Worry About Misuse of AI, Not Superintelligence

The piece argues that while bold AGI timelines persist, the immediate risks in 2025 come from AI misuse: lawyers relying on fabricated outputs, the rapid spread of non-consensual deepfakes enabled by accessible tools, the "liar’s dividend" undermining authentic evidence, and companies selling unreliable AI decision systems; it cites recent cases, platform guardrail failures, and emerging legislation, warning that distinguishing real from synthetic content will become increasingly difficult.