Recently, a company specializing in translation claimed that the phenomenon of “technological singularity” was going to happen very soon. However, this declaration is rather daring insofar as until now, the concept does not yet have exact criteria.
How to define technological singularity?
First mentioned in the 1950s, the notion of technological singularity received a precise definition some fifteen years later, under the impetus of the British statistician Irving John Good. This singularity would occur when the invention of artificial intelligence would be able to trigger a runaway at the level of technological innovations, and this, after having surpassed the human in all areas. However, this same runaway could cause unpredictable changes in human society.
In short, we are talking about new generations of increasingly intelligent AI, ultimately giving rise to super-intelligence. who would definitively surpass humans. This implies that from this precise point, the AI would have become so intelligent that it alone would enable future progress. Thus, the role of the human being himself could be called into question. There is little in a statementscientists from the translation company Translated have claimed that this technological singularity is about to arrive by 2030.
“Many artificial intelligence researchers believe that solving the language translation problem is the closest thing to producing Artificial General Intelligence (AGI)”can we read in the document.
The status of Artificial General Intelligence
Researchers point out that language is a natural human ability that remains quite difficult for machines to assimilate. Thus, it is possible that the ability of AIs to catch up and then eventually overtake humans in terms of translation could be a motive for achieving AGI status. AIs would thus clearly become capable of carrying out any job instead of a human.
« By moving from automated estimations to measures of human cognitive effort, we reassign the assessment of quality to those traditionally in charge of the task: professional translators”explain the scientists.
The researchers based their hypothesis on Time to Edit (TTE), a metric representing the time it takes for human experts to correct translations made by an AI compared to human-generated translations. However, if in the future, the translation generated by AI becomes more advantageous than the human translation, the famous technological singularity tipping point could be achieved.