Error Message - No Access to ChatGPT

[m1sp0rquer14s]

When I try to connect to ChatGPT, I am getting this message, " Free Research Preview. ChatGPT may produce inaccurate information about people, places, or facts. [ChatGPT August 3 Version]", and I can’t type anything, or I get a blank window.
I

[elmstedt]

Search the forum.

This has been asked and answered many times.

[m1sp0rquer14s]

I did… found many posts about errors but not this specific one, otherwise I would have not posted.

[elmstedt]

Did you though?

First result:

One of many solutions listed in the thread:

Link: I am having this message and cant go further"Free Research Preview. ChatGPT may produce inaccurate information about people, places, or facts. ChatGPT August 3 Version" - #48 by RonaldGRuckus

[m1sp0rquer14s]

Well, thank you. We are not all that good with this app. Maybe bring the rudeness down a notch.

[elmstedt]

@m1sp0rquer14s

We are not all that good with this app.

The forum?

It’s just a very simple search for exactly what you described seeing.

You created your topic 110 seconds after creating your account. So, generously you could have spent—at most—90-seconds looking for the answer.

I’m not saying you didn’t search at all, I’m just saying you didn’t search very hard.

Even a simple Google search brings up the relevant topics in the forum,

So, you can suggest,

@m1sp0rquer14s

Maybe bring the rudeness down a notch.

and you wouldn’t be wrong, I could be nicer—but it’s also “rude” to not adequately try to find the answer yourself here first.

There are ostensibly some 342,000 members of this forum. Of those, some 60,500 signed up in the last 30 days, but only 68,000 members have been active in the last 30. So, of the 282,000 members who signed up more than a month ago, only around 7,000 of them were active in the last month, about 2.5%.

The overwhelming majority of these users pop in, ask a question which has already been answered and disappear never to be seen again—until there is another widespread, temporary issue, when they come back and ask another question which has already been answered.

Now, that’s not all on you. You just came in and asked one question which was already asked and answered twice already. In and of itself isn’t that big of a deal. But, to those of us who are here every day, it’s a tremendous annoyance to have to answer the same questions over, and over, and over.

Now, I could have simply ignored your topic altogether. But, instead I told you to search because it had already been asked and answered.

Even if you really had already searched and found nothing, that should have been a clue for you to search again—knowing the answer was there.

But, instead of searching again you… I don’t know? Protested the suggesting because you did already?

So, I don’t know friend, I think suggesting you search and questioning if you really did is pretty far from rude—especially given the context of the situation.

[m1sp0rquer14s]

Lol… wow! Like you said, i created the account right before asking the question (not really since i have been using this for at least 3 months, but whatever shows there must be right!). Can you believe i posted the same thing you did on the search section and nothing came out? The last thing i wanted to do was to “bother” the busy people that frequent the forums… i will see myself out of this! thank you for the lecture (and help!)

[PLANECRASH]

Maybe bring the rudeness down a notch.

I agree! this is now the #1 search result on duck duck go, for this error message, and now we’ve all got to scroll through @elmstedt’s frustrated diatribe before getting to the answer. Dude’s spent way too much time on this forum.

[RonaldGRuckus]

@elmstedt

There are ostensibly some 342,000 members of this forum. Of those, some 60,500 signed up in the last 30 days, but only 68,000 members have been active in the last 30. So, of the 282,000 members who signed up more than a month ago, only around 7,000 of them were active in the last month, about 2.5%.

The overwhelming majority of these users pop in, ask a question which has already been answered and disappear never to be seen again—until there is another widespread, temporary issue, when they come back and ask another question which has already been answered.

This point should make it clear why. This forum is honestly a “support overflow” forum. Whenever there is a noticeable issue there is almost never any official response and people flood the forums starting new threads, new posts without bothering to see that there are multiple potential solutions available.

We try to help but get overwhelmed and buried by seagulls.

You know, it’s frustrating and so common to see people do rude things (usually from ignorance) and then get flabbergasted when the response isn’t as kind as they expect. Then they walk away thinking “man, these people are so rude”

Also

@PLANECRASH

now we’ve all got to scroll through @elmstedt’s frustrated diatribe before getting to the answer

The answer is post #4, less than 2 paragraphs away. It’s ironic, really. He’s complaining about all of this noise that active members need to scroll past to find meaningful discussions and you are complaining about the same thing

[elmstedt]

@rickscorpio64

2nd This is not an ERROR.
it’s a dumbass message. It took me a few hours to decide to look this up. In it’s current state

Errors occur in multiple forms. Two of the most common are,

1. A process fails entirely
2. An unexpected result is produced

This is an example of the latter.

@rickscorpio64

First time it came up for me. And the tab stays up always on my mobile.

So this is server side not client side issue I do not believe people did anything anything on their mobile device different, it did this in less than an hour on mine.

Ahhh… I see. It’s clear you have a fundamental misunderstanding about what happens after you type a URL into your browser’s address bar.

This is 100% a client-side issue. Here’s how it works,

1. Entering a URL: Imagine you want to visit your friend’s house. You need to know their address, right? In the same way, when you want to go to a website, you type its “address” into the computer.

2. DNS Lookup: Now, let’s say you ask your mom how to get to your friend’s house. She looks it up on a map and tells you the way. The computer does something similar; it looks up the website’s real address.

3. Retrieve HTML: Once you know the way, you go to your friend’s house and say hi. The computer goes to the website’s address and picks up a list of things it needs to show you the website.

4. Collect Assets: Imagine your friend has toys, snacks, and a TV in their house. You gather all these things to play. The computer does the same; it collects pictures, videos, and other stuff from the website.

5. Additional DNS for Assets: Sometimes, you might need to go to the neighbor’s house to borrow some toys. The computer might also need to look up more addresses to get extra stuff for the website.

6. Render the Page: Finally, you and your friend set up all the toys and snacks and start playing. The computer puts all the things it collected together and shows you the website.

I hope that makes it easier to understand!

Now, what is happening in this case is there is a specific URL your computer is failing to resolve, specifically cdn.oaistatic.com.

There are three primary reasons why this would happen,

1. The URL is bad and does not correspond to a valid, registered domain:

   • Imagine you want to visit a friend named “Tommy”, but instead of writing “Tommy’s House” on your map, you accidentally write “Tummy’s House”. There’s no friend named “Tummy”, so you won’t find the house.
   • Similarly, if you type a wrong or non-existent website address (URL) into your computer, it won’t know where to go because that website doesn’t exist.

2. The DNS server used by the user is out of date and does not have the current IP address for the domain:

   • Let’s say you have an old map that shows your friend “Lucy” living on “Pine Street”. But Lucy moved to “Oak Street” last year, and your map hasn’t been updated. So, you’ll end up at the wrong house.
   • NS servers are like maps for the computer. If they are outdated, they might direct the computer to the wrong place because they don’t have the current address of the website.

3. The domain is blocked for the user either at the client level (the user’s browser has an extension which is blocking the domain—an ad-blocker or privacy extension) or at the network layer (the user has a firewall or custom DNS server which nullifies the request to the domain either explicitly or by a heuristic):

   • Imagine you’re about to visit your friend “Jake”, but your big brother stops you and says, “You can’t go to Jake’s house today.” Your big brother is trying to protect you, maybe because he heard Jake is sick and doesn’t want you to catch a cold.
   • Some browser extensions (like ad-blockers or privacy tools) and network devices (like firewalls) act like protective big brothers. They might block certain websites or parts of websites to keep your computer safe or to avoid showing you ads.

Now, we can rule out issue number one, because the domain is a valid registered domain.

That leaves the other two problems—both client-side from the perspective of OpenAI, because OpenAI cannot fix them, only you can!

Let’s tackle the first,

An outdated DNS server.

Now, there are almost as many computer setups for connecting to the Internet as there are computers on the Internet, so this requires a bit of technical know-how.

Most users use the DNS server provided by their Internet service provider because it’s simple, easy, and requires zero effort or knowledge on their part. This is usually perfectly fine, DNS records propagate throughout the Internet usually within 48-hours and almost always within 72-hours. Some, rare, providers are slower than others though and this seems to be the root cause of the issue.

The solution is to set a specific DNS server for your computer to use. There are three verified, free, and common DNS servers which people are using,

• Google
  • 8.8.8.8
  • 8.8.4.4
• CloudFlare
  • 1.1.1.1

Setting Custom DNS Server on Various Platforms

Here are the steps to set your DNS server on different platforms:

Windows:

1. Open the Control Panel from the Start menu.
2. Click Network and Sharing Center.
3. Click Change adapter settings.
4. Right-click on your active network adapter and choose Properties.
5. Choose Internet Protocol Version 4 (TCP/IPv4) and click Properties.
6. Enter the new DNS servers into the boxes under "Use the following DNS server addresses"².

OSX:

1. Go to System Settings. You can find it by pressing Command + Space on your keyboard and typing System Settings.
2. Go to Network.
3. Select a network service.
4. Select Details.
5. Go to DNS.
6. Under DNS Servers, select Add⁵.

Linux:
The process of changing DNS server settings on Linux depends on the distribution you are using. However, here is a general approach:

1. Open Terminal.
2. Edit the network configuration file using a text editor such as nano or vi.
3. Locate the line that starts with nameserver.
4. Replace the existing IP address with the IP address of your desired DNS server.
5. Save the file and exit.

iOS:

1. Open Settings on your iOS device.
2. Tap Wi-Fi.
3. Tap the information icon (i) next to your Wi-Fi network name.
4. Scroll down and tap Configure DNS.
5. Tap Manual.
6. Tap Add Server¹.

Android:

1. Open Settings on your Android device.
2. Tap Wi-Fi & Internet or similar options depending on your device model and Android version.
3. Tap the Wi-Fi network you are connected to.
4. Tap Advanced or similar options depending on your device model and Android version.
5. Scroll down and tap IP settings or similar options depending on your device model and Android version.
6. Change the setting from DHCP to Static, then scroll down and enter the IP address of your desired DNS server¹.

Please note that these instructions are general guidelines, and there might be slight variations depending on your specific device or operating system version.

Now, let’s deal with the second,

Domain is blocked by the client

If you are running an ad-blocker or privacy extension there are three ways to resolve this,

1. Disable the extension(s) which is(are) blocking the domain cdn.oaistatic.com
2. Disable filtering on the domain chat.openai.com
3. Whitelist cdn.oaistatic.com

If you have a network appliance like a firewall you’ll need to whitelist the domain there.

There are too many ad-blockers and firewalls to provide comprehensive instructions on how to complete these tasks for all of them, but this is a good article to start you on that path,

How to whitelist in ad blockers.

@rickscorpio64

So maybe be nice and check dev side and see what changed and alert people. Obviously it is an issue as you pointed out that it has been answered before. An answer is not the same as a Fix

1. I don’t work for OpenAI, I’m a Statistics PhD student who does some development work on the side using OpenAI’s platform.
2. What changed is OpenAI registered the domain oaistatic.com at 2023-09-18 23:34:43 UTC and some time after the DNS records for that domain propagated throughout the Internet, they updated the web page for chat.openai.com to access assets at this new domain.
3. Some users have DNS servers which update especially slowly and as a result their computers are unable to locate these assets in their new location.
4. Without these assets loaded, the website becomes non-functional.

@rickscorpio64

Still showing up? they need to fix it.

There is nothing they can really do to “fix” it. The issue is on your end of the line, your set-up is broken, you need to fix your problem.

@rickscorpio64

Don’t be rude to people. We already have that crap world wide on all issues. So ya, don’t tone it down. Turn it off!

At no point was I “rude” to anyone.

I, in order,

1. Told the OP to search the forum because the answer is here
2. Responded to the user’s protests that they did search by,
   1. Pointing out they couldn’t have put much effort into the search
   2. Performing the search myself and demonstrating that,
      1. This shouldn’t have been a new topic as there were already two identical topics
      2. The solution to the issue was already described
   3. Linking directly to a post describing the solution
3. Responded to the OP’s claim of rudeness by,
   1. Pointing out that it was their behavior which is actually rude
   2. Illustrated why their lazy behavior is so vexing

Since this topic was posted there have been no less than three new topics created complaining of the same issue.

We should not need to answer the same questions over and over and over again for people who think they are the main character and can’t be troubled to search the forum.

Why do you think the forum is persistent? Why aren’t topics immediately deleted once a solution is provided?

Could it possibly be so that others can benefit from the knowledge presented so they can immediately get the help they need?

I responded about three-hours after the OP posted their topic.

If they had, at any point during that time, bothered to search the forum, they would have found their solution nearly immediately.

A bit after that they claimed they had already looked, but didn’t find anything.

I wasn’t able to respond until about another three-hours after that.

All told they could have had their solution about seven-hours sooner if they had just bothered to look!

So, I apologize if I’m occasionally short or terse when others fail to put forth the absolute bare minimum of effort to find the solution that already exists, right here on this very forum—even after they’re told directly that the solution is here.

I hope the post is sufficiently clear and detailed enough that anyone reading will be able to fix the client-side issues which are causing them to experience an inability to use the ChatGPT website.

If not, I take solace in knowing that (at least for those with a passive problem, that is an outdated DNS server) the underlying issue will resolve itself eventually—likely sooner rather than later as DNS servers are (more or less) being constantly updated.

[PLANECRASH]

Stackoverflow unemotionally marks questions as duplicates with a hyperlink to the answer using much less verbosity.

Edit: For example:

If I were allowed to post links, you’d notice none of the participants in that thread had a mental breakdown.

[stephenschultz0901]

Same people on here as the rude Stackoverflow experts who claim to be here to teach and help beginner individuals but whine that this process is similar to customer service. What did you think this type of volunteering was? They’ll argue that they were helping teach users a lesson and base it behind their God complex.

[blinkingcloud]

But you did not need to answer this post if you did not want to, did you? Are you sure you are not the one that feels like the main character? Ughh, this reminds me of one of my lecturers who gave me an attitude when I asked him to explain the difference between two similar functions and he said “just Google it”. I know how to Google but that’s not what I come to the labs for, I can Google at home. I know it’s not exactly the same situation, but it’s the antisocial attitude that’s completely irritating. Basically admitting you know the answer to the question but refuse to give it away out of pure pettiness.