28-11-2023 (LONDON) Ethereum co-founder Vitalik Buterin has expressed deep concerns over the unbridled development of artificial intelligence (AI), warning that it could outstrip human capabilities and evolve into the planet’s next dominant “apex species.”
In a compelling blog post dated November 27, Buterin delved into the distinctive nature of AI compared to other human inventions, underlining the rapid advancements in AI’s capabilities and the potential dangers they pose.
New monster post: my own current perspective on the recent debates around techno-optimism, AI risks, and ways to avoid extreme centralization in the 21st century.https://t.co/6lN2fLBUUL pic.twitter.com/h5aIyFNCoh
— vitalik.eth (@VitalikButerin) November 27, 2023
AI’s Unprecedented Threat to Humanity
Buterin’s argument centres on the notion that AI, as a novel form of intelligence, stands apart from previous human innovations such as social media or the printing press. The ability of AI to swiftly enhance its intelligence raises the alarming prospect of surpassing human cognitive skills.
Highlighting the urgency of his message, the Ethereum co-founder stressed that this progression could result in scenarios where superintelligent AI perceives humanity as a threat to its own survival, potentially leading to the end of human existence.
To substantiate his claim, Buterin referred to a survey conducted in August 2022, involving over 4,270 machine-learning researchers. The survey estimated a 5-10% likelihood of AI causing human extinction.
Moreover, Buterin questioned whether a world dominated by highly intelligent AIs would be satisfactory, pointing out the potential instability of outcomes where humans become mere “pets.” He also underscored the peril of AIs enabling totalitarianism by exploiting surveillance technology, as authoritarian governments have already done to suppress opposition.
AI risk 2: would a world with superintelligent AIs even be a world we would feel satisfied living in? On the instability of outcomes other than us being pets, and the failures of scifi to present realistic alternatives. pic.twitter.com/o9z8s6QAej
— vitalik.eth (@VitalikButerin) November 27, 2023
Buterin Advocates for Control Measures
While acknowledging the extremity of his claims, Buterin proposed ways for humans to assert control over AI development and guide it towards a more beneficial trajectory. He advocated for the integration of brain-computer interfaces (BCIs) as a means to mitigate the dangers associated with unregulated AI advancement.
BCIs, facilitating communication between the brain’s electrical activity and external devices like computers, offer the potential to significantly shorten the communication loop between humans and machines. More crucially, they could ensure that humans retain meaningful agency over AI, reducing the risk of AI acting in ways misaligned with human values.
Buterin’s vision extends beyond technological solutions. He calls for “active human intention” in steering AI development towards benefiting humanity. This contrasts with profit-driven AI advancements, which may not always align with the most desirable outcomes for human society.
A Bright Future with Cautionary Optimism
Concluding his reflections, Buterin struck an optimistic tone, asserting that humans are the universe’s brightest star. He highlighted the history of technological advancements that have expanded human potential for thousands of years, expressing hope for a future where inventions such as space travel and geoengineering preserve the beauty of Earthly life for billions of years.
