How to destroy your AI competition ...?

API
1

The other day I posted a link to our (open source) framework that’s using OpenAI’s API. We were using the word “ChatGPT” to describe our product, specifically we wrote things like “Create a ChatGPT-based chatbot”.

I was first politely told that I had to remove the word “ChatGPT” from our website. Then the conversation escalated a bit, and I was told that unless I complied I could have my OpenAI API account revoked.

We’re a small family-based company, with 3 of our 4 employees literally being family (I’m the father, working with two of my children) - The fourth employee is a 25 year old female Indian entrepreneur, amazing girl.

Now ignoring the fact that I only wanted to share our open source platform with the users here, the fact that you might have your OpenAI API account revoked for such a thing SHOULD be quite scary for everybody else in here, simply because not everybody is as honest as me.

Imagine you have a company, you’re one of the 10,000+ companies on the planet that’s using the word “ChatGPT” to describe your product. One of your COMPETITORS sets up a “burner email address”, registers here, and starts bragging about YOUR product, that’s using the word “ChatGPT” in its description. Then these same two mods first politely tells you to “update your website and remove the word ‘ChatGPT’ and use ‘Powered by OpenAI’ instead”. At which point you tell them to go føkk themselves, so they elevate the discussion to employees at OpenAI, and you get your OpenAI API account revoked, for being in “violation of the community guidelines”, and your product is ipso facto destroyed as a consequence, and you need to lay off your employees, people go without income, and everything goes down the drain, because you built your business on OpenAI’s API and violated their “Don’t use ChatGPT community guideline part”.

Of course your competitor that posted the link to YOUR website is at this point quite happy. He eliminated his competitor, and gets to double his revenue.

To see the magnitude of this problem, go to ProductHunt and search for the word “ChatGPT”. I don’t want to post a link, since I’m simply not that guy myself - But I assume the employees at OpenAI can clearly see the problem here, and clarify to the rest of us if we really risk losing our OpenAI API account because of posting a link to a website where the word “ChatGPT” is being used to describe the product or not.

I personally actually REMOVED the word from our product description, and exchanged it with simply “AI”, because after all, I need to make sure my children have an income, in addition to our lovely Indian 4th employee.

However, I lost 95% of our organic search volume from Google (TM!) as a consequence, so I would REALLY appreciate if somebody OFFICIAL from OpenAI could enlighten us here, since it’s obviously a problem …

1 Like
2

The moderators on this forum don’t work for OpenAI. You shouldn’t comply with requests like that unless you can verify that they’re authorized by OpenAI. That said, I believe there might be a misunderstanding here. Moderators on the forum board enforce rules that don’t necessarily apply to your website.

In any case, it doesn’t sound like something you should be too concerned about. Just make sure to follow the ToS and you don’t really have anything to worry.

3

I found the post you were talking about:

This isn’t official OpenAI communication. If you know that you’re following the Brand Guidelines then you should continue what you’re doing. Users and volunteer moderators might chime in to give their two cents though :slight_smile:

3 Likes
4

Not an OpenAI employee, but you need to understand what you agreed to when you created an account.

Part of the Terms of Use includes adhering to the Brand Guidelines.

The relevant bit here:

Powered by OpenAI. A general, simplified way to describe products that are built on OpenAI models or multiple OpenAI models. Where possible, we recommend and prefer that you use the “Powered by OpenAI” badge provided above. Please do not say “powered by ChatGPT” or “built with ChatGPT” to refer to products built with our APIs.

Your text,

Is, at least to me, clearly a violation of this term.

The Terms of Use also includes a section on termination of service,

Termination and Suspension

Termination. You are free to stop using our Services at any time. We reserve the right to suspend or terminate your access to our Services or delete your account if we determine:

  • You breached these Terms or our Usage Policies.
  • We must do so to comply with the law.
  • Your use of our Services could cause risk or harm to OpenAI, our users, or anyone else.

So, while none of us who engaged with you work for OpenAI, all of us are developers of some form or another and are familiar with the terms of service.

None of us are personally invested in what you decide to do or not do with your website, rather someone noticed it and gave you a friendly “heads up” because they care about this community and its members and wanted to help you avoid any potential fallout from violating the brand guidelines.

Your response to that was… Not great.

So while you’re free to follow this advice,

I think you’d do better to,

  1. Understand the advice was offered to help, not control you.
  2. Review the situation with as objective of a lens as you can muster (or enlist a trusted, knowledgeable outside point of view).
  3. Make the decision you think is best or right given your new understanding.

Since you’ve already made the changes, it’s a bit of a moot point, I just wanted to point out that none of this should have been terribly shocking.

Access to OpenAI models requires abiding by the terms of service, the terms of service require adhering to the brand guidelines, and one possible consequence of not doing so is suspension or termination of the account you use to access the models.

This is no different than any other service offered in the world—if you break their rules they can, and often will, take their ball and go home.

5 Likes
5

I am not an OpenAI employee but a community moderator. As @elmstedt mentioned, you need to adhere to the Brand Guidelines. First, ensure you are in official communication with OpenAI staff members (for email communication, they use the domains openai.com and openai.intercom-mail.com).
The chance of speaking with an official from OpenAI is very slim. The other community moderators here are knowledgeable and can provide excellent assistance; they are your best bet.
I suggest not escalating the conversation further. You have limited influence over the situation and OpenAI has the power to do much more than merely revoking your API account.

I am going to ping @Foxalabs as he would probably know what to do best in this situation.

1 Like
6

There’s a few misunderstanding’s and misinterpretations going on here, and I’d like to clear those before we continue.

What I told you was that the specific phrasing “Put a ChatGPT Chatbot on your website” is against the brand guidelines, and that you can’t use the name “chatGPT” to sell your products.

I did this as a courtesy, because I could see that you’ve spend significant amount of effort on making a site that actually looks good.

As previously mentioned

And included both a request for a cease & desist letter, threats of doxing and a request to be banned.

As far as I’m aware, non of these things have happened.

OpenAI may choose to terminate your account if, it violates their brand guidelines or the TOS, it has nothing to do with this forum in particular.

7

I would think for SEO purposes you can have it on the page, but not in the domain or trying to claim you’re officially OpenAI.

Best bet would be to reach out to help.openai.com and wait for an official response about how you want to use the term.

5 Likes
8

Indeed, which seems to solve 50% of the problem, notice how many companies have a blog on their site now? :wink:

The other 50% of the problem is the attitude, but that’s not my problem :rofl:

2 Likes
9

Weird to get into specific details about your business matters. Pls don’t try to pander for brownie points.

Wouldn’t care if your business was saving sea turtles.

I think the advice given was good. Any representative will advise you not to jaywalk.

If the referee isn’t looking tho :eyes:.

I have also seen way too many popular sites advertising themselves as ChatGPT bots. So :man_shrugging:

3 Likes
10

Lmao, I think it’s generally good advice to remember that search engines might find whatever you post about your business online, so not telling people to “go F* themselves”, and acting like a brand ambassador, is probably in your best interest :rofl:

1 Like
11

Parts of this reply is quoting N2U too, since I didn’t want to spam with “a bajillion” comments. You know which is which I presume …

It didn’t feel like that, but I understand there might be some communication breakdown here. For the record, I did not threaten anyone, at least not according to my understanding - I did however offer to help with making sure this part of the ToS would be known by others by offering to redact the full name and email address if I got an “official warning letter”, which I personally think would have been a better solution.

none of this should have been terribly shocking

Actually it was very shocking. I suspect everybody else affected is also quite shocked, something you can understand the gravity of by searching for the phrase “ChatGPT” at ProductHunt, or Google for that matter.

I did this as a courtesy, because I could see that you’ve spend significant amount of effort on making a site that actually looks good.

Thank you, it’s sometimes difficult to see the difference :slight_smile:

threats of doxing

Errh, no! Redacted is the keyword here, besides, how can it be doxing when you don’t post with your full name or anything else like that? The closest thing I could find that’s “personal” for you in your profile for instance is that you’re in Denmark. If OpenAI employees sent me an official email, explaining their ToS, I assume it would be in their best interest to have more people read it anyways …

That is literally the definition of “helping” - However, if you or N2U experienced is as a “threat of doxing” I apologise, that was definitely not my purpose …

it has nothing to do with this forum in particular

That is not how it came across when you explained it the first time. I suspect you agree here too.

Anyways, I see there are some misunderstandings here, and I take the high ground and I apologise. However, we are on the internet, bad things are happening on the internet, and I’m the only one here who is not anonymous so far. Sometimes it’s wise to chose words carefully - Especially when dealing with people who do not hide their true identity, for obvious reasons.

But I apologise and I accept the explanation, and I’m grateful for it :slight_smile:

12

I didn’t do that for the record, I use it as an example of what could happen, if your original answer was correct …

People are impersonated every single day on the internet. If you go to places like 4chan, they’re even “swatting” others to “get revenge”. Imagining somebody posting a link to the competitor here to apply some damage is not that far fetched actually. Even though I don’t use these schemes myself, for weird reasons I happen to be quite good at identifying them …

… 41 years of software development tends to result in that I presume :smiley:

13

I owe you and @elmstedt and apology. I apologise, but I completely freaked out. It wasn’t your fault, you did me a service, and once I got to sleep on it, I realised that - Thank you :slight_smile:

8 Likes
14

I’m happy that we figured it all out in the end :laughing:

Thanks, I asked you to clarify this, because the phrasing you used might be misinterpreted as such without further explanation, I’m happy this wasn’t the case.

Now that we’ve got that cleared up, allow me to educate the wider community, as a few people have already asked me privately:

what does it actually take to get banned from the community forum, and what happens then?

To get banned you have to be persistently annoying, not only towards moderators, but also towards the rest of the community. Being threatening, spammy or breaking the guidelines in some other way, may get a post flagged and removed, but this is usually the worst that could happen. I can count the number of actual ban’s on one hand, and an OpenAI employee will always be involved in that process.

6 Likes
15

So this is probably the top piece of copy pasta i’ve read on this forum :rofl: :rofl:

Dude, you had people whos salaries were depending on you so your best decision making was to instead of change a string of text which should take like…3 minutes, you chose to tell members of OpenAI to f*** themselves, and then you blamed them after doing so for terminating your account?

I think you mean unhinged

Also, what is meant by this? :face_with_raised_eyebrow:

17

@thomas11 can we close this topic? I think there is nothing left to add.

2 Likes
18

“Close”? What do you mean? Are posts closed when they’re resolved? Isn’t this like a standard forum where stuff just stays up? If the post should be closed, then sure, go for it I’d say - However, it’s kind of an important topic, and I suspect it would be in other peoples’ benefit to read it also in the future …

1 Like
20

Even after it is closed, people can still read the posts in this topic.

2 Likes