Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
You can assure me that none of my data has been used to train the model? Are you sure? Are you completely sure?

What has it got to do with me? I just told you what their legal, contractual pledge is.

No, but the company assures you of that. They are working with many major companies who obviously have stringent governance requirements around this.

They’d be opening themselves up to unbelievable massive reputation damage and to endless lawsuits if they violated their public commitment. It’s in the contracts. Of course their clients need to be sure their propriety data is secure.

You’re free to whatever you like of course, but I don’t think the concern is warranted and if you think the tiny bit of code you represent is of so much value to OpenAI - lol.

They’re not alone in offering this guarantee btw. Anthropic don’t train on user/customer data AT ALL.
 
What has it got to do with me?

No, but the company assures you of that. They are working with many major companies who obviously have stringent governance requirements around this.

They’d be opening themselves up to unbelievable massive reputation damage and to endless lawsuits if they violated their public commitment. It’s in the contracts.

You’re free to whatever you like of course, but I don’t think the concern is warranted and if you think the tiny bit of code you represent is of so much value to OpenAI - lol.
Wait, so the model can use my data that’s out there for training. With zero money coming to myself??

If the tiny bit of code to OpenAI is of much value? But if it uses a tiny bit of code after training on millions of people’s hard work… that’s a lot of code all together

It’s like saying, oh don’t worry that transaction in error on your account for a dollar isn’t anything, but if it’s done to everyone on the planet? That’s billions of dollars 😂
 
Wait, so the model can use my data that’s out there for training. With zero money coming to myself??

If the tiny bit of code to OpenAI is of much value? But if it uses a tiny bit of code after training on millions of people’s hard work… that’s a lot of code all together

It’s like saying, oh don’t worry that transaction in error on your account for a dollar isn’t anything, but if it’s done to everyone on the planet? That’s billions of dollars 😂

I’ve just told you what the policy is. I think the default for non Teams/Enterprise/API use is that training on your data (as a user) is on by default.

But it can be switched off in seconds if you’re not comfortable with that. Do whatever you like, literally nobody cares.

I’m just pointing out the options and saying that OAI obviously have glaring incentives to not violate this pledge. Imagine the lawsuits from their corporate clients.
 
Last edited:
Really don’t want to judge about that stuff or anyone using it. Coding since 3 decades my opinion is: if you want or need to code, then simply learn coding, then practice, practice, practice. These tools look nice, but should be used - in opposite to the impression they would help or support inexperienced users or beginners - by those who know what they do, want and at least - but most important to me - UNDERSTAND what they are proposed from these bots. Used as some sort of brainstorming or tipping in the right direction for specific functions, methods or algorithms one doesnt implement regularly, ok. But you have to understand if the proposed code is efficient, safe, and doing what you want to achive. Copy Paste and try compile some snippets you have no clue about, better dont try start or learning coding that way. Old guy over and out :]
It's good to learn to code that way but some of us (e.g., scientists) need functioning code without having the time or money to spend years learning coding. I've been doing various forms of coding for decades, but tools like this allow me to produce excellent, workable, and correct code in much less time than it used to take me. That way I can get more done in less time. This frees me up to plan out more sophisticated analyses of my data.

If coding is a full-time job for someone, spend the time to really know how to code. What tools like this allow is the democratization of coding. Some people are disturbed by that, others are not. You can end up with insecure, junk code this way (especially if people don't really know what they are doing), but that was also the case with the 'old way'. That's not really an issue for me because my code is almost always for a single project and then will never be used again. It does get posted on Github for others to potentially use, but the number of people in the world who would ever use most of my code for research is likely less than 10 (and probably 0, not counting my use of it -- but if people want to check and replicate my work, the code is there for people to do that).
 
Last edited:
  • Like
Reactions: erthquake
I’ve just told you what the policy is. I think the default for non Teams/Enterprise/API use is that training on your data (as a user) is on by default.

But it can be switched off in seconds if you’re not comfortable with that. Do whatever you like, literally nobody cares.

I’m just pointing out the options and saying that OAI obviously have glaring incentives to not violate this pledge. Imagine the lawsuits from their corporate clients.
Yeah, but what data have they trained the models on prior to you being a user. My data is already out there.

These companies don’t explicitly state what these models have been trained on.

So my data could have been used whether or not I’m a user

😀
 
Yeah, but what data have they trained the models on prior to you being a user. My data is already out there.

These companies don’t explicitly state what these models have been trained on.

So my data could have been used whether or not I’m a user

😀

That’s a different issue.

The concern you expressed was about being a ChatGPT user and having your clients code being used for training, that’s the issue I was replying about.
 
That’s a different issue.

The concern you expressed was about being a ChatGPT user and having your clients code being used for training, that’s the issue I was replying about.
What if I’ve already got code out there, and it’s used without the correct acknowledgment

And that should concern everyone. What were these models trained on?
 
I've the latest (Version 1.2025.057 (1741115522)) version (and Team plan) but it seems the app can only read VSCode content and produce something which I have to manually copy-paste back. There is no button or switch to bring the generated code back.
 
Really don’t want to judge about that stuff or anyone using it. Coding since 3 decades my opinion is: if you want or need to code, then simply learn coding, then practice, practice, practice. These tools look nice, but should be used - in opposite to the impression they would help or support inexperienced users or beginners - by those who know what they do, want and at least - but most important to me - UNDERSTAND what they are proposed from these bots. Used as some sort of brainstorming or tipping in the right direction for specific functions, methods or algorithms one doesnt implement regularly, ok. But you have to understand if the proposed code is efficient, safe, and doing what you want to achive. Copy Paste and try compile some snippets you have no clue about, better dont try start or learning coding that way. Old guy over and out :]
The amount of code being generated with little to no thought about security...😬
 
  • Like
Reactions: xpxp2002
Are the apps like this using Apple Intelligence or leveraging the NPU? Or at they just regular programs which need to connect through the API to run everything off the mother ship and pop back in?

I ask, because I don't fully understand the benefit of full integration of "AI" in to the actual desktop experience if the same thing can be done via API.

My cynical accusation being that Big Tech is quick to jump on blanketing all kinds of things under "AI," including things that exist to silo customers for their own data collection purposes.
 
Swift Assist is still missing on Xcode even though Apple said will release it at end of 2024.
 
I used grok (I know Elon…) yesterday and built a viable app within 2 hours.

It’s helped me understand coding because it explains what it does fully when errors present, and it fixes those errors. What a time to be alive.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.