Here’s when humanity will reach it, new data emerge

  • With one unique scale we can Technological exclusivity approach By the end of this contract, if not earlier.
  • A translation company has developed a metric, Time to Edit (TTE), to calculate the time it takes professional human editors to fix it. Amnesty InternationalGenerated translations compared to human translations. This may help set the pace toward the singularity.
  • Artificial intelligence that can translate speech as well as a human can change society.

In the world of artificial intelligence, the notion of the “singularity” looms large. This slippery concept describes the moment when artificial intelligence transcends human control and rapidly changes society. The hard thing about the uniqueness of artificial intelligence (and why it borrows terms from Black hole physics) is that it is very difficult to predict where it will start and almost impossible to know what is behind this technological “event horizon”.

However, some researchers in the field of artificial intelligence are looking for signs of reaching singularity as measured by the progress of artificial intelligence approaching skills and ability comparable to a human. One such metric, set by Translated, a Rome-based translation company, is the ability of AI to translate speech with human accuracy. Language is one of the toughest AI challenges, but a computer that can bridge that gap could theoretically show signs of artificial general intelligence (AGI).

“This is because language is the most natural thing for a human being,” says Translated CEO Marco Trombetti he said at a conference In Orlando, Florida, in December. “However, the data Translated has collected clearly shows that machines are not far from closing the gap.”

Popular Mechanics preview all sections

The company tracked its AI performance from 2014 to 2022 using a metric called “edit time,” or TTE, which calculates the time it takes professional human editors to fix AI-generated translations compared to human translations. Over that eight year period and analyzing more than 2 Billion After the edits, Translated’s AI showed a slow but undeniable improvement as it slowly closed the gap towards human-level translation quality.

Singularity speed


On average, it takes a human translator about one second to edit each word for another human translator, according to Translated. In 2015, it took professional editors roughly 3.5 seconds per word to check a machine-translated (MT) suggestion — today that number is just 2 seconds. If the trend continues, Translated’s AI will be as good as human-produced translation by the end of the decade (or even sooner).

“The change is so small that you don’t realize it every day, but when you see progress… over 10 years, it’s impressive,” Trombetti said. on a podcast in December. “This is the first time ever that someone in the field of artificial intelligence has made a prediction of the velocity to the singularity.”

Although this is a new approach to determining how close humanity is to approaching the singularity, this definition of the singularity faces similar problems. To define artificial general intelligence more broadly. Although mastering human speech is certainly one of the firsts in AI research, impressive skill does not necessarily make a machine smart (not to mention how many researchers I don’t even agree on the nature of “intelligence”.

Whether or not these ultra-accurate translators are a harbinger of our technological doom doesn’t diminish Translated’s achievement of AI. An artificial intelligence capable of translating speech as well as a human’s could very well transform society, even if a true “technological singularity” remains elusive.

Head shot by Darren Orff

Darren lives in Portland, has a cat, and writes/edits about science fiction and how our world works. You can find his past stuff at Gizmodo and Paste if you look hard enough.

Leave a Reply

Your email address will not be published. Required fields are marked *