A well-known philosopher and historian Yuval Noah Harai has penned a comment for The Economist, decrying the advent of AI as the end of human history. He talks about language as the foundation of human culture and civilization, and AI as a rising dark power that (according to Harari) has already hacked its very fabric. He warns that AI’s uncanny ability to mimic human speech will lead (and in many instances has already led, as Elon Musk has discovered in the process of auditing Twitter) to humans’ interaction regarding the most divisive issues with bots, not actual people. And this, in turn, leads to endless manipulation, because while you can convince another human of your point of view, you are powerless against a bot, which can be programmed to repeat its take on things ad nauseam, and in the process hone its message so that it stands a chance of convincing you. (Not to mention the devaluation of representation when “likes” don’t come from real people.)

Unfortunately, Mr. Harari doesn’t quite practice what he preaches. He warns of the danger of people seeing AI as an oracle that can be consulted on each and every thing, but elsewhere in the text he himself deifies AI. Meanwhile, the AI that he talks about is nothing more than a very large generative language model, not a sentient being that he perceives it to be. It can be an excellent assistant, when used properly and with professional oversight, but it does not have an intellect or a consciousness, despite everything the hype is about.

If we stop deifying AI after falling victim to the media hype, and see it for what it is, if we use critical thinking instead of uncritically echoing “opinion leaders” and PR persons looking for funding, we will see that it is an instrument that can relieve us of many mundane tasks. But can it hack our democracy and culture all on its own? Not unless it has guidance from the humans, or unless humans believe the hype. And this is true of all the other facets of AI’s application. The truly impactful texts need human authors and truly persuasive messages conveyed and created in a target language by human professionals. They may use AI assistance to speed up their work, but they are the ones who create the message and put the final touches and the stamp of approval on any AI-generated text.

_____________________________________________________________________________________________________________

Send your comments to rd@logrusglobal.com with YOUR case, opinion, example, problem or inquiry on this fast-changing, magic technology in the translation industry.