Cardboard robot, please avoid water
It says something about lockdown that upon reading "cardboard robot" my first thought was "That would be a good idea for a videogame".
Lawyers at Google have been criticised by the company’s AI researchers for watering down negative language in their academic papers. A leaked email memo to Reuters revealed that words like “concerns” were changed to “considerations”, “dangers” became “risks”. Other words like “breach” or “sensitive” were frowned upon too. The …
Upon reading 'Cardboard Robot' the image that my brain grabbed was of an animated depiction of a forlorn robot made of cardboard boxes being told a definition positive thinking by a blue-eyed anime-style girl in a classroom.
The reason for this specificity is that I watched just such a scene in part 6 of Adam Curtis's new series on iPlayer, I Can't Get You Put of My Head.
When most researchers write a paper, the language and words they use are usually well considered and intended to convey as accurately as possible, their conclusions up to the point of writing.
When someone else comes along and rewrites those papers in a way that devalues any of the researcher's statements, the paper is no longer fully valid because it doesn't represent the conckusions arrived at.
Lawyers may deal with words but few of them have anything better than a glancing knowledge of most of the things they may address from time to time other than the law.
Lawyers should stick to screwing people in court not second guessing the intent of researchers.
As an academic who writes papers I couldn't agree more. I spent ages today agonising with a co-worker over how we were reorganising a sentence because my original sentence was crap to read, but his sentence was subtly less correct. We eventually converged on a better sentence. The point is thought, this was a discussion over improving clarity of the writing while maintaining accuracy. This was not about PR.
The problem is not that Google's legal and PR teams police their academic output. It is that a company with a clear bias and motivation for a particular result is allowed to claim it is doing academic research. If Google really cared about the research it could fund these same researchers to work in academia and give them full freedom over their publications. The thing is Google does research because it is good PR to do research, and so it can attempt to control the academic dialogue. Journals that allow them to publish, and academics that work for them have sold out the very concept of academia. To cry foul when the obvious happens, displays either obvious naivety or total hypocricy.
Right. It seems like the "academics" just realized what Google has been doing for 20 years with every "service" they offer. Google's search is the the foremost example of watering down, rewriting, altering, etc. for corporate profit. I'll end with this: If you complain about censorship in journalism, well you have to first find the articles to complain about, so how are you finding those articles?
That is not a legal team. That is Goebble's PR team hard at work.
There is no legal reason to replace the 'concerns' with the word 'considerations'. Legally, that has no justification.
Propaganda-wise, though, it is a very good decision.
Except that, irony, the propaganda department is working against some of the best engineers in the field, who have chosen their words with utmost care.
Team Goebbel doesn't stand a chance.