Vitalik warns AI must empower humanity not disempower it

Ethereum co-founder Vitalik Buterin has urged for careful development of artificial intelligence (AI), saying it should help humanity grow stronger, not create systems that could take control away from people. Buterin wrote on the future direction artificial intelligence is taking. In a series of posts, the Ethereum co-founder made it clear that AI should empower humanity – not disempower it. AI should help humans, not take over He warned that “poorly designed AI systems might produce something that’s more or less a non-controllable, self-replicating entity.” AI done wrong is making new forms of independent self-replicating intelligent life, Buterin wrote. “AI done right is mecha suits for the human mind. If we do the former without the latter, we risk permanent human disempowerment. If we do the latter, flourishing superintelligent human civilization.” He believes AI should augment human thinking, creativity, and so on, kind of like a “mecha suit” which enhances the user’s strength. This is contrasted by the dangers of autonomous AI systems without human oversight, according to him, which may eventually lead to a loss of control over the technology we create. Buterin further suggests that there should be the use of “distilled human judgment” as a collective means of improvement in decision-making. He further described this as creating systems that could imitate high-quality human decision-making processes but perform more efficiently than humans. Thus, humans could maintain control with the benefits reaped from the capabilities of AI. The latter can be done at the collective level through distilled human judgment, he explained; the collective human intelligence can provide guidance to help AI amplify, rather than substitute for, human abilities. The risks of uncontrolled AI systems Buterin goes on to clarify that the emerging AI tools – such as this real-time video generation software – have an example AI system that indeed empowers, while the “way of ‘passive system’ like, for example, the text-to image model, of which user influences are very minimum.” Prompt and wait doesn’t work that way, aim and adjust to a constant real-time base,” he puts in, suggesting a more intimate and empowering dynamic between humans and AI. He then talked about AI agents: the programs that will do tasks without human input. He explained two types: one that does long-term, complex planning, and the other simpler AI tool like a chatbot, replacing GUIs. The latter, he said, “Chatbot replaces GUIs as the interface to everything. The latter is great!” Ultimately, Buterin advocates for responsible AI development to empower humanity. He is not in favor of an uncontrolled, unbridled drive towards autonomous systems and instead sees tools that serve to enhance human abilities. “If done right,” he thinks, “AI can lead to a flourishing superintelligent civilization.” From Zero to Web3 Pro: Your 90-Day Career Launch Plan