That is probably true in many case, buy to be fair there are some jobs that have had huge impacts due to LLM’s, like translation.
These jobs have been changing quite a lot before this AI bubble mainly because advances in speech-to-text, but I see the LLM’s as final step. The translator need doesn’t fully disappear, but the workflow changes quite drastically and some labor heavy parts are going away.
Note: I am not talking as AI being some hype AGI, just the LLM tech, which is basically advanced auto complete.
At my work we just did a project that included translating a website for multiple regions for a company. We used a translation service that doesn’t use humans. The Belgian, Dutch, German, French, and Italian teams complained that the translation was extremely weird and they had to manually overwrite the automated translations for the majority of the site (at least dozens of thousands of words) before launch.
Using a bad translation/transcript as base for professional translator is still better than nothing. Like I said, translators are still going to be needed, but lots of the heavy manual work can be now automated.
Also often when very domain specific language is used, the translation made by human can be bad, because they don’t know the proper terms. Of course good professional translators will ask these. It is also something that must be done with these dummy LLM models, you cannot just throw text into it and expect good results.
That is probably true in many case, buy to be fair there are some jobs that have had huge impacts due to LLM’s, like translation.
These jobs have been changing quite a lot before this AI bubble mainly because advances in speech-to-text, but I see the LLM’s as final step. The translator need doesn’t fully disappear, but the workflow changes quite drastically and some labor heavy parts are going away.
Note: I am not talking as AI being some hype AGI, just the LLM tech, which is basically advanced auto complete.
At my work we just did a project that included translating a website for multiple regions for a company. We used a translation service that doesn’t use humans. The Belgian, Dutch, German, French, and Italian teams complained that the translation was extremely weird and they had to manually overwrite the automated translations for the majority of the site (at least dozens of thousands of words) before launch.
We’re still a ways off, judging by that anecdote.
Using a bad translation/transcript as base for professional translator is still better than nothing. Like I said, translators are still going to be needed, but lots of the heavy manual work can be now automated.
Also often when very domain specific language is used, the translation made by human can be bad, because they don’t know the proper terms. Of course good professional translators will ask these. It is also something that must be done with these dummy LLM models, you cannot just throw text into it and expect good results.