No it's not. Spell checking relies on very precise rules. Fact checking relies on assessing context and, more often than not, the speaker's intention.
For example, let's take the statement "There are 300 million people in the US". Is this false? Well, it depends where I'm going with it.
If I was only relying on the figure as an approximation, I can say something like "therefore deaths from terrorism only affect 0.0x%, which is not something we should worry a lot about". This is obviously a factually valid argument, even though you might disagree with the conclusion b/c you think other facts are more relevant etc.
On the other hand, my argument can also be "and there were 300 million in 2007, therefore the population has been stagnant and we need to do something about it". This is factually quite false, since the population has in fact increased by ~18M.
Now should the original statement "there are 300 million people in the US" be rated true or false? A hostile fact checker may rate the first instance false, since there are actually 318+ million. A friendly fact checker will rate the second instance true b/c I obviously meant the rate of growth was low by historical standards etc.
I think this all a little besides the point and a bit of a red herring. The analogy, which is quite apt in this case (you can have fact-checking and something that is still overall misleading), much like how a spell/grammar-checker can ensure you have correct spelling & grammar and still a flawed sentence.
Your comment seems to nitpick at the flaws in the analogy rather than the issue itself. No analogy is going to be perfect at describing something and will break down under enough scrutiny.
You can put spell checking into an algorithm. You can't reduce fact checking to an algorithm. How is this besides the point? The algorithm they have is "whatever websites X, Y and Z say".
He's not talking about the process of fact checking, like you are. OP was using the analogy to say fact checking is not enough.
Here it is slightly different: just because you spell check a sentence doesn't mean it is grammatically correct. Just like: just because you fact check a document doesn't mean it is without bias.
Remember, we're talking about Google introducing fact checking. In this context, "fact checking is not enough" is a very different position from "fact checking is subjective".
Fact checking isn't subjective, though. You either do fact-checking or you don't. If you do fact-checking, you're either checking if the statement is factually correct or not. There is no real in-between and no real "this is true because I think it is."
At least from a human perspective. You're right in raising the concern that from a machine perspective this is hard to achieve, but again that's a red herring to the OP.
At some point the machine fact-checking relies on data input by humans. Or does the machine interpret data directly from cameras on the street to determine, e.g. that suspect A shot victim B with weapon C? Does it interpret a historical textbook and assess the veracity of its sources and claims? Or does it build a time machine to go into the past and acquire raw data to verify claimed facts?
> At some point the machine fact-checking relies on data input by humans.
But so does nearly every ML model? In the case of a spell checker it is using corpora made by humans. If the majority of humans start spelling words differently, then facts about the correct spellings change with them.
> Or does the machine interpret data directly from cameras on the street to determine, e.g. that suspect A shot victim B with weapon C?
If the military gets their way this will happen sooner than later. It is not technically infeasible to do activity detection from drone footage.
> Does it interpret a historical textbook and assess the veracity of its sources and claims?
Yes. Just like a journalist would when fact checking an article about WWII.
> Or does it build a time machine to go into the past and acquire raw data to verify claimed facts?
Raw data is both an oxymoron and a bad idea. Data is brought into existence by human-made measuring devices.
It is color? Or colour? Should my English dictionary include "haiku"? What about "teriyaki"? "Shamisen"? Is it "Internet" or "internet"?
---
Everything has some wiggle room, but both fact checking and spell checking rely on precise rules to a great exent. You just have to recognize when something is really an opinion. Many things on Politifact, etc. are inferences and conclusions composed of facts, but not facts themselves.
"Aaron Burr was vice-president of the US in 1803." Fact.
"The Sun convert light elements to heavier ones via fusion." Fact.
"It is legal for a licensed gun owner to open-carry in California." Falsehood.
"The US is in an economic depression." Too vague, can't be fact checked.
"The US has lower unemployment than in 2010, and higher unemployment than in 2000." Fact.
Dude. We're trying to talk about things in an absolutely neutral and objective way and you come up with that sentence? How can anyone ever prove what definition is "always used"???
The "unemployment rate" is a specific number produced by the government. Whether it's useful or accurate compared to other numbers (e.g. labor participation rate) is another matter.
Take the definition that is given by the relevant government agency, which is the BLS's U-3 measure.
(Now you might argue that U-6 is also a valid measure etc etc, but only a Breitbart writer or somebody willfully ignorant would confound unemployment with labour force participation.)
It's completely irrational to exclude "the percentage of people who aren't employed" from the set of possible definitions of "unemployment." Each of the various measures of unemployment captures meaningful information about the economy.
Labor force participation rate lets you look at important trends in the economy. For example, labor force participation went way up when barriers to women entering the work place were removed. Women didn't want to all be stay-at-home moms, but the economy didn't offer them any other opportunities. These days, drastically more of the population is going to post-secondary education, which tends to drive down the labor force participation rate. But that's really a paradoxical phenomenon when you think about it, because at the same time, the costs of such education are skyrocketing. Why do people eat the cost (including the opportunity cost) of going to college? Is it because of the love of learning? No, it's because the economy doesn't offer them alternatives.
Of course you can't count every student or retiree or stay at home mom as unemployed. Many don't want to be working. But it's equally irrational to assume (as the BLS does), that none of them want to be working.
That last one about unemployment isn't from the realm of checkable facts because the data are vast enough to be difficult to verify, and because the definition of unemployment isn't settled or universally agreed. And you can't rely on government figures to overcome the first problem, because of the second. ("Official" definitions of unemployment are swayed by political motivations and are known to under-report certain kinds of unemployment that legitimately should be counted.)
This is pretty naive. Do you think Google is now cross-checking news stories with the encyclopedia entries on vice presidents and nuclear fusion? Please go to factcheck.org, which is what they'll actually use. They're dealing with stuff that's almost exclusively like your item #4, and they'll definitely give it ratings. #3 and #5 are also much more open to argument than you want to admit here.
Wikipedia isn't a source of facts, but if you read your own link, you will learn that open carry used to be permitted, but since 2012 open carry of all firearms, loaded or unloaded, is prohibited.
"No license or permit is required to openly carry a loaded firearm in unincorporated areas where discharge is not prohibited by local ordinance."
...Maybe every county has adopted local ordinances prohibiting open carry in unincorporated areas?
"Prior to January 1, 2012, it was legal to openly carry an unloaded handgun in public. In October 2011, Governor Jerry Brown signed a bill that modifies the law on openly carrying an unloaded firearm to match the restrictions for openly carrying a loaded weapon.[63] Legislation was later signed by Governor Brown to expand these restrictions to long guns and shotguns, except while hunting."
This seems to be making the law for unloaded carry the same for loaded carry, and doesn't address the legality of loaded carry.
> "It is legal for a licensed gun owner to open-carry in California." Falsehood.
Except that it is true. Well, "licensed gun owner" isn't actually a thing at all (licenses aren't needed to own guns), but the natural interpretation of it as "gun owner who also has an appropriate license" makes the statement true, it's just that there are only limited places in California where you can be appropriately licensed for open carry of handguns (counties with a population under 200,000.)
"licensed gun owner to open-carry in California", as I read it, is "a gun owner licensed to open-carry in California".
Regardless, you can't get a license to open carry in California, as no such license exists, and open carry is illegal in California under California Penal Code Section 26350[1]. So, you may be able to legally carry a firearm in California, but unless you're a law enforcement, you may not carry openly. This has been adjudicated in court in Peruta v. San Diego[2], and the courts found that an outright ban on open carry for the citizenry comported with the second amendment.
So, as a private citizen, you might be able to carry a firearm, but you would have to do so in a concealed fashion, and only after being licensed to do so. In order to obtain a license, you have to apply with your county sheriff, they have to evaluate whether you are of good moral character, whether you have a legitimate reason for carrying a firearm, that you have completed 16+ hours of training, and that you are a resident of their county. Noting that 'self defense' is not a valid reason for issuance, most applications are denied, though that undoubtedly varies from county to county.
A fantastic rebuttal, and you're probably correct. I don't know what it means in the wake of Peruta, but your probably correct for firearms (like rifles, shotguns) that aren't capable of being concealed.
Note though that the parent refers specifically to handguns. Regardless, thanks for the addition/correction.
>Note though that the parent refers specifically to handguns.
Also there are limitations on the locations where open-carry of unloaded handguns aren't allowed. Specifically, [1] states those locations as:
(A) A public place or public street in an incorporated city or city and county.
(B) A public street in a prohibited area of an unincorporated area of a county or city and county.
(C) A public place in a prohibited area of a county or city and county.
...it would seem like you could open carry unloaded handguns in unincorporated counties where it isn't specifically prohibited. Are their any of those places in California? National Forest areas would be one area I'd first look at. A google search comes up with unincorporated areas:
> Regardless, you can't get a license to open carry in California, as no such license exists
Open carry licenses do exist, and you can get them, from the Sheriff of a County with population under 200,000; see Penal Code section 26150(b)(2).
> and open carry is illegal in California under California Penal Code Section 26350
26350 applies to unloaded firearms, and there are a whole slew of exceptions to it in sections 26361-26391, including, most critically, an exception for anyone permitted to openly carry a loaded firearm on the same circumstances (Section 26362) Which includes, naturally, people with open carry permits as described above. (see Section 26010)
Rather than linking each individual section mentioned above, I'll just link the Table of Contents for Division 5 of Title 4 of Part 6 the Penal Code, which covers all of them.
>Open carry licenses do exist, and you can get them, from the Sheriff of a County with population under 200,000;
It looks like the 30 California counties with population of less than 200,000 have a combined land area of 73,063 square miles, while the 27 counties with populations over 200,000 have a combined land area of 83,022 square miles. So about 47% of the state is composed of counties with a population less than 200,000.
There's either an error in one of the counts or a county left out somewhere in there, because California has 58, not 57, counties, but it probably doesn't change the percentage of land area much either way.
Good catch. I apparently was off by one in counting the over 200,000 population counties, but there are actually 28 counties in my spreadsheet, so the area should be correct (but someone might want to double check anyway).
The alternative explanation is that God created stars. And that He did it such that they seem to create heavier elements through fusion. But that's just an illusion. Because, you know, the Universe is only 5000 years old.
And that alternative will pass fact checking, unless they exclude literal readings of the Bible.
But as others have pointed out, choosing accepted fact checkers is subjective and contentious. Do we then need fact checkers about fact checkers?
Many claims have moved from conspiracy theory to fact in recent decades. Now there's a class-action case about CIA experiments with LSD, and the principal investigator will testify. Before Snowden's leaks, how many people thought that the NSA was intercepting so much stuff? Consider allegations about using poor people to study effects of terminal syphilis, measure Pu excretion rates for body-burden calculations, or look at symptoms of lethal full-body irradiation. Which of those (if any) are unconfirmed conspiracy theories?
Because some people believe them to be the ground truth? Any "fact" that contradicts their understanding of their favourite religious passage is obviously fake news. Dinosaurs? Fake. Evolution? Blasphemy. Round Earth? Idiocy...
Curious you drop the adjective "overwhelming" in your quote and confuse scientific theory with hypothesis.
https://en.wikipedia.org/wiki/Theory
"In modern science, the term "theory" refers to scientific theories, a well-confirmed type of explanation of nature, made in a way consistent with scientific method, and fulfilling the criteria required by modern science. Such theories are described in such a way that any scientist in the field is in a position to understand and either provide empirical support ("verify") or empirically contradict ("falsify") it. Scientific theories are the most reliable, rigorous, and comprehensive form of scientific knowledge, in contrast to more common uses of the word "theory" that imply that something is unproven or speculative (which is better characterized by the word 'hypothesis')."
Seems like a fair point to me. If science is really about reproducibility, then this is a fair question. How can we say with absolute certainty until we run a falsifiable experiment?
Note that I'm not disputing our understanding of how stars work. The point is whether our understanding is based on reproducibility.
Spell checking is distance between a word and the closest word in a set. But it's not like the set of allowable spellings for words is well-defined. You generally choose a dictionary (or /usr/dict/words) and call that canonical-enough. Which dictionary you use, and whether your spelling is actually a better choice for your context than what the dictionary says aren't generally considered. Even though they do fuzz the idea of what's properly spelled.
EDIT: I'm not trolling or trying to start a flame war on spellers. Just point out that almost every part of human life has some subjectivity to it.
Note that for many languages there is an institution (often governmental) that decides what the grammar should be like, and which constructions are allowed. That English has many different language authorities, each with their own opinion is more of an anomaly.
No it's not. Spell checking relies on very precise rules. Fact checking relies on assessing context and, more often than not, the speaker's intention.
For example, let's take the statement "There are 300 million people in the US". Is this false? Well, it depends where I'm going with it.
If I was only relying on the figure as an approximation, I can say something like "therefore deaths from terrorism only affect 0.0x%, which is not something we should worry a lot about". This is obviously a factually valid argument, even though you might disagree with the conclusion b/c you think other facts are more relevant etc.
On the other hand, my argument can also be "and there were 300 million in 2007, therefore the population has been stagnant and we need to do something about it". This is factually quite false, since the population has in fact increased by ~18M.
Now should the original statement "there are 300 million people in the US" be rated true or false? A hostile fact checker may rate the first instance false, since there are actually 318+ million. A friendly fact checker will rate the second instance true b/c I obviously meant the rate of growth was low by historical standards etc.