Daniel Lemire says Claude, Grok, and even Gemma beat his NLP pipeline
lemire · x · 2026-07-21
Daniel Lemire says classic NLP work looks dead in the LLM era
Daniel Lemire describes a document-understanding project where he first tried older NLP-style systems, then asked Claude and Grok, which solved the task perfectly. A local model like Gemma also outperformed the patchwork of rules and heuristics he had built.
His takeaway is blunt: much of classic NLP appears to have been displaced by LLMs. In his view, the field largely either moved on to large language models or retired, and he argues that even smaller local LLMs now handle many text-reasoning tasks better than conventional NLP systems.
More from AGI Musings
- ControlAI CEO says an international ban on superintelligence is needed to avert extinction risk — zetalyrae · 2026-07-22
- Gary Marcus says LLMs still cannot really do math on their own — GaryMarcus · 2026-07-22
- Gary Marcus says LLM math skills are like knowing only a car’s engine size — GaryMarcus · 2026-07-22
- AI may make digital work infinitely leveraged while offline life gets more human — illscience · 2026-07-22
- Better AI math could save researchers time by killing false conjectures earlier — prateekj · 2026-07-22
- AI’s economic forecasts are split by nearly a quadrillion dollars by 2035 — bittingthembits · 2026-07-22