4 ms·

A notable feature here is "no training on your business data or conversations" which really shouldn't have to be a feature. (requests using the ChatGPT API alre

A notable feature here is "no training on your business data or conversations" which really shouldn't have to be a feature. (requests using the ChatGPT API already aren't trained upon)

andrewstuart2 3y ago
This is ChatGPT, not OpenAI's API with the gpt4 models. This is allowing your team to use chat.openai.com together, rather than having to build or deploy your own with the API.
- ashu1461 3y ago
  Yes but even in ChatGPT the training is not done ?
  - sjwhevvvvvsj 3y ago
    Be default they train on user data and cripple the product in various ways if you opt out.
    - alwa 3y ago
      For some time I thought the primary way to opt out was to disable chat history. That made it impossible to use custom instructions. But I later found the account-level opt-out request, which, assuming they respect the weakly-worded “ask us not to train on your data” language, preserves the product’s full set of features: https://privacy.openai.com/policies https://privacy.openai.com/policies
      - Roark66 3y ago
        There is literally nothing on that page (nor in the linked help article that describes how to export data) that says how to tell them to "not use my data for training" and not loose history at the least. If you know more about this feature please do tell.
        sjwhevvvvvsj 3y ago
        I also breaks the speak to chat in the app if you turn history off.
        swores 3y ago
        > 'There is literally nothing on that page that says how to tell them to "not use my data for training"...' You can opt out of ChatGPT using your conversations for future training, without disabling any features like convo history. Either some browser plug-in like an adblock is hiding the button from you, or you're not noticing and clicking it (I'm guessing the former [1]). For me, on iPhone, there's a black button with white text "Make a Privacy Request" which sort of hovers bottom centre of the page the way "Chat Live with Support!" buttons often hover. Click on that button to get to this - https://privacy.openai.com/policies?modal=take-control https://privacy.openai.com/policies?modal=take-control - which allows you to either delete your account, or: "I would like to: Step 1 of 2 Do not train on my content Ask us to stop training on your content" They then tell you it applies to content going forward, not stuff already done. But that's the opt out that doesn't require losing ChatGPT conversation history. [1] On iOS Safari with 1Blocker enabled, I could see the button without it being hidden as an annoyance or widget or whatever, however when I tried entering email for opting out to check it still works it gave me an error message that suggested adblock type things might be the issue. I opened the page in Firefox for iOS (so same browser engine as Safari, but without 1Block) and it worked with no error message.
        sumedh 3y ago
        Never knew about this link, I have requested not to train on my data but can we even confirm if they will honor it?
        swores 3y ago
        In addition to the usual risk any company has with breaking the law that a whistleblower might be brave and speak up, it could also come out in discovery during a court case, of which there are likely to be quite a few brought against them regarding what data they train on. The benefit of training on data of people they've explicitly agreed not to train on (which is probably a very small % of even paying users yet alone free ones) is unlikely to be worth the risks. They'd be more likely just to not offer the opt-out agreement option. But ultimately, we often can't know if people or companies will or won't honour agreements, just like when we share a scam of a passport with a bank we can't be sure they aren't passing that scan onto an identity-theft crime ring. Or course, reputation matters, and while banks often have a shit reputation they generally don't have a reputation for doing that sort of thing. OpenAI have been building up a bit of a shit reputation among many people, and have even got a reputation for training on data that other people don't believe they should have the right to train on, so that won't help get people to trust them (as demonstrated by you asking the question), but personally I still think they're not likely to cross the line of training on data they've explicitly agreed with a customer not to use.
sjwhevvvvvsj 3y ago
For some reason “Confidentiality Tax for Small Business” has less of a ring to it than “Teams”.
alwa 3y ago
Similarly you can opt your individual account out on the ChatGPT side. [0] Although by default they do seem to use both vanilla and Plus conversations for training. [0] https://privacy.openai.com/policies https://privacy.openai.com/policies
- MacsHeadroom 3y ago
  If you opt out you lose access to basically every feature you're paying for. No conversation history, no access to plugins, etc.
  - tmerse 3y ago
    Are you sure about that? Parent was not referring to the "APP -> Data controls -> Chat History & Training" Radio Button which would delete your history older than 30 days, but rather to a specific form you can use to opt out. Coincidentally I just used this form yesterday and got the confirmation about opting out.
    - creesch 3y ago
      Edit: This was previously a comment where I said I thought it did no longer show history. As it turns out the website version of chatGPT is just having a ton of issues. So it is very slow to load and when it does finally seems to load the sidebar doesn't show any content. After refreshing a few times the sidebar does show up. So if people are going through the steps now they might indeed think that they no longer have access to advanced features.
ParetoOptimal 3y ago
Does anyone really trust openai isn't training on their data given their views on copyright? It would make more sense for them to just train on it anyway.
- dvngnt_ 3y ago
  if there EULA says one thing and they do another that opens them up to huge liabilities. though anything is possible
  - ParetoOptimal 3y ago
    Well "huge liabilities"... until recently I haven't seen anything resembling something a company wouldn't just say "pay the fine and do it anyway, it'll be better for our bottom line".
- phh 3y ago
  My personal guess is that they don't put it in training data, BUT they still have human read what you send, to see 1. what are the innovative uses that they can try to copy/integrate; 2. optimize (both in score and throughput) for their consumer's usage.
ttul 3y ago
I don’t agree with you here. OpenAI should be free to train on your data assuming you agreed to that in the terms of service (and yes, you did). If they ask for a little more money in exchange for not having that valuable information in trade, that seems fair. If you want an entirely free and open LLM experience, you can also run one of the ever-improving open source models on your own hardware. But for many if not most companies, paying $25/mo per user for something as amazing as ChatGPT-4 is a bargain.
- unshavedyak 3y ago
  > I don’t agree with you here. OpenAI should be free to train on your data assuming you agreed to that in the terms of service (and yes, you did). If they ask for a little more money in exchange for not having that valuable information in trade, that seems fair. Yea, another way to word it would be to imagine that they _only_ had a more expensive "no train" option. Now ask if it would be okay to offer a lower priced but yes-train version.
johnfn 3y ago
Google trains on all your searches. Why is OpenAI held to a higher standard?
- dmonitor 3y ago
  OpenAI's product might just print out the info you put in verbatim to another user if asked politely
- matheusmoreira 3y ago
  Google should not be allowed to do that either.

A notable feature here is "no training on your business data or conversations" which really shouldn't have to be a feature. (requests using the ChatGPT API already aren't trained upon)

andrewstuart2 3y ago

This is ChatGPT, not OpenAI's API with the gpt4 models. This is allowing your team to use chat.openai.com together, rather than having to build or deploy your own with the API.

ashu1461 3y ago
Yes but even in ChatGPT the training is not done ?
- sjwhevvvvvsj 3y ago
  Be default they train on user data and cripple the product in various ways if you opt out.
  - alwa 3y ago
    For some time I thought the primary way to opt out was to disable chat history. That made it impossible to use custom instructions. But I later found the account-level opt-out request, which, assuming they respect the weakly-worded “ask us not to train on your data” language, preserves the product’s full set of features: https://privacy.openai.com/policies https://privacy.openai.com/policies
    - Roark66 3y ago
      There is literally nothing on that page (nor in the linked help article that describes how to export data) that says how to tell them to "not use my data for training" and not loose history at the least. If you know more about this feature please do tell.
      - sjwhevvvvvsj 3y ago
        I also breaks the speak to chat in the app if you turn history off.
      - swores 3y ago
        > 'There is literally nothing on that page that says how to tell them to "not use my data for training"...' You can opt out of ChatGPT using your conversations for future training, without disabling any features like convo history. Either some browser plug-in like an adblock is hiding the button from you, or you're not noticing and clicking it (I'm guessing the former [1]). For me, on iPhone, there's a black button with white text "Make a Privacy Request" which sort of hovers bottom centre of the page the way "Chat Live with Support!" buttons often hover. Click on that button to get to this - https://privacy.openai.com/policies?modal=take-control https://privacy.openai.com/policies?modal=take-control - which allows you to either delete your account, or: "I would like to: Step 1 of 2 Do not train on my content Ask us to stop training on your content" They then tell you it applies to content going forward, not stuff already done. But that's the opt out that doesn't require losing ChatGPT conversation history. [1] On iOS Safari with 1Blocker enabled, I could see the button without it being hidden as an annoyance or widget or whatever, however when I tried entering email for opting out to check it still works it gave me an error message that suggested adblock type things might be the issue. I opened the page in Firefox for iOS (so same browser engine as Safari, but without 1Block) and it worked with no error message.
        sumedh 3y ago
        Never knew about this link, I have requested not to train on my data but can we even confirm if they will honor it?
        swores 3y ago
        In addition to the usual risk any company has with breaking the law that a whistleblower might be brave and speak up, it could also come out in discovery during a court case, of which there are likely to be quite a few brought against them regarding what data they train on. The benefit of training on data of people they've explicitly agreed not to train on (which is probably a very small % of even paying users yet alone free ones) is unlikely to be worth the risks. They'd be more likely just to not offer the opt-out agreement option. But ultimately, we often can't know if people or companies will or won't honour agreements, just like when we share a scam of a passport with a bank we can't be sure they aren't passing that scan onto an identity-theft crime ring. Or course, reputation matters, and while banks often have a shit reputation they generally don't have a reputation for doing that sort of thing. OpenAI have been building up a bit of a shit reputation among many people, and have even got a reputation for training on data that other people don't believe they should have the right to train on, so that won't help get people to trust them (as demonstrated by you asking the question), but personally I still think they're not likely to cross the line of training on data they've explicitly agreed with a customer not to use.

sjwhevvvvvsj 3y ago

For some reason “Confidentiality Tax for Small Business” has less of a ring to it than “Teams”.

alwa 3y ago

Similarly you can opt your individual account out on the ChatGPT side. [0] Although by default they do seem to use both vanilla and Plus conversations for training. [0] https://privacy.openai.com/policies https://privacy.openai.com/policies

MacsHeadroom 3y ago
If you opt out you lose access to basically every feature you're paying for. No conversation history, no access to plugins, etc.
- tmerse 3y ago
  Are you sure about that? Parent was not referring to the "APP -> Data controls -> Chat History & Training" Radio Button which would delete your history older than 30 days, but rather to a specific form you can use to opt out. Coincidentally I just used this form yesterday and got the confirmation about opting out.
  - creesch 3y ago
    Edit: This was previously a comment where I said I thought it did no longer show history. As it turns out the website version of chatGPT is just having a ton of issues. So it is very slow to load and when it does finally seems to load the sidebar doesn't show any content. After refreshing a few times the sidebar does show up. So if people are going through the steps now they might indeed think that they no longer have access to advanced features.

ParetoOptimal 3y ago

Does anyone really trust openai isn't training on their data given their views on copyright? It would make more sense for them to just train on it anyway.

dvngnt_ 3y ago
if there EULA says one thing and they do another that opens them up to huge liabilities. though anything is possible
- ParetoOptimal 3y ago
  Well "huge liabilities"... until recently I haven't seen anything resembling something a company wouldn't just say "pay the fine and do it anyway, it'll be better for our bottom line".
phh 3y ago
My personal guess is that they don't put it in training data, BUT they still have human read what you send, to see 1. what are the innovative uses that they can try to copy/integrate; 2. optimize (both in score and throughput) for their consumer's usage.

ttul 3y ago

I don’t agree with you here. OpenAI should be free to train on your data assuming you agreed to that in the terms of service (and yes, you did). If they ask for a little more money in exchange for not having that valuable information in trade, that seems fair. If you want an entirely free and open LLM experience, you can also run one of the ever-improving open source models on your own hardware. But for many if not most companies, paying $25/mo per user for something as amazing as ChatGPT-4 is a bargain.

unshavedyak 3y ago
> I don’t agree with you here. OpenAI should be free to train on your data assuming you agreed to that in the terms of service (and yes, you did). If they ask for a little more money in exchange for not having that valuable information in trade, that seems fair. Yea, another way to word it would be to imagine that they _only_ had a more expensive "no train" option. Now ask if it would be okay to offer a lower priced but yes-train version.

johnfn 3y ago

Google trains on all your searches. Why is OpenAI held to a higher standard?

dmonitor 3y ago
OpenAI's product might just print out the info you put in verbatim to another user if asked politely
matheusmoreira 3y ago
Google should not be allowed to do that either.