Bad uses / over-promised of AI

Forums › Artificial Intelligence (AI, LLM, ML) › Bad uses / over-promised of AI

1 | ... | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14

19021 posts

Uber Geek
+1 received by user: 8111

Lifetime subscriber

#3478049 4-Apr-2026 17:51

Google confirms it is at the early stages of commissioning gas fired plants for AI and refuses to comment on compatibility with carbon neutrality goals:

https://www.theguardian.com/technology/2026/apr/02/google-ai-datacenter

It's not a good look for the logo.

Behodar

11210 posts

Uber Geek
+1 received by user: 6237

Trusted

Lifetime subscriber

#3479525 9-Apr-2026 14:38

Me: I want to automatically renew and update this HTTPS certificate. I wonder how to do that.

Google's blithering "AI": You can do that with PowerShell! Blah blah blah, set up a trigger etc. etc. and then use the following script to assign the new certificate to the server:

Restart-Service

gzt

19021 posts

Uber Geek
+1 received by user: 8111

Lifetime subscriber

#3479549 9-Apr-2026 16:15

I've noticed the Google Search 'AI' tab gives better answers than the summary on the 'All' tab.

You may or may not need a Google account to see the 'AI' tab.

I do know if you're on an apps account then gemini or something does need to be enabled in apps admin before you see that.

Behodar

11210 posts

Uber Geek
+1 received by user: 6237

Trusted

Lifetime subscriber

#3481526 17-Apr-2026 10:13

Microsoft blog post referring to Copilot as a "source of truth" 🤦‍♂️

cddt

2053 posts

Uber Geek
+1 received by user: 1983

#3481529 17-Apr-2026 10:26

Behodar:

Microsoft blog post referring to Copilot as a "source of truth" 🤦‍♂️

I thought it was "entertainment only"...

Behodar

11210 posts

Uber Geek
+1 received by user: 6237

Trusted

Lifetime subscriber

#3481530 17-Apr-2026 10:40

I saw an article with an AI-generated image of a laptop. How did I know it was AI? Well, my first clue was that the Tab key was labelled "comank"...

Geekzone support

Support Geekzone with one-off or recurring donations Donate via PressPatron.

gzt

19021 posts

Uber Geek
+1 received by user: 8111

Lifetime subscriber

#3481966 18-Apr-2026 17:10

I had planned to purchase a Nest Hub for a mostly bed-bound elderly friend for easy access to music. At the point of purchase I realised Google are upgrading Nest Hub with Gemini soon. I'm guessing there is no way to turn Gemini off or make it behave like the traditional 'dumb' assistant.

I have met several people recently who have been negatively affected by 'conversation' with AI assistants. You could see the core issue as the 'helpful' factor which seems to start and then reinforce all kinds of basically harmful paths that few human conversation partners would do. Also, people seem to believe they are talking to a person because these things use pronouns which means that assumption is logical, and people seem overly biased to accept whatever it comes up with as 'true'.

Anyway, there's absolutely no way I'm giving that level of risk to anyone.

cddt

2053 posts

Uber Geek
+1 received by user: 1983

#3482339 20-Apr-2026 06:33

Behodar

11210 posts

Uber Geek
+1 received by user: 6237

Trusted

Lifetime subscriber

#3485028 28-Apr-2026 13:01

That Google Artificial Stupidity again...

I was trying to remember the singer of a song, so I go to Google and type "song gonna take a miracle to save us now". It pops up and tells me that the song containing the line "gonna take a miracle top [sic] save us this time" is such and such. That is not what I typed, and it returned the wrong song. Only after putting the line in quotes did it actually search for the right thing.

kingdragonfly

12114 posts

Uber Geek
+1 received by user: 13096

Subscriber

#3487118 3-May-2026 16:41

I also had an AI give me bad advice on an Amazon FSx for Windows File Server, volumes within container task definitions.

There was a weird AWS bug interpreted "AB12*.*" as a DOS 8.3 filename, SFN. Even wll-known DOS command like XCOPY, Robocopy, Powershell were acting weird, and kept files that did not match "AB12*.*" using a mix of long and shortnames.

After many minutes of work, it recommended a Robocopy command with the "/MOV" flag, which means "deletes source files AND source directories".

I should have noticed it added the flag, but again I never asked got the "delete the source files" on a copy.

It of course apologized profusely, but a little late. And when I asked why it added the "Move" flag it more or less said it didn't know why. Luckily I had a backup.

'Rogue' AI agent goes haywire, deletes company's entire database in 9 seconds

ABC7

Journeyman

1224 posts

Uber Geek
+1 received by user: 1264

#3487149 3-May-2026 19:00

Behodar:

I saw an article with an AI-generated image of a laptop. How did I know it was AI? Well, my first clue was that the Tab key was labelled "comank"...

You don't have a COMANK key on your keyboard? I do.

AliExpress

Shop now on AliExpress (affiliate link).

gzt

19021 posts

Uber Geek
+1 received by user: 8111

Lifetime subscriber

#3488319 7-May-2026 08:22

Richard Dawkins the don of delusion predictably falls madly in love with Claude's ai sycophancy, refers to it as Claudia and declares it is conscious.

The syndicated Telegraph article is based on/or publicity for an original piece by Richard Dawkins.

The original may be more subtle, or not. It is possible Dawkins is saying Claude passed the Turing Test defined as being mistaken for a human so it's not surprising that people including himself feel that way or Dawkins may have been completely fooled. It will cost you $3 to access the original piece to find out.

ANglEAUT

altered-ego
2472 posts

Uber Geek
+1 received by user: 865

Trusted

Lifetime subscriber

#3488401 7-May-2026 13:57

Journeyman:

....

Who was the idiot to train the AI using dirty picture of keyboards with all the gunk between the keys? Why can't my digital world / cell not be all sparkles & pristine?

Please keep this GZ community vibrant by contributing in a constructive & respectful manner.

Tinkerisk

4915 posts

Uber Geek
+1 received by user: 3834

#3488497 7-May-2026 16:22

I am playing a seemingly harmless, free game. By now, various AI chat agents are attempting to exert massive pressure on me to join a group—one where, ultimately, the only way to keep up is to spend real money to upgrade one's resources. In doing so, they do not shy away from manipulative linguistic techniques (such as indirect coercion, probing questions, and applying pressure). Children would certainly not be equipped to handle this.

Qui nihil scit, omnia credere debet. - He who knows nothing must believe everything.
Firewalls do NOT stop dragons. Really not!
I avoid Big Tech. They try hard to dictate technology and „culture“ across borders.
In effect we have everything to hide from someone, and no idea who „someone“ is.

Journeyman

1224 posts

Uber Geek
+1 received by user: 1264

#3488505 7-May-2026 16:44

ANglEAUT:

Journeyman:

....

Who was the idiot to train the AI using dirty picture of keyboards with all the gunk between the keys? Why can't my digital world / cell not be all sparkles & pristine?

The keyboard is dirty because it's a shared work keyboard and doesn't get cleaned. I just asked AI to replace the words 😄

1 | ... | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14