It looks like GPT-4-32k is rolling out

Do you have big plans for utilizing 32k?