Source: https://front-end.social/@fox/110846484782705013
Text in the screenshot from Grammarly says:
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.
The only way to avoid Grammarly using your data for AI is to pay for 500 accounts
Protip: You can also simply not use grammarly.
Let’s ignore the ethical implications of this for a moment.
Grammarly is training it’s AI off of the poorly written grammar of it’s users that it has to already adjust?
It seems like this would be a flawed set of training data. It’s training on what it already either produced or on something written by someone who may not have used proper grammar in the first place.
Am I to expect this AI will improve over time?
Depending on how they’re training it, they’re likely looking at when grammarly corrections were accepted or rejected and the context around that. That’s what I’d be using from the dataset anyhow
Remember that people have said GPT4 is getting dumber because of interacting with humans.
On average, people’s grammar is correct, kind of by definition.
Grammarly is basically a keylogger anyway, with every stroke send to their servers. Why ANY business even allows their employees to use this is really beyond me.
I work in the field of data science and I really get why data is needed and why it makes sense to collect use data but enforcing this and now allowing to opt out free of charge is simply not okay
For anyone looking for an alternative, LanguageTool though not perfect, has shown itself to be far more privacy respecting
Came here to say this. I’ve been using LanguageTool for a while, but they’ve also recently started implementing AI into the product.
Using AI itself isnt a problem if the engine they’re using is completely proprietary. But they’re likely using some third-party engine to send the data to. But I’d love to be proven wrong by them open sourcing the code for it so I can take a look at it myself.
I’m quite certain this policy is illegal in some jurisdictions (read: the European Union).
So… grammarly is problem to help you write an email or a document you will send via gmail or publish online?
Or you use grammarly for private diary?
Also they are not selling data back to you, they are using it to train a model and provide better service.
I don’t get it, from when is reading a book and using that knowledge unethical and illegal.
Most people consider email private. Plus a lot of people use Grammarly for work documents.
The same people will use Gmail without batting an eye and we all know what Google has been doing for years with emails in Gmail.
It’s quite funny that now that there are “AI” that everyone can use, people get all worried about their data being used to train them but nobody cared before when it was Google or amazon using them to train their models.
There is a second way, legend has it. The ancient ones tell a tale of the one that does not use the service, and does not train someone else’s shitty models for free.
The third way is like the second way: we learn to write good without crutches.
Why is it necessary to post this to so many different cimmunities?