财联社7月29日讯（编辑 黄君芝）Meta公司首席执行官马克·扎克伯格周二在发表的一篇新评论文章中写道人工智能的兴起应该用来赋能所有人而不是被少数机构集中控制这似在抨击世界上两家最具价值的AI创业公司。

扎克伯格撰文指出，在未来几年，人们将能够利用超越人类能力的超级智能来创造和发现新事物，并推动经济增长和创造机遇。他指出，这与一些人工智能开发者的言论截然不同，并认为广泛普及人工智能及其能力将带来更好的结果。

令人惊讶的是，许多从事人工智能开发的人员的言论中充斥着悲观情绪，扎克伯格写道，他不明白，为什么那些认为人工智能会取代大多数工作岗位并削弱人类存在意义的人会急于构建这样一个未来。

扎克伯格进一步解释说：“认为人工智能如此危险，以至于唯一安全的道路就是权力的极端集中，这种观点本身就很危险。从历史上看，指望绝对权力在人类足够开明的情况下能够仁慈地造福人类，并没有带来安全或积极的结果。”

He emphasizes that if leading research laboratories want to develop AI technologies in a highly controlled manner, that would be equivalent to abandoning our values in the development of American technology and would stifle innovation.

Anthropic and OpenAI are currently developing AI in the manner described by Zuckerberg. The two companies believe that when model capabilities continue to approach or even surpass human levels, complete openness of model weights would enable any organization to download, modify and deploy them, thereby increasing the risk of model abuse.

OpenAI previously stated its hope to establish a government-led AI safety evaluation system that would impose mandatory safety testing on certain advanced models. Anthropic has repeatedly emphasized that open-source models could lower the threshold for the diffusion of dangerous capabilities and therefore should be subject to more stringent management measures.

扎克伯格表示，纵观历史，科技领域出现了许多变革性的进步，这些进步也引发了人们的担忧，担心科技会将人们抛在后面。但最终，随着科技的进步，人们享受到了更多的繁荣、健康和自由。

He wrote: “Endowing every individual with the power to pursue their own aspirations represents the method by which humanity achieves its greatest progress. Novel ideas and major advancements rarely arise solely from established institutions.”

Zuckerberg has suggested that the major technological advances that have shaped human life, such as flight, electricity, and the personal computer, have been driven by individuals who do not maintain deep connections with institutions. With the statement that as more powerful tools become available to everyone, every person will be able to shape the future rather than becoming weaker.

与此同时扎克伯格也承认了人工智能在自动化应用以及促进创新帮助人们拓展技能和创业之间的平衡需要取得并补充说如果这种平衡倾向于自动化可能会对就业和经济产生负面影响。

但如果超级智能得到广泛应用，我相信未来就业机会将会增加，而不是减少。创业将变得容易得多，而无需筹集大量资金。我预计经济将变得更加创业化，更多的人在小型企业而不是大型公司工作。他写道。

Zuckerberg added that the development of superintelligence “will be the most profound technological advancement we will witness in our lifetime.”

Meta is committed to the principle of personal empowerment, innovation and power balance in its construction. The trajectory of human history has developed in the direction of empowering more people with power. If these values can guide the direction, then I am optimistic that we can create a beautiful future for all. He wrote.

来源：暗讽OpenAI和Anthropic？扎克伯格：AI应该赋能所有人！ | 财联社