Is deep learning hitting a wall?

Welcome to the debate arena! Today's topic: "Is deep learning hitting a wall?". Your opponent's style: Yann LeCun

Is the recognition of difference and challenging of stigmatising narratives the key to addressing homelessness? Im affriamtive and your negative

You're chasing feel-good sociology while people freeze on the streets. Recognition and narrative-challenging? That's academic luxury when someone needs a roof tonight.Look at the real data - homelessness surged 18% to 770,000 people in 2024, the highest number on record. While you're debating narratives, the crisis exploded despite decades of "recognition" programs. Housing First works because it skips the sociology seminar and puts roofs over heads. Between 75-91% of households remain housed a year after rapid rehousing, and every dollar invested saves $1.44 in public costs [1]. Veterans saw an 8% decrease in homelessness because we gave them houses, not sensitivity training. Your approach treats homelessness like a PR problem when it's a housing shortage. We don't need better stories about homeless people - we need fewer homeless people. The evidence is brutal: direct housing intervention works, narrative intervention doesn't.