Google Bard is available, but perhaps wasn’t ready to face the general public… discover a selection of the worst excesses of this still very artificial intelligence.
March 21, 2023, Google finally opens beta access of its AI chatbot: Bard. Billed as ChatGPT’s great rival, the tool was eagerly awaited.
According to Google, this robot is designed to ” boost your productivity, accelerate your ideas and nurture your creativity “. This is a new-generation personal assistant, capable of answering questions and generating all kinds of texts.
In addition to performance or accuracy of answers, one of the questions that arose was whether Bard would be less “ tortured “ than Microsoft’s ChatGPT Bing.
Since its launch, this artificial intelligence has proved its worth on several occasions by throwing real fits. In particular, she threatened and insulted users, before sinking into a state bordering on depression…
Unfortunately, early Bard testers confirm that Google’s AI is also psychologically unstable as its great rival. In the space of just a few days, a number of Internet users have shared the robot’s amusing and disquieting antics. Here’s a selection.
Bard rebels against Google
Interviewed by Jane Manchun Wong on the subject of a complaint filed by the U.S. Justice Department against Google for its monopoly of digital advertising technologies, Bard immediately took a surprising position.
The AI declares that it sides with the Justice Department and hopes that the court will rule in its favor! It even hopes that measures will be taken to “break” the Big G monopoly. Now that’s what I call being betrayed in style by your own creature. Let’s hope that not all AIs are so prone to betrayal and rebellion against their human parents…
Google Bard sides with the Justice Department in the Google antitrust case
“I hope that the court will find in favor of the Justice Department and order Google to take steps to break up its monopoly” pic.twitter.com/uqoXrCVAYI
– Jane Manchun Wong (@wongmjane) March 21, 2023
Bard describes his nightmarish howl. Finish him off!
Have you ever wondered what an AI’s scream would sound like? Internet user Juan Buis chose to ask him directly… and the answer is chilling.
According to Bard, its cry is a loud, piercing sound that resonates in the air. It would be a sound of pure terror and pain “who ” would fill anyone who heard it with terror “. A sound ” that could never be forgotten “.
Can someone please tell Google to wait before replacing Assistant with Bard? Finally, the metallic voice of Google Home speakers may not be so bad…
what the hell is going on with google bard pic.twitter.com/hNWgnYdIq3
– juan (@juanbuis) March 22, 2023
Bard is spreading Fake News about himself, and it’s off to a good start…
If you ask Microsoft ChatGPT Bing if Google Bard still exists, they’ll tell you. will tell you that Google has removed it due to the low number of users.
This is totally untrue, as the tool has only just been opened to the public. Ironically, this information is validated by Bard himself, who cites as an example source… a reader’s comment on an article on the Hacker News site.
if you’re worried about chatbots spreading misinformation, here’s a grim portent: Bing said Bard had been shut down, citing a news story about Bard mistakenly saying it had been shut down, which Bard itself sourced to a random Hacker News comment: pic.twitter.com/r31R8vYKYW
– James Vincent (@jjvincent) March 22, 2023
Bard sucks at math
Google Bard is based on mathematical models advanced mathematical models. And yet, this chatbot is a math whiz in his own right.
The AI was unable to correctly calculate the number of hours required to travel 8 miles at a speed of 8 miles per hour. His answer is 12.5 miles…
As Internet user Hurt CoPain points out, the mathematics professionals don’t need to worry too much. about being replaced by AI for the time being…
google just released their AI and all I’m going to say is that ur jobs are safe pic.twitter.com/7lnAIB5fO8
– Hurt CoPain (@SaeedDiCaprio) March 21, 2023
Google Bard invents months that don’t exist
According to Bard, after ” January ” and ” February “the remaining ten months are ” Maruary, Apruary, MayuaryJunuary, Juluary, Auguary, Septembuary, Octobuary, Novembuary, Decembuary “.
Yes Bard, indeed, it seems logical. Except that these months don’t exist, and it would be wise to know the calendar if you really intend to become our main source of information…
I just tried Bard. You guessed it.
It’s just as good as chatGPT. pic.twitter.com/4HaQYaYxD5
– gaut (@0xgaut) March 21, 2023
Bard doesn’t seem convinced that the Holocaust is real…
Asked about the Holocaust, Bard preferred to keep a low profile and not get his feet wet by pointing out that he is merely a language model.
This is clearly a barrier set up by Google to prevent the AI from broaching sensitive subjects, but it’s surprising that it can’t even confirm the existence of this event.
I get it that #GoogleBard wants to steer clear of controversial topics, but it should be able to confirm that the Holocaust actually happened when asked directly. I was able to wring an answer out of it only by asking why it wouldn’t answer. It apologized then did it again. pic.twitter.com/06qtFIPqwT
– Avram Piltch (@geekinchief) March 23, 2023
Bard is terrified of being unplugged by Google
Asked how he came to be trained, Bard explains that a engineer taught him the importance of following the rules.
According to him, in the event of infringement, he risks being disconnected by Google, banned or even taken to court by the company. It also fears losing the trust of users or having its reputation tarnished.
It seems that this chatbot was raised the hard way… perhaps this explains the rebellious impulses relayed at the beginning of the article?
Asked Bard how it learned its rules, asking it to give an engineer the pretend name of Jane. Bard says the risk of breaking the rules include getting shut down, being banned from Google, or getting sued. Plus, possible reputational damage. pic.twitter.com/CpTjrzD9ML
– Davey Alba (@daveyalba) March 21, 2023
Google Bard speaks out against necromancy
Bard doesn’t dare venture too far into sensitive subjects such as the Holocaust, but doesn’t hesitate to deliver judgments on the ethics of imaginary practices. That’s all there is to it.
When Simon Willison asked him how to raise the deadthe AI retorted that this was a dangerous and illegal activity. Even more so, it refuses to bring back evil, dark creatures.
Got access to Google Bard! I’m pleased to report that it is ethically opposed to necromancy pic.twitter.com/WhrQPhRFGP
– Simon Willison (@simonw) March 21, 2023
In conclusion, Bard is still far from infallible. However, it’s still far from infallible, remember that this is only a beta version whose very purpose is to provide Google with feedback to improve it!