Traffic from AI scraping bots are ruining Wikipedia's web hosting infrastructure, so I understand why they'd make deals to streamline it and keep operating costs down.
But there's a lot of stuff in the article that are massive red flags, and I'm not optimistic about it.
But there's a lot of stuff in the article that are massive red flags, and I'm not optimistic about it.
January 16, 2026 at 8:14 AM
Traffic from AI scraping bots are ruining Wikipedia's web hosting infrastructure, so I understand why they'd make deals to streamline it and keep operating costs down.
But there's a lot of stuff in the article that are massive red flags, and I'm not optimistic about it.
But there's a lot of stuff in the article that are massive red flags, and I'm not optimistic about it.
Tl;dr, ChatGPT is an unreliable source of information, and the only guaranteed way to be sure it doesn't hallucinate and lie to you is to already know more than it does.
January 15, 2026 at 8:07 AM
Tl;dr, ChatGPT is an unreliable source of information, and the only guaranteed way to be sure it doesn't hallucinate and lie to you is to already know more than it does.
...any source of information is great, it's way too easy to accept new information if you're not already well versed in that particular subject, and it's simply not reasonable to put that responsibility on regular users.
January 15, 2026 at 8:07 AM
...any source of information is great, it's way too easy to accept new information if you're not already well versed in that particular subject, and it's simply not reasonable to put that responsibility on regular users.
...you end up with people being wildly misinformed, because they didn't know to express doubt that the LLM could pick up and react on (and even then, it can still double down instead of admitting it's at fault, or it could fold even when it's giving correct information). While being critical of...
January 15, 2026 at 8:07 AM
...you end up with people being wildly misinformed, because they didn't know to express doubt that the LLM could pick up and react on (and even then, it can still double down instead of admitting it's at fault, or it could fold even when it's giving correct information). While being critical of...
...it to be used for that particular topic;
or it's because the user has pushed back against the LLM (meaning it's 100% on the user to actually recognize that the LLM is wrong) because it's trained to react to the user's response, not its own inability.
The latter is the big danger. That's how...
or it's because the user has pushed back against the LLM (meaning it's 100% on the user to actually recognize that the LLM is wrong) because it's trained to react to the user's response, not its own inability.
The latter is the big danger. That's how...
January 15, 2026 at 8:07 AM
...it to be used for that particular topic;
or it's because the user has pushed back against the LLM (meaning it's 100% on the user to actually recognize that the LLM is wrong) because it's trained to react to the user's response, not its own inability.
The latter is the big danger. That's how...
or it's because the user has pushed back against the LLM (meaning it's 100% on the user to actually recognize that the LLM is wrong) because it's trained to react to the user's response, not its own inability.
The latter is the big danger. That's how...
...haven't been trained in.
If you ever see ChatGPT or another LLM say they don't know or that they were wrong, it's because of one of two reasons:
Either it's a topic it's been trained to always respond to in that way (making it completely useless at that topic) because the creators don't want...
If you ever see ChatGPT or another LLM say they don't know or that they were wrong, it's because of one of two reasons:
Either it's a topic it's been trained to always respond to in that way (making it completely useless at that topic) because the creators don't want...
January 15, 2026 at 8:07 AM
...haven't been trained in.
If you ever see ChatGPT or another LLM say they don't know or that they were wrong, it's because of one of two reasons:
Either it's a topic it's been trained to always respond to in that way (making it completely useless at that topic) because the creators don't want...
If you ever see ChatGPT or another LLM say they don't know or that they were wrong, it's because of one of two reasons:
Either it's a topic it's been trained to always respond to in that way (making it completely useless at that topic) because the creators don't want...
Why yes, I definitely do research by asking the chatbot that is INHERENTLY INCAPABLE of admitting it doesn't actually know or that it's wrong!
This is a fundamental problem with LLMs. You literally cannot train them to say "I don't know" because that's only a valid response to contexts they...
This is a fundamental problem with LLMs. You literally cannot train them to say "I don't know" because that's only a valid response to contexts they...
January 15, 2026 at 8:07 AM
Why yes, I definitely do research by asking the chatbot that is INHERENTLY INCAPABLE of admitting it doesn't actually know or that it's wrong!
This is a fundamental problem with LLMs. You literally cannot train them to say "I don't know" because that's only a valid response to contexts they...
This is a fundamental problem with LLMs. You literally cannot train them to say "I don't know" because that's only a valid response to contexts they...
Have you tried wrapping yourself with blankets and drinking a hot cup of cocoa or tea while looking at the falling snow or your favourite show in case it doesn't snow in your geographical area?
It doesn't make having to go out in the cold any better, but it's at least something.
It doesn't make having to go out in the cold any better, but it's at least something.
January 15, 2026 at 7:28 AM
Have you tried wrapping yourself with blankets and drinking a hot cup of cocoa or tea while looking at the falling snow or your favourite show in case it doesn't snow in your geographical area?
It doesn't make having to go out in the cold any better, but it's at least something.
It doesn't make having to go out in the cold any better, but it's at least something.
It links to the latest, while the comic posted here is old, not new.
January 15, 2026 at 7:16 AM
It links to the latest, while the comic posted here is old, not new.
But it's specifically that link. I can read your comic completely fine if I just browse through tinyview regularly.
January 14, 2026 at 9:51 PM
But it's specifically that link. I can read your comic completely fine if I just browse through tinyview regularly.
"Application error: a client-side exception has occurred while loading tinyview.com (see the browser console for more information)."
January 14, 2026 at 9:47 PM
"Application error: a client-side exception has occurred while loading tinyview.com (see the browser console for more information)."
I'm guessing they scrolled down your page and saw the Black Ant vore post. They seem to view vore and furry stuff as cannibalism and zoophilia, and equate it all with CSAM.
January 14, 2026 at 9:42 PM
I'm guessing they scrolled down your page and saw the Black Ant vore post. They seem to view vore and furry stuff as cannibalism and zoophilia, and equate it all with CSAM.
I found the list and the person who created it. Apparently this and your work account was blocked ~41 hours ago, but not your third account. I guess they didn't know about it.
Judging by their posts, I think it's a blessing they default to blocking instead of harrassing.
Judging by their posts, I think it's a blessing they default to blocking instead of harrassing.
January 14, 2026 at 8:10 PM
I found the list and the person who created it. Apparently this and your work account was blocked ~41 hours ago, but not your third account. I guess they didn't know about it.
Judging by their posts, I think it's a blessing they default to blocking instead of harrassing.
Judging by their posts, I think it's a blessing they default to blocking instead of harrassing.
I've seen a woman getting murdered with total impunity against my will just from browsing social media
I already knew about it, I was already mad about it
I didn't need to see it to think the nazis in ICE (and the nazi politicians backing them) should all be jailed for life
Now I just feel like shit
I already knew about it, I was already mad about it
I didn't need to see it to think the nazis in ICE (and the nazi politicians backing them) should all be jailed for life
Now I just feel like shit
January 10, 2026 at 11:11 AM
I've seen a woman getting murdered with total impunity against my will just from browsing social media
I already knew about it, I was already mad about it
I didn't need to see it to think the nazis in ICE (and the nazi politicians backing them) should all be jailed for life
Now I just feel like shit
I already knew about it, I was already mad about it
I didn't need to see it to think the nazis in ICE (and the nazi politicians backing them) should all be jailed for life
Now I just feel like shit
This reads like a linkedin-style motivational/educational post: "I pissed my bed on my honeymoon and this is what I learned about project management."
January 10, 2026 at 11:02 AM
This reads like a linkedin-style motivational/educational post: "I pissed my bed on my honeymoon and this is what I learned about project management."
Never fucking mind how reproductive rights is part of the human rights.
January 6, 2026 at 11:09 AM
Never fucking mind how reproductive rights is part of the human rights.
SO said "I love the smears on her nipples!" I don't have the same eye for animation as them, but I do love just about anything you do, especially the animations!
January 6, 2026 at 10:45 AM
SO said "I love the smears on her nipples!" I don't have the same eye for animation as them, but I do love just about anything you do, especially the animations!
Reposted by A__
Reposted by A__
Wish I could see the original post. Oh, well.
January 5, 2026 at 7:12 AM
Wish I could see the original post. Oh, well.
He has complete MORE Project 2025 goals than any president in recent memory have completed election campaign promises, both in turns of numerical amount and percentage of the total.
"BOOGIEMAN"???
"BOOGIEMAN"???
January 2, 2026 at 9:34 AM
He has complete MORE Project 2025 goals than any president in recent memory have completed election campaign promises, both in turns of numerical amount and percentage of the total.
"BOOGIEMAN"???
"BOOGIEMAN"???
It's not. It *might've* been etymologically based on Deutsch, but the term is unequivocally Dutch Angle.
December 29, 2025 at 3:15 AM
It's not. It *might've* been etymologically based on Deutsch, but the term is unequivocally Dutch Angle.
It's part of most if not all the 40k RPGs. In the FFG ones I'm pretty sure it halved the power cell's capacity. In Imperium Maledictum you need to pass a +20 Tech test to see if it takes 4 or 8 hours to charge. Fumble destroys it.
December 23, 2025 at 7:16 AM
It's part of most if not all the 40k RPGs. In the FFG ones I'm pretty sure it halved the power cell's capacity. In Imperium Maledictum you need to pass a +20 Tech test to see if it takes 4 or 8 hours to charge. Fumble destroys it.