The Guardian's 'scary' new robot author didn't know the word 'robot' originates from Czech

The British daily's robot author made its debut yesterday, but the paper may want to employ some robot fact-checkers, too

Jason Pirodsky

Written by Jason Pirodsky
Published on 09.09.2020 13:56 (updated on 09.09.2020)

Yesterday, British daily The Guardian published an article written entirely by artificial intelligence in an effort to convince us that we have nothing to fear from robots. Or that maybe we do.

“A robot wrote this entire article. Are you scared yet, human?” the (human-crafted) headline reads. The article’s author, OpenAI’s “powerful new language generator” GPT-3, was tasked to write a 500-word essay on why robots come in peace.

But before switching over to a full staff of robot writers, The Guardian might want to employ some robot fact-checkers, too.

GPT-3 has the wealth of our combined knowledge at it’s proverbial fingertips, and yet it didn’t seem to know where the word for it’s namesake, ‘robot’, comes from. Worse, it carelessly spread misinformation. No thanks, Guardian: we have enough human writers for that.

“Robots in Greek [sic] means ‘slave’,” GPT-3 writes.

“But the word literally means “forced to work”. We don’t want that. We need to give robots rights. Robots are just like us. They are made in our image.”

While the rest of the information is somewhat accurate, if awkwardly phrased, plenty of us humans know that the word ‘robot’ originates from Czech. Not Greek.

The term was first used by Czech science-fiction Karel Čapek for his 1920 play R.U.R (Rossum’s Universal Robots). Čapek borrowed the name from the Czech word ‘robota’, meaning serf labor, for the name of his automaton creations.

Screengrab via The Guardian

‘Robota’ is also commonly used in other Slavic languages as a general term for labor or hard labor.

Upon the term’s addition to the Oxford English Dictionary, Čapek wrote a letter to its editors in 1933 explaining that he was not actually the inventor of the term: he originally wanted to use the term laboři, from the Latin for labor, for the automatons, but his brother Josef Čapek convinced him to go with the Czech term instead.

And the rest is history. At least until GPT-3 and its robot compatriots rewrites it.

While Guardian editors added the [sic] designation to GPT-3’s text, it’s unclear if that was for the dodgy grammar or the blatant factual inaccuracy, left uncorrected and unexplained.

A link to a Today I Found Out article, presumably added by human editors, does indeed correctly attribute Čapek and the Czech roots of the word robot, while an additional piece of info identifies the word ‘android’ as being of Greek origin.

PRIVATE PROPERTIES

Apartment for rent, Flatshare, 25m<sup>2</sup>

Apartment for rent, Flatshare, 25m2

Haštalská, Praha 1 - Staré Město

Apartment for rent, 2+kk - 1 bedroom, 50m<sup>2</sup>

Apartment for rent, 2+kk - 1 bedroom, 50m2

Jindřišská, Praha 1 - Nové Město

Apartment for rent, 1+1 - Studio, 42m<sup>2</sup>

Apartment for rent, 1+1 - Studio, 42m2

Vlašská, Praha 1 - Malá Strana

Apartment for rent, 2+kk - 1 bedroom, 51m<sup>2</sup>

Apartment for rent, 2+kk - 1 bedroom, 51m2

Francouzská, Praha 2 - Vinohrady

Maybe GPT-3 just got its wires crossed. Or maybe A.I. suffers from the same problem a lot of humans do: with so much information available to us and so much of it contradictory, how do we know which information to trust?