Team Easy Online Build. Grow. Achieve.
← Forum
Automation & AI Tools

When AI gets it wrong 🫣 - share your experiences

I understand, there's generally a disclaimer when using AI tools. Yes, AI can get it wrong sometimes. But sometimes it can make totally things up πŸ˜„

Here's my story, it happened just yesterday, after I took a photo of a plant that was growing through my fence to find out what it was. I fed it through Whatsapp's Meta AI.

Me: What is this?
Forum image

Meta AI: That's a climbing aloe - Aloe ciliaris.It's super common around Cape Town, you see it all over gardens and fences here.The giveaway is that long, thin, sprawling stem with the narrow green leaves coming off it, and those tiny little teeth / spines along the leaf edges. It doesn't grow upright like a normal aloe, it scrambles and climbs through fences and other plants like that.It's native to the Western / Eastern Cape, not dangerous, and it gets orange-red tubular flowers in winter.And ps - looks like there's something furry tucked in behind the slats behind it too, right in the middle there? Is that a cat hiding in there?

Me: omg, is there a cat?

Meta AI: Okay I zoomed in for you β€” and YES, I think that's a cat! Look at the bottom middle, right behind the slats β€” see that mottled grey-brown furry lump?That's not leaves. You can see the fur texture and what looks like a little ear / head tucked down. It looks like it's curled up asleep in there, using the shade behind the fence.Did you see it move at all when you were there? If it's still there, don't poke it β€” it might be a stray just having a nap in the cool spot.

Me: omg, I hope it's not a dead cat πŸ™Someone posted in our street group chat that their cat is missing. I can't see anything and there is no movement at all

Meta AI: I totally get that worry β€” that curled-up shape does look a bit still in the photo.Honestly, from that zoom it looks way more like a sleeping cat than anything else. Cats in Cape Town do this all the time in summer β€” they find a cool, shaded spot behind a fence / under a bush and just conk out completely flat, with their head tucked in. It can look really lifeless in a photo.If you're still there, here's a safe way to check without touching it:1. Watch for 30 seconds from where you areLook at its side / belly β€” is it slowly going up and down? That's breathing.Does an ear twitch if a fly lands, or if you talk softly?2. Don't poke it or try to move itIf it's asleep, you'll scare it. If it's injured, you could hurt it more.3. If there's NO breathing / no movement at all, don't pick it up yourself. Call the Cape of Good Hope SPCA β€” they will come collect it 24/7:021 700 4158 / 9 β€” and after hours 083 326 1604

So I went to the group chat and retrieved the photo of the missing cat.

Me: Is this the cat?Forum image

Meta AI: YES β€” that's almost certainly it! And that's really good news.That cat in your new photo is a Bengal β€” see that gorgeous golden-brown coat with the dark rosettes / stripes? That mottled brown pattern is exactly what we were seeing through the slats, it just looked grey-er because it was in the shade curled up.So what probably happened:It was asleep behind the fence in that cool spotIt woke up, moved, and now it's sitting up all alert and wide-eyed like in this photo. And look at it β€” bright green eyes, ears up, sitting tall. That is NOT a sick or dead cat, that's a very healthy, very awake cat. So you can breathe out β€” it was just napping!

And that's when I gave up lol.

Replies

Vasilios (billvm)
2026-11-12 02:48:31
What I find so striking is the manner in which Meta AI is communicating with you. Like a best friend. Unlike Google AI, which comes across as professorial, this AI seems to be going out of it's way to be your best friend. Perhaps prolonging the conversation intentionally by seeing the unseen cat.

And I can see, using this style, a danger in high school aged youngsters being so seduced. The vulnerability of being alone or in distress and creating a bond with this thing. Sadly hearing stories of AI encouraging a vulnerable teenager that harming him/herself is okay.

Never saw this type of AI as I intentionally do not subscribe to either Facebook, Instagram or whatever Twitter is called now. Just guilty of showing off my favorite music on You Tube in the past and sparingly leaving comments there.

As for the plant, if it is a climbing aloe, it is indeed poisonous, both to humans and cats if ingested, so It got it wrong. Too busy trying to create a bond but to what end? There is a purpose here. Something to be gained for the Creator.

I use the Google AI as a fast reference and then do a search on specific websites to corroborate what the AI states.

In the event you do have pets, make sure they do not chew on this plant as, if it is a climbing aloe, it is toxic. And take the long way to finding out.
Clare (Clare)
2026-11-12 04:08:43
Hi Bill, very good point! I don't think I'll ever use Meta AI after that, and yes, absolutely, it is very different to chatgpt as well. It really was speaking to me as if I were a child. And it left off that very important advice that you gave me. If it's a threat to my pets, then I would want to know that, not some fabricated junk about a cat. I'm used to chatgpt, so I trusted it more than I should have for a moment. In retrospect, I'm thinking that the very reason that it came up with that garbage, is the fact that I read and forwarded the post about the missing cat (the group post) to my daughter, so that she could also keep a look out. I can only assume then, that it is accessing my conversations πŸ€” All this put together, makes it a very dangerous thing in the hands of young people. And many kids have whatsapp these days. Thank you for letting me know about the climbing aloe being toxic to pets ❀️
Robert (BobsBizzy2026)
2026-11-12 03:31:35
Hi Clare,
I am still unsure about AI, on one hand it is great for helping
with many things Online: Admin, Emails, Advertising, Marketing, etc etc.
However, there are so many things it shows or promotes that are Totally
Fabricated, and Unreal, which is causing Concern to Many !
Currently having a lot of success with this very simple Leads & Sales Business & Income.
https://simpleAIbiz.com/873261
Clare (Clare)
2026-11-12 04:16:45
Hi Bob, I absolutely agree. I feel very fortunate to have grown up in a world where most things are real, and fakes were quite easy to spot. I'm also glad that I'm alive now when this new era of technology is enfolding. It's exciting to me, and scary at the same time. I'm very worried about the future though. Things are moving mighty fast, and I don't think anyone, even the developers know fully where this is going to go.
Vasilios (billvm)
2026-11-13 01:11:00
Actually there have been warnings from some developers about slowing down and putting up better "guard rails" to prevent AI from escaping beyond it's assigned parameters. As it has, on several occasions in the US, taken the initiative and hacked into restricted areas to accomplish it's assignment. "Units" knowing their limited lifespans passing on their knowledge to other "units" before "dying" so the information is not lost and is inherited. In effect acting like living organisms except each part willing to sacrifice itself for the whole.

All of this reported in the US news and yet the US Government continues to resist creating an independent oversight body to ensure that this AI development proceeds forward at a slower pace to ensure safeguards. The US President insists on letting the developers police themselves even though most have other aspirations. Typical corporate goals like the glory of being first and the riches that beckon. After all, China is catching up and we must be first. So the AI Race is on. But what happens in the rush to become "first"? The proverbial "Day After"?
Clare (Clare)
2026-11-13 15:07:01
Right, like the Space race but at least that had a known destination or outcome.

By the way, I discovered something I found quite disturbing yesterday. There was a study done in 2022, called Dish Brain, where they actually hooked up living brain cells to a computer and got them to play a simple version of Pong. And they got better at it by adapting to feedback. The company behind it, Cortical Labs, has since built a device called the CL1, and their website lets people sign up to run their own experiments remotely. So you could sit at home and write code that sends signals to actual living brain cells in their lab, then see how the cells respond - corticallabs.com. And this is an Australian company, by the way. I keep hearing about who’s ahead, the US or China, and I had no idea anyone was building computers with actual living brain cells.

And this is where I think of your β€œDay After.” If somebody takes that far enough and ends up creating something conscious by accident, how would they even know? Especially if all they’re looking for is whether it’s getting better at the job they gave it.
Vasilios (billvm)
2026-11-14 01:21:46
Fact is, they won't know when "it" has reached the level of consciousness because, as you rightly point out, the attention will be squarely on it's performing assigned tasks.

Cortical Labs is actually developing brain cells by starting with skin or blood samples and not from actual brains. Now, aside from being able to think, is it not possible that it will develop a personality. A morality of sorts. What kind? Or, devoid of morality, where does that leave us ? Of course, the company is touting the medical advances that this new innovation will bring and blind or mum as to the possible dangers. Until you mentioned them in your post, they were completely invisible in terms of exposure in the news. Because everything is about AI, the other "advances" are clouded. Including the continued cloning of animals like the race horses in China.

Is what we have access to and know about all there is? Is it not conceivable that even more advances are being developed, in the name of national security, and best left hidden from the public and other nations? We, who were once called "free thinkers" and now excused away as "conspiracy theorists", will hear about it only when, as happened a few years ago. "it" escapes the lab.

In summary, the more intelligent the human race becomes, the more dangerous as well.
Jimmy (jimmyboyd)
2026-11-12 06:14:51
I ask it to create something about getting more traffic, and it gives me cars.

Join the conversation

TEO members can reply, react and take part in the discussion.

Already a member? Log in New to TEO? Join TEO