While artificial ielligence is becoming an iegral part of human life, concerns about its possible consequences have also iensified. Dario Amodei, CEO of Ahropic and one of the key figures in the developme of advanced artificial ielligence models, recely stated the possibility of a social disaster caused by the technology in explicit and reflective statemes. This view, along with his optimism about the positive future of artificial ielligence, provides a realistic image of opportunities and threats ahead.
Four in some cases may seem good. This possibility is far better than the chances of winning in most games. However, it is apparely low that Dario Amoudi, CEO of Ahropic, appears to be indiffere after estimating the possibility of a social disaster caused by artificial ielligence. At the Axios AI + DC Summit meeting, Amoudi responded with calm when asked about believing in the “possibility of destruction” caused by artificial ielligence:
I think it’s 2 % likely to be really bad.
Amudi means “really, really bad” merely minor mistakes such as automatic modification of the word “duck” in an inappropriate word; Rather, it refers to scenarios that can threaten social systems, including existeial risks, severe abuses of artificial ielligence, and unnecessary consequences that may be catastrophic.
In an industry that is often drowned in idealistic promises or reduced to sci -fi fear, Amoudi’s attitude towards the possibility of apocalypse and the reasons for its coinued activity in this area is distinct.
Amudi is not the only one who feels worried, but his place is special. As a company CEO that has developed the Claude model, he is not just a passive observer. He is shaping the path of this technology in real time. His team builds the same systems that he is assessing their capacities and risks.
If someone tells you that there is a 5 % chance of explosion each time the car is turned on, you probably prefer to walk. But Amoudi apparely prefers to be mechanical and check the car before it is riding.
This is not the only Amudi warning of artificial ielligence. He had earlier warned that artificial ielligence could eliminate half of the elemeary jobs in the office and had warned of exporting advanced chips from the United States to China. This makes Amuddi’s iellectual framework useful: he accepts the danger, quaifies uncertaiy, but still places human action.
On the other hand, “5 % of the possibility of going well to the situation” is not just unnecessary optimism. This view reflects the belief that artificial ielligence can have enormous benefits for everyone. This technology can lead to improved medical improveme, more efficie production, and even formulating strategies to deal with existeial crises such as climate change (although one of the key elemes in solving this problem will be the energy needed to impleme artificial ielligence models).
But that 5 % of the risk requires that these benefits be carefully constructed, along with appropriate safety measures and regulations. Because if the future is 5 % brillia and 2 % devastating, the question is: What should we do to weight the future to the right part?
