Yes, I read the disclaimer.
But if nobody of the company replies to my support questions, via help chat and email, what should I do.
I subscribed for the api to use gpt4, but I always get the error, if I use my key, “there is no service like GPT4”.
We, my employer and me, although couldn’t figure out, how he could pay for GPTplus and I’m allowed to use it.
I have a lot of understanding for the situation of OpenAI, there overwhelming growth in the last few month, all the media hype, etc.
But if you are not able to deliver the products you offer, don’t offer.
First: ChatGPT Plus has nothing to do with GPT-4 API other than it comes from the same company. You don’t buy access to GPT-4 you are selected to get it.
Second: your API Access to GPT-4 is not granted the moment you create an Account. It takes some time (not long ago you needed at least one completed payment of 1$ for API usage in last months as well). They roll out access but not everyone gets it immediatly.
If you wanted to buy access you got to contact sales and ask for your own machine. I think I’ve read something like 1.2 million $ per year. And I am pretty sure that even those guys won’t get that thing the next day they pay the money. And on azure we had to wait a months to get an appointment with the AI team and had to wait another ~6 weeks to get GPT-3.5. And now we would need to make another appointment there to request GPT-4 but also with a waittime of 4 weeks.
I got GPT-4 even without being a plus user back then. So I can confirm it has no relation. I registered for the wait list and got selected (without a notification on why I was selected too - just got a mail).
Took ChatGPT Plus later to get a chance to become plugin developer.
So, all you can do is wait. You can try prompts on GPT-4 in ChatGPT already.
I wouldn’t beat yourself up over it. They haven’t really made it very clear to be honest. They have GPT-4 and ChatGPT-4 named pretty similar things and it can be confusing unless you have been hanging around for a while and listening. But if you have just decided to jump in and use it, they have not made it clear that the two GPT-4s have no relation between each other.
In general, they have really poor messaging, which is hilarious considering how many times I’ve seen Sam Altman tell startups about how to shape their visibility, messaging, and user flow.
Not sure I agree with that assessment, it has been a very rapid rise in popularity, a rise so quick it was at least an order of magnitude faster than anything prior. No system put in place would be able to cope with that. Accouriding to interviews 80% of Sam Altman’s time is spend on onboarding new staff to the team to help level up the support to all users. There are also initiatives like the community champions who will be trying to bridge that gap somewhat. I get your view on there not being enough people to assist with account related issues and that is frustrating, I don’t see what else you could do.
When millions of people join a service in rapid succession there is only a limited response to that, you onboard staff as fast as possible until such time as the demand is met. This has been the case for lots of companies that experience sudden growth, it’s a difficult thing to manage. You have on the one hand restricting users from accessing services due to insufficient support for them, and that will cause complaints from people wanting to access the services, or you can allow everyone to use the services and then get complaints about there being insufficient staff to support them… it’s a catch 22 situation that I don’t see any easy solution to, other than time and more people.
Well, you can have poor messaging because you can’t keep up. I’m not saying there is no reason for it. I’m just saying they have poor messaging. Nothing you have said really refutes that point. It gives a reason, a fairly reasonable one to be honest, but it still needs to be said they have taken a few missteps and it’s not the fault of the users for misunderstanding the relation of all of the moving pieces.
