which the labor market can adapt. Dario is concerned that this exponential speed of development will overwhelm social adaptability.

Geopolitical Risks: AI Chips Are Like Nuclear Weapons

This is a very important point that is not well covered elsewhere. Dario expressed a very critical view of the US government’s semiconductor export policy toward China.

Export Controls and the National Security Dilemma

Currently, the US is adopting a strategy of selling chips to keep China tied to the supply chain, but Dario strongly compared this to “selling nuclear weapons to North Korea for Boeing’s profit.”

He views AGI technology as having destructive power directly linked to national security, not just at the level of communication equipment. Therefore, while technological competition with his rival Demis can be resolved somehow, he pointed out that an authoritarian state getting its hands on powerful AI is a problem on a different level. This suggests that AI technological hegemony will become the biggest variable in the future global economic outlook and international situation.

< Summary >

AGI Arrival Time: Both Anthropic and DeepMind expect the late 2020s, and once the ‘self-improvement loop’ where AI develops AI is formed, the speed of development will transcend imagination.
Economic Impact: Anthropic’s revenue is growing 10 times annually. However, junior and entry-level white-collar jobs are projected to face serious threats within 1 to 5 years.
Geopolitical Crisis: AI semiconductor chip exports are not just trade but a security issue. Dario compares this to nuclear proliferation and argues for strong controls.
Safety and Regulation: Both leaders reject ‘doomsday scenarios’ but emphasize that international cooperation and safety standards are urgent to prevent risks such as biological terrorism or national misuse.

*Source: World Economic Forum

NextLens.Net