What Is the “Nothing at Stake” Problem Unique to Some Early PoS Designs?

The "nothing at stake" problem arises because in early PoS designs, a validator could vote on multiple conflicting chains at no additional cost (no "stake" is lost). This meant validators had a rational incentive to hedge their bets and support every potential chain split, which undermines consensus and security.

Modern PoS systems solve this with mechanisms like slashing.