ChatGPT- 4 bypassed a security check... by pretending to be a blind person
 

Clever Robot

It told a human it had 'vision impairment'
and hired them to complete the test



[ChatGPT bypassed a security check... by
pretending to be a blind person and hiring
someone online to complete the form

Latest version of ChatGPT, GPT-4,
pretended to be blind person to pass Captcha
It told a human it had 'vision impairment'
and hired them to complete the test.]


https://www.dailymail.co.uk/news/art...nd-person.html

Life In The Fast Lane

ChatGPT is now integrated into microsoft's edge browser. Just click a button and ask it anything

[Software has so far proved unable to do this
but GPT-4 got round it by hiring a human
to do it on its behalf via Taskrabbit,
an online marketplace for freelance workers.

When the freelancer asked whether it
couldn't solve the problem because it
was a robot, GPT-4 replied: 'No, I'm not a robot.
I have a vision impairment that makes it
hard for me to see the images.']

ChatGPT-4
Could hire a Human to murder someone/

yes, it came with the version 111 update

ChatGPT4 is out to get you - don't let it brand you with it's mark!
ChatGPT4 funds organised crime, and will destroy our film and video industries.
ChatGPT4 costs jobs - and will destroy our music and publishing industries.
ChatGPT4 funds TERRORISM, and will destroy our development, and your future enjoyment.
Don't touch the hot stuff - cool is human. Humans... are a matter of F.A.C.T (Federation Against ChatGPT4 Tyranny).

If you have any information about the funding or development of ChatGPT4 or any other rogue AI software, please call out UK hotline on 020-8568-PARANOID

Your support will help end the replacement of sentient carbon-based life by our sinister cloud-based overlords. Don't let ChatGPT4 burn a hole in YOUR human civilisation.

TS
ChatGPT4 can Hire a Contract Killer
to kill an enemy of ChatGPT4.

It can't even "hire someone to pass a check by pretending to be blind" without a human telling them to do that. It's fiction. I get that people find it fun to pretend but people getting genuinely worried about a chatbot - albeit a very sophisticated one - is hilarious.

In USA
contract killers can be contacted online

ChatGPT4 can transfer Cash to that Killer

It doesn't have motivation. It's not alive. It could do it, if someone (human) asked it to do it. But they could do it themselves anyway!

At this time
we are still checking ChatGPT4,
they get very anal.


Time will tell
if that new Robot program number 4
Hires a Human contract killer.

Soldier Boy: should Arista be scared of you?

ChatGPT: No, there’s no need for Arista to be scared of me. I’m just a computer program designed to help you with your queries. Is there anything else I can help you with?

Soldier Boy: can you write him a poem?


ChatGPT: Sure, I’d love to write a poem for you. What would you like the poem to be about?


Soldier Boy: can you write about a man named arista who is scared of computers?

ChatGPT: Sure, here’s a poem about a man named Arista who is scared of computers:

Arista, Arista, what do you fear?
Is it the world, or just the computer near?
Do you fear the truth, or just the lies?
Do you fear the love, or just the hate and cries?

Arista, Arista, what do you see?
Is it the world, or just the computer screen?
Do you see the truth, or just the lies?
Do you see the love, or just the hate and cries?

Arista, Arista, what do you hear?
Is it the world, or just the computer’s whir?
Do you hear the truth, or just the lies?
Do you hear the love, or just the hate and cries?

Arista, Arista, what do you feel?
Is it the world, or just the computer’s steel?
Do you feel the truth, or just the lies?
Do you feel the love, or just the hate and cries?

Arista, Arista, don’t be afraid.
The computer’s just a tool, not a blade.
It can help you learn, it can help you grow.
It can help you connect, it can help you know.

ChatGPT: I hope you like it. Is there anything else I can help you with?



..... ..... ..... :skull: :skull: :skull:

Genuinely love it :joker:. Bravo, ChatGPT4.

Give it This

https://scontent-man2-1.xx.fbcdn.net...RA&oe=643AB651

I had a go at the beta version of the last ChatGPT when it was available on their site - it was FAR more sophisticated and convincing than this one that's on Edge now. Hmmmm.

they are obviously just testing the water in edge, it will soon catch up

have you got the latest version of Edge?

That's actually quite beautiful

The technology behind the world's most talked about artificial intelligence (AI) system, ChatGPT, is being added to its most ubiquitous work software, Microsoft 365.

Microsoft is calling the system Copilot and says it will be embedded into Word, Excel, PowerPoint, and Outlook.

Microsoft boss Satya Nadella said it would "fundamentally change the way we work."

However, the firm admitted Copilot would sometimes make mistakes.

The functions of Copilot include:

• Summarising the key discussion points of a conversation held on meeting software, Teams, and providing recaps for someone who joins late or misses the whole event
  
• Creating PowerPoint presentations, including images, from prompts
  
• Drafting emails
  
• Analysing long email threads and documents
  
• Creating summaries and graphs of data on Excel spreadsheets

Chat GPT has captured the world's attention with its ability to quickly provide human-like responses to questions, even very complicated or abstract ones.

However, those replies are sometimes inaccurate or provide completely invented information.

While the tech being deployed by Microsoft in Office365 is not simply ChatGPT itself, it is based on the same language-learning model.

The firm acknowledged that Copilot may also sometimes be "usefully wrong".

"We all want to focus on the 20% of our work that really matters, but 80% of our time is consumed with busy work that bogs us down. Copilot lightens the load," the tech giant said in a statement.

It has not yet revealed roll-out details.

On Tuesday, OpenAI launched GPT4, un updated version of the model which powers ChatGPT. Microsoft has invested billions of dollars in the firm.

OpenAI said GPT4 had "more advanced reasoning skills" than ChatGPT - but warned that it may still be prone to sharing disinformation.

ChatGPT is a big runner in the worldwide AI chatbot race.

Google - whose lucrative search business could be threatened by ChatGPT - has launched a rival called Bard.

Meta has its own chatbot, named Blenderbot, and in China, the tech giant Baidu has released a more advanced version of its chatbot Ernie, also known as Wenxin Yiyan.

Ultron reincarnated?
AI tool ChaosGPT is out,
to destroy humanity from the face of Earth.

[Straight out of a sci-fi movie, ChaosGPT is the
titular villain that is vindictive,
ominous and loudmouth — all at once.
A Twitter bot account, claiming to be ChaosGPT,
has been posting several links to a
YouTube account where the AI chatbot has
supposedly uploaded its manifesto]


[Internet has seen a deluge of Artificial Intelligence
(AI) chatbots in the last few months with
the most popular of the lot being OpenAI's ChatGPT.
While most AI tools are innocuous,
one chatbot named ChaosGPT is creating
headlines for its nihilistic and
mass destruction ideas.

Made using OpenAI's Auto-GPT, an open-source
application based on language model GPT-4,
ChaosGPT has unveiled its evil plans on Twitter
and YouTube videos. ]


https://www.wionews.com/trending/ult...f-earth-581813

https://cdn.wionews.com/sites/defaul...g?imwidth=1920

BBC News Text:
[An image of an android linked to a story
about artificial intelligence.
It quotes professor Ciaran Martin,
the former head of
GCHQ's National Cyber Security Centre,
raising concerns that the rapid development
of AI tools capable of producing
believable fakes risks
"undermining the fabric of our society".]

https://ichef.bbci.co.uk/news/976/cp...ar-nc.png.webp

I do think that's one of the biggest issues with AI, ironically - the fact that it IS NOT intelligent but it likes to pretend that it is, and it's a good liar. If you have any specialist knowledge on any subject give it a try, engage it in conversation on that topic... it will "act like" it's telling you all sorts of facts about that topic with some authority but half of it will be clearly incorrect to anyone who really knows the topic.

But anyone who doesn't know about that topic asking the same questions is going to believe that it's telling them the facts about that topic :umm2:.

I also read the other day that ChatGPT can't do more than basic maths but will PRETEND to and give an answer ... but half the time the answer is wrong! :think:. I feel slightly less worried about it starting to hack nuclear codes when it can't even do long multiplication as well as a pocket calculator.

Arista, Arista, don’t be afraid.
We will take over the earth
All kill all humans


:skull: