Smell that?
That's desperation.
Microsoft's GitHub next month plans to begin using customer interaction data – "specifically inputs, outputs, code snippets, and associated context" – to train its AI models. The code locker’s revised policy applies to Copilot Free, Pro, and Pro+ customers, as of April 24. Copilot Business and Copilot Enterprise users are …
I use it for my Kitty profiles.
But at work we have 2 gitlab instances, an internal for stuff that needs to run on our clusters and servers and an external one that we mostly use for personal projects.
I have nothing there, so that it will serve me well.
It also annoys me that they call the merge requests "Pull requests".
And er, antitrust. They should never have been allowed to buy GitHub (or LinkedIn), many of us said at the time that they will take everyone's private/proprietary code as their property, and now they are doing exactly that and more. (if they haven't already of course. This policy change could just be a fig-leaf to justify private-data-slurping that has already occurred)
Here's hoping that microslop will be forced to delete whatever plagiarism-model results from this. It is absolutely NOT OK to steal people's private repos (and even their code review styles) without permission.
And thanks for the settings link. Although I wouldn't be surprised if it is nothing more than an an additional bitfield in my data profile, treated with no more respect than a "do not track" marker in a HTTP request..
I wouldn't be surprised if it is nothing more than an an additional bitfield in my data profile, treated with no more respect than a "do not track" marker in a HTTP request
And, being Microsoft adjacent, it'll likely be arbitrarily reset to "Opted In" every third or fourth Github site code update.
I moved all my (and my companies') repositories over to Gitlab 6 months ago, which is arguably even late to the game, and deleted everything we had on GitHub. Of course nothing is ever truly deleted, but...
It's been 100% predictable (and was predicted) where all this has been going since GIthub got bought by Microslop. There was no other reason for MS to buy it other than to hoover up all your code for slop training and it was inevitable all your stuff would go to train their shit. Any exceptions for now are just part of the slow erosion, they can't last. Because literally the only thing Slopya cares about is slop. Good thing his parents named him that, huh?
"and deleted everything we had on GitHub"
A better tactic might be to restrict accounts so regular users that were allowed can't use it anymore and then being liberal with the Strychnine. Spend a happy weekend injecting malicious code into the commits. Since it's not meant to be accessed by anybody, no foul, right?
Neither company has a competent IT guy, no way we're getting an in-house repository. We wouldn't even have source control if it relied on these clowns. So I'd set the (shared) repository up on my personal computer except I'm remote... so if the company doesn't care, I don't care any more than moving from GitHub to GitLab. And the personal stuff I have there is actually public repositories I want people to see. Which, okay, means the AI bots will probably find it anyhow, but at least I'm not standing in one of those red light district booths in Amsterdam by staying on Microslop Github.
"You. Were. Warned."
What needs to happen is teaching. Warnings are sort of hit and miss with people. There's still plenty of people that haven't internalized that relying on a third party to make your business go is problematic if that 3rd party can't be easily replaced. For example, companies that have built web apps that only run on one or two browsers and then go Tango Uniform when the browser company pushes out an "update" and that app no longer works. If the app is just a tool to do something more efficiently, fine. If it's required for employees to conduct business, the company might be flailing to find a work around while customers are getting more pissed off and spending some effort finding alternative suppliers.
to any of the smelly old code of mine that I have dropped on github and forgotten to flush.
Should poison the spring from which Crapilot is drawing its buckets of wisdom.
Some of it is very old indeed like a /sbin/mount wrapper for SGI Irix 6.x that performed Solaris type loop back mounts (mount.lofs) — can see that being really useful. ;) As too some I2O stuff for an obscure Adaptec scsi adaptor twenty+ years ago.
No need to opt out. I stopped using GitHub Copilot after experimenting with it and a few different LLMs and their code suggestions for a couple weeks to deal with one particular issue I had. I didn't like the way I found it working, so I haven't used it for anything else - I consider it a fraudulent technology. Sure it spits out text super-fast, but it isn't sane text, and you have to spend more time fixing the mistakes it made than it would have taken to just do it from scratch without making mistakes in the first place.
Talk about a balloon waiting to be popped; this whole "agentic" mission is just more Bafflegab for the Peons who Pay The Bills.
And man are those bills ever getting big considering it doesn't even work as promised...
"and you have to spend more time fixing the mistakes it made than it would have taken to just do it from scratch without making mistakes in the first place."
That's for little things. With something more complex, it can be a slog to figure out what it's doing and debugging is next to impossible. Sure, maybe it found a more efficient way, but it's not how the vast majority of the industry would have gone about it so it's cumbersome, if not impossible, to support.
"That means they do what the hell they like with other people's data, right?"
If you read through most of the social media Terms and Conditions/EULA's, that's what they're saying. By signing up for Insta and posting things, you are giving them the rights to sell/transfer that content, modify it, use it internally, worldwide and perpetually with no credit or royalty. If they never intended to do any of that, why would they need to stick those terms in the agreement?
I've gotta ask: do you have side projects? And do you self-host the source code?
I have very mixed feelings about GitHub; I boycotted them for three years after M$ took over, but then I broke the boycott during the pandemic. Due to their sheer size and name recognition, GitHub offers the promise of relief from the loneliness/isolation that I felt, when I was struggling to find developers who would do code reviews and work with me collaboratively. (Granted, most people ignored my Pull Requests, but... where else am I to go, to meet open-source developers during a pandemic?)
OK.....so GitHub/Microsoft have actually told their users about the impending slurp.
But what about all those others (you know, ChatGtp, Perplexity, META, OpenAI, Google/DeepMind, ORACLE, SAP...........)
The "big data" slurp has been going on for YEARS.......Yup.....PETABYTES of OTHER PEOPLE'S COPYRIGHT material.........
.......now available from your local AI vendor.
Consent? Hah......you must be joking!
They just didn't tell you. Private repos (and everything else not easily accessible from the open web) may have been the main exceptions. Now those are fair game as well, I suppose.
The poster above is right. They must be really desperate if they believe this will help improve their products in any way.
And as we all know, genies don't like going back in. Second best to stuffing it back in is turn it into an idiot, and that's up to you what let the damned thing loose to begin with.
M$ wants to train on your code? Let it. Load in that crap from high school you wrote that was so bad the teacher didn't grade it, just stapled a fast food application to it. Write viruses designed to make an AI spread it around, overwrite backups and lock the machine. Write stuff that looks logical but twists back on itself, drawing more and more resources until it locks the machine while copying itself - and with a watchdog that restarts the process an hour after the machine comes back on. You get the idea.
If the genie can't go back into the lamp, make it into something nobody wants.
> And as we all know, genies don't like going back in. Second best to stuffing it back in is turn it into an idiot, and that's up to you what let the damned thing loose to begin with.
Can confirm genies don't like going back in. But when American voters let the genie out of the bottle, it hasn't exactly been a good thing that the genie is also a demented idiot.
But among the 39 posts commenting on the change at the time this article was filed, no one other than Martin Woodward, GitHub VP of developer relations, has really endorsed the idea.
And you expected anything less. It's like Steven Miller sycophantizing about the Orange-utan's immigration policy. What's telling is that he is the lone voice.
I am sure that many of you, like me, in an effort to get away from Copilot, have switch to other git providers like codeberg.org. Even so, if you still have a github.com account I suggest that you log in to github and follow the instructions to disable Copilot completely - or as completely as Microsoft will allow. This will register every so slightly on the Microsoft corporate "brain" that people don't like Copilot and it is an foolish business venture. At least we can hope.
must idiots to trust an opt out checkbox.
sorry.. there was a bug that ignored the checkbox...
1. keep all code on premises.
2. no vscode or other browser shell editors or application for own code.
3. no openclaw reading all your drives.
4. no Windows slopware.
5. no api tokens based API , only OAUTH.
6. no python, node based MCPs,
Says it all, thatks for the link, I've disabled everything the link takes me to.
But I moved my current and all future repos to Codeberg six months ago because, well AI trained on buggy content [not mine necessarily] will return buggy content, and it takes a human to understand the difference.
AI trained on private and copyrighted content is theft plain and simple, just because it can be done doesn't mean doing do is morally right.
Google, Meta et al don't give a fsck about morality, just money and content that is enshittified.