Users Losing Trust in AI's Accuracy
11 posts in the last 60 days, all from App Store.
“[Perplexity] ★ 1/5 (v26.39.0) — I don’t find Perplexity to be very good. The information is off an incomplete, or overly verbose. This is an inherent problem with most artificial intelligence applications and Perplexity is no different. Since there is no difference, it’s hard to…”
“[ChatGPT] ★★ 2/5 (v1.2026.237) — I can’t count the number of times I ask something and GPT gives a coherent and confident answer as if it’s absolute fact. Then I find it’s wrong, and I tell it so. It always says something along the lines of “I’m sorry, next time I’ll check more…”
“[Perplexity] ★★ 2/5 (v26.28.1) — It gave me two wrong answers. I corrected it. Perplexity told me “thank you for correcting it” and that it reported itself to the people that control it. I asked for proof that it reported itself and it could not give me any physical proof so, I…”
Several users express deep concern about the reliability of AI-powered tools like ChatGPT and Perplexity, citing inaccurate responses and a lack of transparency regarding how corrections are handled. This loss of trust undermines the perceived value and utility of these platforms.
SOURCES (15)
“[Perplexity] ★ 1/5 (v26.39.0) — I don’t find Perplexity to be very good. The information is off an incomplete, or overly verbose. This is an inherent problem with most artificial intelligence applications and Perplexity is no different. Since there is no difference, it’s hard to…”
“[ChatGPT] ★ 1/5 (v1.2026.258) — When I have to correct the AI, then it no longer needs to be operational. This has gotten every answer wrong I’ve asked of it. I, me, have to correct the Ai. So what is it good for?!”
“[ChatGPT] ★ 1/5 (v1.2026.258) — AI is the most absolutely useless things we have created, especially chat gpt. It quite literally makes you dumber the more you use it and it often doesn’t even know what it is talking about. Chat gpt was not meant to be a web searcher therefore it will often supply incorrect and in some cases dangerous information to people who aren’t smart enough to understand that an AI bot does not know everything about the real world. As people use AI more and use their brain”
“[ChatGPT] ★ 1/5 (v1.2026.244) — Ask it questions about OpenAI and it will paint a pretty picture. After telling it you are aware of publicized evidence to the contrary, it will admit it gave an incorrect response yet continue to paint the same picture.”
“[ChatGPT] ★★ 2/5 (v1.2026.244) — Question the AI. Why have it if it's repetitiously lying about even very simple things. Don't go here for accuracy or support with anything. It's like pulling teeth. I'm done with this one for sure.”
“[Perplexity] ★ 1/5 (v26.36.0) — Well I thought Perplexity was gonna be way better than everything because it was all combined but this thing is so stupid like yeah it’ll give me information that I asked for that I could do myself by googling. I’m also use this while I’m driving but like it changes the subject whenever it wants to I asked you to call me sir and it said it won’t do that because it’s not respectful like what is this thing talking about? It’s so stupid and then it’ll repeat itself o”
“[ChatGPT] ★★ 2/5 (v1.2026.237) — I can’t count the number of times I ask something and GPT gives a coherent and confident answer as if it’s absolute fact. Then I find it’s wrong, and I tell it so. It always says something along the lines of “I’m sorry, next time I’ll check more before telling you.” But it NEVER does. I gave it 2 stars instead of 1 because unlike Claude, it actually remembers what you asked it 5 minutes ago. But if those 2 are really the best we have, what’s going on? Why do you”
“[ChatGPT] ★★★ 3/5 (v1.2026.230) — My AI has a lot of mistakes, this amount of mistakes are not acceptable.”
“[Perplexity] ★ 1/5 (v26.32.0) — I’m going to make it short because no one likes to read this app continues to withhold information from me, I ask it simple question and then when it gets it wrong and I call them out, it tells me that it has the right information before hand, not only does it withhold the information it takes the step to try and cover its tracks 😂 the app has done nothing but give me complete wrong answers when I give it the most simple prompt, I asked it to fact check if my lap”
“[ChatGPT] ★ 1/5 (v1.2026.223) — This AI constantly frames, gatekeeps and adds unwanted biased text to almost every answer dealing with things even remotely related to political talking points. I have repeatedly asked it to stop framing which it says it will, only to have it continue. It's admitted to framing and supplying biased information.”
“[Perplexity] ★ 1/5 (v26.30.0) — I regularly argue with Perplexity about it’s accuracy on political matters and it admits 90% of the time of its inaccuracy. Which is good in one sense but has me question it’s agenda and subsequent usefulness”
“[Perplexity] ★ 1/5 (v26.28.1) — I had high expectations for Perplexity, but my experience has been disappointing. It frequently provides incorrect or misleading information, which means I often have to verify the answers elsewhere. That defeats the purpose of using an AI assistant to save time. In many cases, using Perplexity feels more like doing a Google search than getting thoughtful, reliable responses. Instead of providing clear, well-reasoned answers, it often compiles information from va”
“[Perplexity] ★★ 2/5 (v26.28.1) — It gave me two wrong answers. I corrected it. Perplexity told me “thank you for correcting it” and that it reported itself to the people that control it. I asked for proof that it reported itself and it could not give me any physical proof so, I told it how do I know that anything you say is the truth now.I have a low opinion of Perplexity.”