埃隆·马斯克在接受《经济学人》主编的90分钟专访中阐述了其对人工智能发展前景的完整观点。[1]他预测，约5年内AI的综合智力将超过全人类智力总和，10年内人类大概率不再是地球的主导力量。[1]马斯克认为技术加速是不可逆的过程，超级智能的出现属于必然趋势。[1]
在立场转变上，马斯克表示，他曾在三年前签署暂停前沿模型开发的公开信，但如今认为即使有停止按钮也不应该按下。[1]关于更远的未来，马斯克预测到2036年金钱将不再重要，因为数字智能与物理机器人的结合将创造出接近无限的经济产出。[1]
在AI安全治理方面，马斯克提出应建立竞争对手参与的安全机制，即任何突破性模型发布前，必须给竞争对手1-2周的API早期测试权。[1]同时，他指出AI能力的核心公式为"芯片×电力"，并认为中国在电力和工业制造领域具有明显优势，光刻工艺的突破比多数人预想的更为接近。[1]
Elon Musk outlined a comprehensive vision for artificial intelligence development during a 90-minute interview with The Economist's editor, forecasting transformative changes to human civilization within the coming decades [1]. Speaking at Tesla's Texas Gigafactory on July 23, 2024, Musk predicted that within approximately five years, artificial intelligence will surpass the combined intellectual capacity of all humanity [1]. He further contended that within a decade, humans will likely no longer be Earth's dominant force [1].
Musk has significantly shifted his position on AI development compared to his stance three years ago, when he signed an open letter calling for a pause on advanced model development [1]. He now argues that even if a "stop button" existed for AI advancement, it should not be pressed, emphasizing that technological acceleration is irreversible [1]. According to Musk, the convergence of digital intelligence with physical robotic systems will create a near-infinite economy, rendering money obsolete by 2036 [1].
On the question of AI safety governance, Musk proposed a mechanism requiring competitive participation: any breakthrough model released to the public must first provide competitors with one to two weeks of early API testing access [1]. Regarding the competitive landscape, Musk framed AI capability as a function of chips multiplied by available electrical power, noting that China possesses significant advantages in both electrical infrastructure and industrial manufacturing, while also suggesting that breakthroughs in lithography may arrive sooner than most observers expect [1].
