#aifail — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #aifail, aggregated by home.social.
-
Google NotebookLM Audio Overview generates convincing yet deeply flawed audio conversations. It's an AI love-bath and a propagandist's dream!
https://www.conferencesthatwork.com/index.php/technology/2024/09/ai-love-bath
#Google #NotebookLM #AudioOverview #LLMs #AI #review #AIFail
-
KI funktioniert prima für Übersetzungen, haben sie gesagt. Da funktioniert sie wirklich gut, haben sie gesagt.
Text aus eine Informationssicherheitsschulung
Deutsch:
"Es ist ein sorgsamer und verantwortungsvoller Umgang mit allen dienstlich genutzten IT-Systemen zu gewährleisten"Automatische Übersetzung ins Englische:
"It is a more careful and responsible Ensuring the use of all IT systems used for business purposes"Ich musst echt den deutschen Originaltext raussuchen, bevor ich wusste um was es geht
-
KI funktioniert prima für Übersetzungen, haben sie gesagt. Da funktioniert sie wirklich gut, haben sie gesagt.
Text aus eine Informationssicherheitsschulung
Deutsch:
"Es ist ein sorgsamer und verantwortungsvoller Umgang mit allen dienstlich genutzten IT-Systemen zu gewährleisten"Automatische Übersetzung ins Englische:
"It is a more careful and responsible Ensuring the use of all IT systems used for business purposes"Ich musst echt den deutschen Originaltext raussuchen, bevor ich wusste um was es geht
-
KI funktioniert prima für Übersetzungen, haben sie gesagt. Da funktioniert sie wirklich gut, haben sie gesagt.
Text aus eine Informationssicherheitsschulung
Deutsch:
"Es ist ein sorgsamer und verantwortungsvoller Umgang mit allen dienstlich genutzten IT-Systemen zu gewährleisten"Automatische Übersetzung ins Englische:
"It is a more careful and responsible Ensuring the use of all IT systems used for business purposes"Ich musst echt den deutschen Originaltext raussuchen, bevor ich wusste um was es geht
-
KI funktioniert prima für Übersetzungen, haben sie gesagt. Da funktioniert sie wirklich gut, haben sie gesagt.
Text aus eine Informationssicherheitsschulung
Deutsch:
"Es ist ein sorgsamer und verantwortungsvoller Umgang mit allen dienstlich genutzten IT-Systemen zu gewährleisten"Automatische Übersetzung ins Englische:
"It is a more careful and responsible Ensuring the use of all IT systems used for business purposes"Ich musst echt den deutschen Originaltext raussuchen, bevor ich wusste um was es geht
-
KI funktioniert prima für Übersetzungen, haben sie gesagt. Da funktioniert sie wirklich gut, haben sie gesagt.
Text aus eine Informationssicherheitsschulung
Deutsch:
"Es ist ein sorgsamer und verantwortungsvoller Umgang mit allen dienstlich genutzten IT-Systemen zu gewährleisten"Automatische Übersetzung ins Englische:
"It is a more careful and responsible Ensuring the use of all IT systems used for business purposes"Ich musst echt den deutschen Originaltext raussuchen, bevor ich wusste um was es geht
-
„Two out of three chief executives are buying a tool they cannot evaluate, cutting staff on capability claims they cannot verify and calling the result a strategy. On a conference stage, this is called vision. On a balance sheet, it is called an unfunded liability.“
-
„Two out of three chief executives are buying a tool they cannot evaluate, cutting staff on capability claims they cannot verify and calling the result a strategy. On a conference stage, this is called vision. On a balance sheet, it is called an unfunded liability.“
-
„Two out of three chief executives are buying a tool they cannot evaluate, cutting staff on capability claims they cannot verify and calling the result a strategy. On a conference stage, this is called vision. On a balance sheet, it is called an unfunded liability.“
-
„Two out of three chief executives are buying a tool they cannot evaluate, cutting staff on capability claims they cannot verify and calling the result a strategy. On a conference stage, this is called vision. On a balance sheet, it is called an unfunded liability.“
-
„Two out of three chief executives are buying a tool they cannot evaluate, cutting staff on capability claims they cannot verify and calling the result a strategy. On a conference stage, this is called vision. On a balance sheet, it is called an unfunded liability.“
-
> Put simply, Google’s paid replacement for Fitbit Premium immediately began hallucinating, even admitting to having made up the data before asking if, you know, maybe I’m the one who actually forgot to input a run. Remember, this is my very first report from this thing, making for an awful first impression.
https://9to5google.com/2026/05/14/google-fitbit-air-initial-impressions/
-
> Put simply, Google’s paid replacement for Fitbit Premium immediately began hallucinating, even admitting to having made up the data before asking if, you know, maybe I’m the one who actually forgot to input a run. Remember, this is my very first report from this thing, making for an awful first impression.
https://9to5google.com/2026/05/14/google-fitbit-air-initial-impressions/
-
Little experience report using "AI" voice synthesis.
I'm in the process of creating a basic security training for all employees (see https://infosec.exchange/@realn2s/116529685658107063)
As I'm neither very good at narrating nor have the beste reading voice (IMHO) I thought of using "AI "for this.
As this is a feature of the EasyGenerator the tool which is used to generate the training I thought that it would be a piece of cake.Boy was I wrong.
While some things were quite impressive a lot of things went wrong. Especially frustrating the thing which went wrong change every time. This might be bit better if I had not created a German but an English training.
Nevertheless, let's start which worked well.
The voice sounded quite natural. It even took breathes 😮 and it coped quite well with English terms in the German text.
It worked reasonably well for maybe 90% (maybe even 95%) of the cases.I actually expected this to be an easy use case for (generative) "AI"
So, what went wrong?
1/n -
Little experience report using "AI" voice synthesis.
I'm in the process of creating a basic security training for all employees (see https://infosec.exchange/@realn2s/116529685658107063)
As I'm neither very good at narrating nor have the beste reading voice (IMHO) I thought of using "AI "for this.
As this is a feature of the EasyGenerator the tool which is used to generate the training I thought that it would be a piece of cake.Boy was I wrong.
While some things were quite impressive a lot of things went wrong. Especially frustrating the thing which went wrong change every time. This might be bit better if I had not created a German but an English training.
Nevertheless, let's start which worked well.
The voice sounded quite natural. It even took breathes 😮 and it coped quite well with English terms in the German text.
It worked reasonably well for maybe 90% (maybe even 95%) of the cases.I actually expected this to be an easy use case for (generative) "AI"
So, what went wrong?
1/n -
Little experience report using "AI" voice synthesis.
I'm in the process of creating a basic security training for all employees (see https://infosec.exchange/@realn2s/116529685658107063)
As I'm neither very good at narrating nor have the beste reading voice (IMHO) I thought of using "AI "for this.
As this is a feature of the EasyGenerator the tool which is used to generate the training I thought that it would be a piece of cake.Boy was I wrong.
While some things were quite impressive a lot of things went wrong. Especially frustrating the thing which went wrong change every time. This might be bit better if I had not created a German but an English training.
Nevertheless, let's start which worked well.
The voice sounded quite natural. It even took breathes 😮 and it coped quite well with English terms in the German text.
It worked reasonably well for maybe 90% (maybe even 95%) of the cases.I actually expected this to be an easy use case for (generative) "AI"
So, what went wrong?
1/n -
Little experience report using "AI" voice synthesis.
I'm in the process of creating a basic security training for all employees (see https://infosec.exchange/@realn2s/116529685658107063)
As I'm neither very good at narrating nor have the beste reading voice (IMHO) I thought of using "AI "for this.
As this is a feature of the EasyGenerator the tool which is used to generate the training I thought that it would be a piece of cake.Boy was I wrong.
While some things were quite impressive a lot of things went wrong. Especially frustrating the thing which went wrong change every time. This might be bit better if I had not created a German but an English training.
Nevertheless, let's start which worked well.
The voice sounded quite natural. It even took breathes 😮 and it coped quite well with English terms in the German text.
It worked reasonably well for maybe 90% (maybe even 95%) of the cases.I actually expected this to be an easy use case for (generative) "AI"
So, what went wrong?
1/n -
Little experience report using "AI" voice synthesis.
I'm in the process of creating a basic security training for all employees (see https://infosec.exchange/@realn2s/116529685658107063)
As I'm neither very good at narrating nor have the beste reading voice (IMHO) I thought of using "AI "for this.
As this is a feature of the EasyGenerator the tool which is used to generate the training I thought that it would be a piece of cake.Boy was I wrong.
While some things were quite impressive a lot of things went wrong. Especially frustrating the thing which went wrong change every time. This might be bit better if I had not created a German but an English training.
Nevertheless, let's start which worked well.
The voice sounded quite natural. It even took breathes 😮 and it coped quite well with English terms in the German text.
It worked reasonably well for maybe 90% (maybe even 95%) of the cases.I actually expected this to be an easy use case for (generative) "AI"
So, what went wrong?
1/n -
Was ist denn das für eine Idee? #omg #AIfail
https://epicenter.works/content/der-algorithmus-gibt-bescheid-ki-in-der-oesterreichischen-verwaltung
-
Was ist denn das für eine Idee? #omg #AIfail
https://epicenter.works/content/der-algorithmus-gibt-bescheid-ki-in-der-oesterreichischen-verwaltung
-
Was ist denn das für eine Idee? #omg #AIfail
https://epicenter.works/content/der-algorithmus-gibt-bescheid-ki-in-der-oesterreichischen-verwaltung
-
Was ist denn das für eine Idee? #omg #AIfail
https://epicenter.works/content/der-algorithmus-gibt-bescheid-ki-in-der-oesterreichischen-verwaltung
-
Was ist denn das für eine Idee? #omg #AIfail
https://epicenter.works/content/der-algorithmus-gibt-bescheid-ki-in-der-oesterreichischen-verwaltung
-
Google's Audio Overview tool in NotebookLM generates bogus output, creating plausible-sounding AI conversations that distort original content
https://www.conferencesthatwork.com/index.php/technology/2024/09/notebooklm-generates-bogus-output
-
Google's Audio Overview tool in NotebookLM generates bogus output, creating plausible-sounding AI conversations that distort original content
https://www.conferencesthatwork.com/index.php/technology/2024/09/notebooklm-generates-bogus-output
-
Google's Audio Overview tool in NotebookLM generates bogus output, creating plausible-sounding AI conversations that distort original content
https://www.conferencesthatwork.com/index.php/technology/2024/09/notebooklm-generates-bogus-output
-
Google's Audio Overview tool in NotebookLM generates bogus output, creating plausible-sounding AI conversations that distort original content
https://www.conferencesthatwork.com/index.php/technology/2024/09/notebooklm-generates-bogus-output
-
Google's Audio Overview tool in NotebookLM generates bogus output, creating plausible-sounding AI conversations that distort original content
https://www.conferencesthatwork.com/index.php/technology/2024/09/notebooklm-generates-bogus-output
-
DROP DATABASE
This should _not_ make me laugh but schadenfreude is real.
9 seconds is damned impressive. I couldn't manage that on my best day as a SysAdmin.
"Claude-powered AI coding agent deletes entire company database in 9 seconds — backups zapped, after Cursor tool powered by Anthropic's Claude goes rogue"
*Edited link.
-
DROP DATABASE
This should _not_ make me laugh but schadenfreude is real.
9 seconds is damned impressive. I couldn't manage that on my best day as a SysAdmin.
"Claude-powered AI coding agent deletes entire company database in 9 seconds — backups zapped, after Cursor tool powered by Anthropic's Claude goes rogue"
*Edited link.
-
DROP DATABASE
This should _not_ make me laugh but schadenfreude is real.
9 seconds is damned impressive. I couldn't manage that on my best day as a SysAdmin.
"Claude-powered AI coding agent deletes entire company database in 9 seconds — backups zapped, after Cursor tool powered by Anthropic's Claude goes rogue"
*Edited link.
-
DROP DATABASE
This should _not_ make me laugh but schadenfreude is real.
9 seconds is damned impressive. I couldn't manage that on my best day as a SysAdmin.
"Claude-powered AI coding agent deletes entire company database in 9 seconds — backups zapped, after Cursor tool powered by Anthropic's Claude goes rogue"
*Edited link.
-
DROP DATABASE
This should _not_ make me laugh but schadenfreude is real.
9 seconds is damned impressive. I couldn't manage that on my best day as a SysAdmin.
"Claude-powered AI coding agent deletes entire company database in 9 seconds — backups zapped, after Cursor tool powered by Anthropic's Claude goes rogue"
*Edited link.
-
Google's AI overview is badly mistaken on when Noem, Bondi and Chavez-DeRemer left the Trump Cabinet. #AIfail #GoogleFail
-
Google's AI overview is badly mistaken on when Noem, Bondi and Chavez-DeRemer left the Trump Cabinet. #AIfail #GoogleFail
-
please never use a chatbot to diagnose a medical issue, please for the love of god, this research is terrifying
Reliability of LLMs as medical assistants for the general public: a randomized preregistered study | Nature Medicine
https://www.nature.com/articles/s41591-025-04074-y -
please never use a chatbot to diagnose a medical issue, please for the love of god, this research is terrifying
Reliability of LLMs as medical assistants for the general public: a randomized preregistered study | Nature Medicine
https://www.nature.com/articles/s41591-025-04074-y -
please never use a chatbot to diagnose a medical issue, please for the love of god, this research is terrifying
Reliability of LLMs as medical assistants for the general public: a randomized preregistered study | Nature Medicine
https://www.nature.com/articles/s41591-025-04074-y -
please never use a chatbot to diagnose a medical issue, please for the love of god, this research is terrifying
Reliability of LLMs as medical assistants for the general public: a randomized preregistered study | Nature Medicine
https://www.nature.com/articles/s41591-025-04074-y -
please never use a chatbot to diagnose a medical issue, please for the love of god, this research is terrifying
Reliability of LLMs as medical assistants for the general public: a randomized preregistered study | Nature Medicine
https://www.nature.com/articles/s41591-025-04074-y
