Useful Is Not Sufficient (tante.cc)

Linus Torvalds, the leader of Linux kernel development, recently addressed criticism of large language models (LLMs) on the Linux Kernel development mailing list. He argued that AI should not be dismissed as an “anti-AI project” and emphasized its proven usefulness, stating that LLMs are just tools that have become indispensable in various contexts. However, Torvalds acknowledged broader questions about the economic implications of AI, reinforcing that while LLMs have utility, this does not justify their uncritical acceptance. The significance of Torvalds' statement lies in the ongoing debate within the AI/ML community about the ethical implications of technology that is deemed “useful.” He cautions against assuming that usefulness alone is a valid rationale for incorporating AI without considering its broader societal impacts. By drawing comparisons to controversial practices such as child labor, Torvalds challenges the industry to hold itself to a higher standard, suggesting that just because something serves a purpose does not automatically make it ethically acceptable. His perspective underscores the need for critical discourse around the integration of AI technologies within open-source projects, highlighting the responsibilities of developers to maintain ethical and political considerations in their work.