Mrinank Sharma says 'world is in peril' now
In
February 2026, Mrinank Sharma, the former Safeguards Research Lead at the AI
safety company Anthropic, resigned and published a letter stating that the
"world is in peril".
His
resignation has gained significant attention because he led the team
responsible for keeping Anthropic's AI models, such as Claude, safe from
misuse.
Key
Details of the Warning
Interconnected
Crises: Sharma clarifies that the danger is not just from AI or bioweapons
alone, but from a "whole series of interconnected crises unfolding in this
very moment".
Wisdom
vs. Power: He warns that humanity is approaching a threshold where our
collective wisdom must grow at the same rate as our technological capacity to
"affect the world," otherwise we will face severe consequences.
Pressure
Over Values: He hinted at internal and societal pressures that make it
difficult for core values to govern actions, stating he repeatedly saw
"pressures to set aside what matters most".
Distortion
of Humanity: One of his final projects focused on how AI assistants might
distort human reality or make us "less human".
His Next
Steps
Instead
of joining another tech firm, Sharma is returning to the UK to "become
invisible" for a time and pursue a degree in poetry. He expressed a desire
to bridge "poetic truth" with "scientific truth" to address
the modern human condition.
His
departure comes at a high-stakes moment for Anthropic, which is reportedly
seeking funding at a valuation of up to $350 billion.
Sem comentários:
Enviar um comentário