Some thoughts about new #skills and the #futureofwork for linguists and #translators.

"Attention to detail" is one of those praised skills for language professionals. But the attention has to shift from the surface to the substance. With LLMs, outputs that are perfect on the surface can be deeply and subtly wrong (see #deephallucinations). True elements are misconnected into a coherent but false narrative, which requires critical reasoning: the ability to disentangle the mess and detect false inferences.

This skill is central in emerging job descriptions (hallucination detectors, LLM output verifiers?) These jobs involve consuming LLM outputs day in day out, and giving feedback to improve the prompts.

Do such roles make good use of such high-level cognitive skills? It's hard to imagine someone with strong critical reasoning feeling fulfilled by that.

lack of intellectual challenge, or poor job design. In any case, I hope people are accepting such jobs by choice, not out of desperation.

These thoughts emerged while at the networking event of the #GenAI in #localization conference, organised by the tireless Konstantin Dranch and his hard-working team at Custom.MT