People say things like “load the data” then “it won’t load.”
They are referring to a data file.
Sometimes they say “we need more data” and they really mean more data points.
I rarely hear the term “datum” but it’s not technically incorrect, just a bit out of touch.
What is a datum? Any piece of data will have multiple properties, so a singular piece is really a data point, which implies a number of bits of datum. I know it’s hard to understand this stuff.
This used to be completely true. It definitely is not always so in modern American English. But we might be just fighting over prescriptive versus descriptive approach to linguistics.
The Economist is a British publication, and generally tends to follow stricter grammatical rules than what I assume is your American English. "Datum" is singular and "data" is plural, so The Economist have used "are" instead of "is".
You are taking a prescriptive approach (which is fine), but it is just as reasonable to take the descriptive approach to linguistics and say "in modern American English 'data' is not countable, and it is used the way 'water' is used". It is self-consistent and (currently) grammatically correct to say "this data is important" like you would say "this water is wet". I guess this implies "datum" is archaic now. At least for the linguistic/cultural bubble in which I exist.
Even in British English usage as an abstract mass noun with singular verbs is acceptable (possibly even the norm[1]).
Usage as a plural noun is an Economist house style choice[2].
It's hard to tell how well the Economist sticks to that style choice. A Google site: search, with a date range of the last 10 years, shows no usage of "datum" (except for the article about the subject).
An Google site: search for usages of "data is" in the last 10 years returns plenty of results.
While opinion is split on this, these days I don't think you can appeal to strictness alone to justify 'data' being plural, given the way usage has already gone. People (like me) who say 'data is...' are simply using data as a mass noun (https://en.wikipedia.org/wiki/Mass_noun), which seems to be a reasonable direction to be taking the language. I may be influenced by the fact that I usually see 'datum' used to refer to a reference point or marker in some arbitrary space, and thus in my mind is now another word separate from 'data' (albeit with shared etymology). If I had more than one datum, I'd call them 'datums', though I've yet to test how acceptable this would be in a formal setting.
There is a lot of American English, but your assumption is wrong. It would be uncommon to say 'data are valuable' in England too.
If this was ever a debate, it was decided a long time ago. Grammarians may choose to argue the point (as if our language was defined in Ancient Rome) but saying "data are" sounds like simple-minded country bumpkins struggling to come to grips with it all.
Been a while since I looked. My understanding is many of these techniques such as SMPC are useful only when operations are linear (eg no exponents or tan functions) and on simple branches. Others like homomorphic encryption require operations on the whole data set.
I think the best best for some of these workflows will be differential privacy on large datasets with strict privacy budgets.
Still highly valuable, but make sure you understand the limitations and risks if you use these techniques.
What questions could never be answered safeguarding privacy? I can think of one -- the taxes paid by the person in the data with the most assets, if you know who the country's richest person is.
No, that's precisely my point -- maybe the richest person wants to hide that they paid nothing in taxes. If you can ask about the richest person, then they can't.
This piece is mostly about enclaves, which are going to be little-r revolutionary.
Enclaves allow for nearly any kind of processing of an individual's data without making that data directly available to either the algorithm provider or the cloud host. The enclave's public key is provided to the data holder. Their data is encrypted, sent to the enclave, decrypted with the enclave's private key, processed in the enclave, then results are encrypted with the data holder's public key for return to the holder. Other kinds of IO generally not permitted in the enclave, preventing leakage.
A lot of ergonomic details remain to be worked out but the underlying machinery is quite interesting.
24 comments
[ 2.8 ms ] story [ 65.0 ms ] thread"DATA ARE valuable".
Is 'data' like water or like sheep? :) Can whatever-it-is be counted (like sheep) or not (like water). Eg:
Water are valuable. (Wrong!!)
Sheep are valuable. (Perfectly acceptable, phew!)
For me, data is like water.
'Data is valuable'.
'A piece of data is valuable'.
'Lots of pieces are valuable'.
'Data are valuable' <- sounds all wrong!
They are referring to a data file.
Sometimes they say “we need more data” and they really mean more data points.
I rarely hear the term “datum” but it’s not technically incorrect, just a bit out of touch.
What is a datum? Any piece of data will have multiple properties, so a singular piece is really a data point, which implies a number of bits of datum. I know it’s hard to understand this stuff.
Usage as a plural noun is an Economist house style choice[2].
It's hard to tell how well the Economist sticks to that style choice. A Google site: search, with a date range of the last 10 years, shows no usage of "datum" (except for the article about the subject).
An Google site: search for usages of "data is" in the last 10 years returns plenty of results.
[1] https://trends.google.com/trends/explore?date=all&geo=GB&q=%... [2] https://www.economist.com/graphic-detail/2012/07/13/data-or-...
If this was ever a debate, it was decided a long time ago. Grammarians may choose to argue the point (as if our language was defined in Ancient Rome) but saying "data are" sounds like simple-minded country bumpkins struggling to come to grips with it all.
Nit: "Is" is singular because of "piece," not "data." As in, "the pieces of data are valuable."
My educated guess is that <at best> UN is using homomorphic encryption for basic analysis...
or they have achieved a breakthrough and have FHE /s.
I think the best best for some of these workflows will be differential privacy on large datasets with strict privacy budgets.
Still highly valuable, but make sure you understand the limitations and risks if you use these techniques.
But i can't think of any regressions like that.
Enclaves allow for nearly any kind of processing of an individual's data without making that data directly available to either the algorithm provider or the cloud host. The enclave's public key is provided to the data holder. Their data is encrypted, sent to the enclave, decrypted with the enclave's private key, processed in the enclave, then results are encrypted with the data holder's public key for return to the holder. Other kinds of IO generally not permitted in the enclave, preventing leakage.
A lot of ergonomic details remain to be worked out but the underlying machinery is quite interesting.