The Confidence Calibration Trap

Senior AI Engineer interview at Anthropic, and the interviewer asks:

“We’re bleeding money on inference. We want to build a 𝐌𝐨𝐝𝐞𝐥 𝐂𝐚𝐬𝐜𝐚𝐝𝐞 (𝐅𝐫𝐮𝐠𝐚𝐥𝐆𝐏𝐓) system, route easy queries to Llama-7B, and only send the hard stuff to GPT-4. What is the actual engineering bottleneck that makes this unreliable in production?”

Don’t say: “The latency overhead of calling multiple models.”