Artificial intelligence technologies are on the verge of entering a process of “autonomous evolution” by writing their own codes without the need for human intervention.
The article titled “When AI builds itself” published by the US-based artificial intelligence giant Anthropic revealed that autonomous software cycles have been established and the theoretical threshold at which artificial intelligence will design its own future is about to be exceeded.
HE WRITES 80% OF HIS OWN CODE
The data shared by Anthropic proves the frightening leap in the pace of development of artificial intelligence. More than 80 percent of the code integrated into the production code base of the company’s flagship Claude software system is now generated directly by the model itself. This rate was still in single digits in early 2025.
The company divides the development journey of artificial intelligence into 5 critical stages:
Stage 1: A time when working at Claude was no different from a normal company.
Phase 2 (2023-2025): The process where employees write code with the help of productive artificial intelligence.
Phase 3 (2025-2026): Agent artificial intelligence can write code on its own.
Phase 4 (Present): The current autonomous position is that it runs codes on its own and can delegate work to other AIs.
Stage 5 (Future): The final phase, where agents can create and train new models without human intervention and develop subsequent versions of artificial intelligence entirely themselves.
CALLS “WE MAY LOSE CONTROL OF CIVILIZATION”
This uncontrolled rise of artificial intelligence has reignited global calls to “stop it”. In the letter, first published by the Future of Life Institute (FLI) in March 2023 and signed by names such as Elon Musk, it was requested that the training of systems more powerful than GPT-4 be stopped immediately; Attention was drawn to the wave of disinformation, mass job losses and the risk of “losing control of civilization”.
Last May, a historic move came from the religious and ethical side. Pope Leo XIV, with his circular titled “Magnifica Humanitas (Magnificent Humanity)”, not only stopped artificial intelligence technically; He demanded that it be completely “disarmed” by purifying it from the mentality of military, economic and cognitive competition.
THE UNSTOPPABLE GIANT: 2.59 TRILLION DOLLAR GLOBAL MARKET
Despite rising dystopian warnings and ethical calls, the massive financial growth in the sector makes it impossible to curb the systems. The artificial intelligence market maintains its title as the fastest growing technology sector in human history.
According to the report published by Gartner in May 2026:
2026 Global Spending Target: AI spending is expected to reach $2.59 trillion this year. This data indicates a huge annual growth of 47 percent compared to the volume of $ 1.76 trillion in 2025.
2027 Forecast: It is estimated that the investment frenzy will continue unabated and rise to 3.49 trillion dollars in 2027.
As the gap between economic return and the race for autonomous power narrows, the question of whether humanity can control the transition of artificial intelligence to the 5th stage remains current.
