A former lead scientist at OpenAI says he's struggled to secure resources to research existential AI risk, as the startup reportedly dissolves his