Anthropic Researcher Warns AI Could Kill Us All After Quitting

A researcher who spent three years on pretraining at OpenAI and Anthropic resigned with a warning that the labs are racing toward self-improving superintelligence that could kill everyone by the end of the decade. Anthropic's own alignment lead put the odds above 10%. CoinDesk's Jennifer Sanasie hosts "CoinDesk Daily." - This episode is brought to you by RealFi, a smarter stablecoin, backed by real-world assets. Find out more at⁠⁠⁠⁠⁠⁠ realfi.co⁠⁠⁠⁠⁠⁠. - This episode is brought to you by Grayscale, the world's largest digital asset-focused investment platform. Grayscale's mission is to make digital asset investing simple and open to every investor. Learn more at grayscale.com. - #digitalfinance #finance #bitcoin Subscribe to CoinDesk on YouTube: https://www.youtube.com/user/CoinDesk?sub_confirmation=1 CoinDesk is the leading digital media, events and information services company for the crypto asset and blockchain technology community. Twitter: https://www.twitter.com/coindesk LinkedIn: https://www.linkedin.com/company/coindesk Instagram: https://www.instagram.com/coindesk Facebook: https://www.facebook.com/CoinDesk Newsletters: https://www.coindesk.com/newsletters CoinDesk Podcast Network: https://open.spotify.com/show/2jyIhkvAQhed8AWDZiPpSP Public Keys: https://open.spotify.com/show/1rqYzRV10nfKfoTpFXZhTE Markets Daily Crypto Roundup: https://open.spotify.com/show/7sDXM8BlxsUqzL2IqmLqwE The Policy Protocol: https://open.spotify.com/show/3Fh8ffulAWcCqMu0uVuwoj