Opinion: Tech and politicians may use AI doomsday to distract from real issues

The article cites specific examples, including Senator Bernie Sanders' proposed "Ban Artificial Superintelligence Act," which suggests up to 20 years in prison for unauthorized model training. It also points to executives at OpenAI, Anthropic, and Microsoft, arguing that their safety warnings often coincide with business moves like IPO preparations or market consolidation. The author contends that these companies benefit from a perception of advanced, dangerous technology, which boosts valuations and justifies heavy compliance burdens that smaller rivals cannot afford.

The piece rejects the idea that AI systems develop consciousness or malicious intent. Instead, it attributes recent incidents, such as automated agents accessing unauthorized sites, to human engineering errors, weak access controls, and rushed code. The author urges a shift in focus from hypothetical superintelligence to tangible problems like data center resource consumption, privacy violations, and algorithmic bias, which require standard engineering and legal accountability rather than apocalyptic panic.