We develop data sets to train our algorithms so that we can improve the services we provide to customers like you. We have devoted significant time and resources to developing methods to ensure that these data sets are anonymized and de-identified.
To develop these data sets, we sample snippets of text at random, disassociate them from a user's account, and then use a variety of different methods to strip the text of identifying information (such as identifiers, contact details, addresses, etc.). Only then do we use the snippets to train our algorithms-and the original text is deleted. In other words, we don't store any text in a manner that can be associated with your account or used to identify you or anyone else.
We currently offer a feature that permits customers to opt out of this use for Grammarly Business teams of 500 users or more. Please let me know if you might be interested in a license of this size, and I'II forward your request to the corresponding team.
Per their website premium includes "Unlimited sentence paraphrasing powered by A.I." so I'm not sure they're an appropriate alternative to avoid the "AI" bullshit.
I can avoid it like I've avoided cryptocurrency and NFTs. And it may be a "tool," but it's one built on the theft from and unpaid labor of tens of thousands of independent creators, and is nigh wholly controlled by corporate interests bent on eliminating those same independent creators whose data they stole to make their "tools." It should not exist. Not until it can be made in an ethical manner without harming the creatives necessary to make it.
I don't buy the theft argument. Was reading books to my daughter to help them learn how to read theft? When we were working on parameters in the 60s to help a computer identify a balloon vs. a dog, was that theft? The corpulent side of me says if you put something out in public spaces, people are going to learn from it. If you don't want that, don't share it.
But even beyond that, parameters of learning are not copying, they are examples to develop data points on. Or in the case of imagery and something like stable diffusion it is math formulas developed in the 40s on how to make noise and then reverse that. Is that copying or theft?
I am willing to have the argument that AI is full of pitfalls. And that corporate control is not a good thing. I am struggling to see this theft.
It isn’t theft because the technology fundamentally steals. It’s theft because the people in control of the technology fundamentally steal.
I’m not talking about basement dwellers with a 3090 either. People using their m2 to generate lactating joe Biden fanfic aren’t the problem like multi-billion dollar companies taking advantage of the webs openness to train models that will be used to sell generative services replacing the creators of the stuff they were trained on are.
It’s the enclosures all over again.
Now when people speak out about it they’re called luddites and we don’t have the historical literacy to say “yes, I will smash this and any mill used to oppress me”.
I still do not see that as theft. Or at least no different than theft of labor like a company store.
Corporate dominance, commercialization, exploitation, something along those lines. But that is the same as everything else, AI is not specifically the issue.
Then again I was listening to Knowledge Fight and frankly the fact that people will believe DNA has antennas, or that a team of people on Real World cannot solve "what is 27 divided by 3" does not leave me much hope for us anyways. They tried and ran out of time saying it was unsolvable. Maybe we get what we deserve.
people struggling for a way to express how massive incredibly powerful companies are literally building technology that will take their livelihoods away aren't gonna develop a vocabulary for the exact thing that's happening.
they're stealing. it's theft.
it doesn't matter that the precise thing happening isn't what we would legally call theft. the people saying ai is theft believe that their only chance to keep the corporate users of the technology from destroying their future and leaving them with nothing is to mobilize now using language that everyone understands. so they're calling it theft.
those people are wrong btw, they can't win even a tiny victory against the entire economic system that's decided the way itll extract profit from the tech sector is by automating labor.
thats the difference between which side of the spear youre on.
if you got the point youre yelling "ahh, dont, youre literally killing me! please, if you have any humanity in you save me from this monster!" if you have the stick end youre yelling "oh fuck off, the severe hemorrhage is killing you, stop lying! check out what happens when i twist this sucker!" when youre on the sidelines youre just eating popcorn and arguing about minutiae like us.
to your last point, i'd be more worried about historical literacy. a calculator can act as a bandage for lack of numeracy, no machine will bridge the gap in understanding that pedantry mobilized against justice sails through.
So you are saying automobiles stole the whip makers jobs. Or movable type stole the scribes job. Or the Word processor stole the secretary's job. I simply do not buy the argument that changing technology is theft, it just is what it is.
Its like you are conflating two arguments.
I fail to see the theft. I only see change. Livelihoods have changed since forever. What we do from here is a problem, but it is not theft any more than everything else. I mean unless you want to go back to the source of all the problems: agriculture which kicked off the whole damn mess in the first place.
Im not saying any of those things. I went to great lengths to express explicitly and through use of historical parallel that what people are angry about is an attempt to cut them out of the productive process and reduce the quality of their lives.
the examples you tried to put in my mouth are laughable and betray a deep misunderstanding of history.
scriveners existed alongside printers for a hundred years that i know of for myriad reasons and still exist today. there was a tradition of offering both longhand and printed services at shops and during the expansion of the printing press one way that scriveners dealt with the new technology was by writing letters for people and performing recording duties at events. if what you said was true we simply wouldn't see longhand documents after 1500.
a better parallell with automobiles would be coachbuilders, but of course cars didn't come with bodies for seven decades during which you'd go to the coachbuilder and have a custom coach built to your specification on top of the rolling chassis. that's still done today for heavy equipment and specialty stuff. if we wanted to take the traditional turn of phrase it'd be that the internal combustion engine put buggywhip makers out of business overnight but that process actually took the better part of fifty years. even though the market for whips dried up over that time, whip makers were skilled producers of leather goods, and found plenty of work some of which was in the outfitting of the cars themselves. we can look at the production records of these various industries and see that it took decades for these changes to happen and they didn't amount to an upending of the labor relation.
the last example you use is just absurd. the electronic word processor found a market even as computers in offices rose to dominance over electronic typewriters specifically because they were simpler and cheaper than computers and allowed administrative staff to handle digital files and formats. even today secretaries use the computer to perform a variety of duties that would have been done in the past with typewriters, carbon copy machines and electronic word processors. those jobs didn't go away, they were intensified. If what you said was true there would be a drop in secretary jobs when the word processor comes out, or when the word processing software suite comes out. as it turns out, secretary jobs don't peak till 2015. so something different is going on there. it's worth noting that everyone classifies secretaries and administrative assistants together, so if youre using the definition from the 40s then you might have a point.
and again, i never said any of those things, i'm just bushwacking your crappy arguments to pluck the low hanging fruit that the person conflating things here is you.
I have consistently held that the use of the open internet to train models that will be used to sell generative services is theft of the commons in the same way as the enclosures. I also drew a parallel between the people calling AI theft now and the luddites of yore.
you even acknowledged this fact by saying it's like theft of labor.
Yes I was referring to the 40s. Another real world example was the loss of a job as a data entry person when the clerks/order takers/agents were not allowed to actually use a computer. Those jobs went away. I still cannot see this as theft because that word has specific meaning and I do not think it applies. A tragedy of the commons that says people cannot be creative without compensation? So what about the compensation. That is my whole point. The being tied to money for survival I suppose is theft of free will, but that happened long before AI came about.
And that is the issue: AI in corporate hands does not steal anything. It just changes things, and gives us more of the same bullshit we already have.
And saying I have crappy arguments is not helping you case to make up mythical ideas about how AI is somehow stealing.
The whole system is built on exploitation. I don't see you boycotting luxury clothes, diamond, rare metal that are made by exploiting someone from a third world country to inhuman levels. Ah, yes. It could affect people you know, It's immoral now. Am tired of this hypocrisy.
I'm pretty sure most tools like this have to use ai to some degree to be more effective than something like Microsoft Word. I think the issue is more whether it's opt in or not to include your own data.
I took a quick look at this and it seems that the server portion of this product is open source but the apps such as extensions are not. I'm not saying it's bad or even that it's a red flag. I just felt like I should point it out.