What is Singularity with AI and humanity, and when will it happen?
What is Singularity with AI and Humanity? What does it mean? Experts believe that we may reach the singularity within just a few years.
By one metric, we could reach technological AI singularity by the end of this decade.
A translation company has developed a metric, Time to Edit (TTE), to measure the time it takes human editors to fix AI-generated translations.
A machine that can translate speech as well as a human could have a profound impact on society.
The concept of “singularity dominates artificial intelligence.” This slippery concept refers to when AI surpassed human control and rapidly transformed society. The tricky thing about AI singularity (why it borrows terminology from black hole physics) is it’s challenging to predict where it begins and almost impossible to know what lies beyond this technological “event horizon.”
What is Singularity with AI and Humanity
AI researchers, however, are looking for signs that we are nearing a singularity when AI is approaching a level of skills and abilities comparable to human intelligence. According to Translated, a translation company based in Rome, one such metric is the accuracy with which an AI can translate speech. AI faces several language-related challenges, but a computer that can close that gap could theoretically exhibit signs of Artificial General Intelligence (AGI).
“Language is the most natural thing for humans,” Translated CEO Marco Trombetti said at a December conference in Orlando, Florida. “However, the data Translated collected shows that machines are not that far away from closing the gap.” This is What is Singularity with AI and Humanity and how it works.
To track the AI’s performance from 2014 to 2022, the company used a metric named “Time to Edit,” or TTE, it measures the time it takes human editors to fix AI-generated translations. Over those eight years and analyzes more than 2 billion post-edits, Translated’s AI displayed a slow but undeniable major improvement as it slowly closed the gap toward human-level translation quality.
Approximately one second is required for a human translator to edit each word of another human translator, according to Translated. It took professional editors about 3.5 seconds in 2015 to review a machine-translated (MT) suggestion — today, it takes just 2 seconds. By the end of the decade, Translated’s AI will be as good as human-produced translation. That seems to explain What is Singularity with AI and Humanity
Trombetti said on a podcast in December that even though the changes are so small, they are impressive when you see progress over a period of 10 years. In this paper, someone has predicted the speed to singularity for the first time ever in the field of artificial intelligence. So again, this a great example of what is singularity with AI and humanity.
Although this is a next-level approach to quantifying exactly how close humanity is to approaching singularity, this definition of singularity runs into the same problems as identifying AGI more broadly. Although perfecting human speech is indeed a frontier in AI research, it doesn’t really make a machine intelligent (not to mention how many researchers don’t even agree on what intelligence is).
Translated’s AI accomplishment is not diminished by the possibility that these hyper-accurate translators are harbingers of our technological doom. Although the true “technological singularity” remains elusive, an AI that can translate speech and a human could change society very well.
We may not be convinced that a true transformation in the power of human thinking will ever be achieved. Our past tech developments have usually followed S-shaped curves. Predictions made at the beginning often don’t stand up to reality. As well as this, without humans learning to become better thinkers and problem solvers themselves, how can technology truly help us? Speed isn’t a substitute for quality; machines can just reach their conclusions quickly – right or wrong.