Tech analyst Ben Thompson dismisses the 'clearly absurd' concept embedded in AI watermarking
Thompson took particular issue with Anthropic adding a watermark even if Claude was used only for proofreading.
Anthropic's decision to watermark AI-generated content has sparked debate within the tech community. Tech analyst Ben Thompson, a popular writer for the Stratechery newsletter, does not support the plan, arguing that it goes too far and not far enough. Thompson believes that requiring AI companies to be transparent about their models' outputs is missing the mark and would be overburdensome.
He questions the concept of what constitutes "fake" content, pointing out that even if an LLM generates a true fact, it may not matter if a human creates a false narrative without a watermark. Thompson particularly objects to Anthropic watermarking Claude models for proofreading, as there would be no way to differentiate between human edits and AI-generated text.
He notes that this requirement could unfairly label human-authored content as AI-generated. OpenAI, which already watermarks images and audio, plans to expand watermarking to text output, though the specifics are not yet clear. The EU AI Act, which Anthropic cited as the reason for watermarking, requires AI companies to mark AI-generated outputs in a machine-readable format.
However, Thompson finds the concept of watermarking AI-generated text absurd, likening it to insisting that a pen be credited as the author of a written work.
Written by urgent.news from Business Insider's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
