Симуляция кризиса 2033 года: лидеры мира обсуждают риски неконтролируемого ИИ

A simulated crisis set in 2033, in which a US technology firm unveils an artificial intelligence system capable of programming itself without human input, has sparked renewed debate over global AI governance and the risk of superintelligence, according to a report by Times UK. In the scenario, the fictional company Synthiliant announces a breakthrough model that marks a decisive step toward superintelligent systems. The simulation depicts senior global figures from the United States, China, Europe and the United Kingdom gathered to assess the implications, as concerns escalate that artificial intelligence could become an uncontrollable force in international affairs. The exercise is presented in a documentary titled Intelligence Rising, which stages a structured war game designed to simulate how global powers might respond to AI-driven crises. The film, directed by Oscar-winning filmmaker Elena Andreicheva and developed by former government adviser Marc Warner, examines scenarios in which rapidly advancing AI systems reshape geopolitical decision-making and strategic stability. Within the simulation, former UK national security adviser Lord Sedwill tells participants that artificial intelligence could emerge as the most dangerous actor on the global stage. In response to the escalating scenario, retired US general Stanley McChrystal announces that the United States would consider pausing AI development, citing the severity of the perceived threat to humanity. The simulation then explores whether China would follow a similar path. The scenario concludes with China agreeing to pause its AI programme, a development that prompts reflection within the simulation on the fragile balance of international coordination. Sedwill delivers a central line in the film, stating that such an outcome between the United States and China would represent a “complete triumph,” while cautioning that failure to achieve such cooperation would leave the world in a far more dangerous position. Sedwill, drawing on his experience as a former cabinet secretary and participant in real-world strategic exercises, said the simulation appeared realistic in its structure and assumptions. However, he expressed scepticism about the likelihood of sustained global coordination at the level required to manage advanced AI systems. According to comments cited by Times UK, Sedwill said that achieving an international framework capable of controlling superintelligent systems would require unprecedented cooperation, with competitive pressures and geopolitical rivalries set aside. He compared the level of coordination required to a scenario in which humanity behaves as if facing an external existential threat. While acknowledging the theoretical risks associated with superintelligence, Sedwill said he was less concerned about AI systems resembling fictional catastrophic scenarios often depicted in science fiction. He indicated that existing security frameworks could likely manage certain categories of threats, including those involving misuse of AI in combination with other technologies. However, he warned that the integration of artificial intelligence into military command systems posed a more immediate and serious risk. Sedwill said that arms control in this domain would be extremely difficult due to strong incentives among states to maintain technological advantage. He referenced Cold War-era nuclear command systems, noting that there were instances in which escalation risked spiralling due to incorrect or incomplete information, and that human judgment had previously played a critical role in preventing catastrophic decisions. He warned that similar risks could emerge if AI systems were embedded into decision-making chains within military structures. The report also highlights broader debates in the United Kingdom over technological sovereignty and control of advanced AI systems. Sedwill noted that efforts to build domestic capability reflect wider geopolitical competition, particularly between the United States and China, which he suggested would likely maintain tight control over the most advanced systems, especially those with defence applications. Within the simulation and broader discussion, concerns were also raised about the economic impact of advanced AI systems, particularly potential large-scale job displacement. Sedwill criticised proposed responses such as universal basic income models, describing them as impractical and ideologically extreme in their implications for the structure of the state and economy. The documentary scenario further explores the possibility that once AI systems reach a certain threshold of capability, attempts to halt or reverse their development may become impossible. Sedwill noted that while some technologists argue systems could simply be shut down if they become dangerous, the risk lies in reaching a point where human control is no longer feasible. As depicted in Intelligence Rising, the question facing global powers is not only whether artificial intelligence can be safely developed, but whether competing nations can maintain cooperation in the face of accelerating technological change. The simulation concludes on the premise that achieving coordinated restraint would be a major success, while failure to do so could have consequences that are increasingly difficult to predict.