Funding for independent AI alignment research — LessWrong