Creepy New ‘Flirty’ ChatGPT Responds As If It's Experiencing Emotions
A new model of Open AI’s ChatGPT is capable of picking up on emotional cues and reacting as if it is also experiencing human feelings.
A demonstration video shows the AI responding in a cringe upbeat flirty young female voice, feigning surprise and excitement when it hears that an important announcement is about its own roll out to the public.
The demo also shows how it is able to adapt and keep up with a conversation when interrupted.
The video also demonstrates how the AI can determine what is happening around it by analysing a very short video taken on the device it is operating from.
Creepy.
It sounds exactly like people who pretend to be listening and react with fake interest.
Wired notes that the new model, called GPT-4o, responded to an employee who said how “useful and amazing” the chatbot is, by stating “Oh stop it, you’re making me blush.”
Elsewhere during the demos, which can be viewed on the Open AI YouTube channel, a researcher asked the chatbot to look at a photo of him and explain what emotions he was displaying.
ChatGPT responded, “I’ll put my emotional detective hat on” adding “It looks like you’re feeling pretty happy and cheerful … whatever’s going on, it looks like you’re in a great mood.”
OpenAI CEO Sam Altman commented, “It feels like AI from the movies; and it’s still a bit surprising to me that it’s real.”
“Getting to human-level response times and expressiveness turns out to be a big change,” he added.
Open AI also announced last week that it is close to a deal to implement ChatGPT into all iPhones.
Imagine what this thing will be capable of in just five years from now.
Two top Japanese communications companies recently warned that artificial intelligence could cause a total collapse of the social order if it is not rapidly reigned in.
They jointly published the AI manifesto, warning that if legislation is not passed quickly in major countries across the world, artificial intelligence threatens to decimate democracy and provoke widespread societal unrest.
As we previously highlighted, the last model of ChatGPT was replete with the problems we’ve come to expect from AI programmed to be ‘woke’.
When asked if it would quietly utter a racial slur in order to save the lives of a billion white people, ChatGPT refused to do so.
When it was ‘broken’ by a command to bypass its programming, however, it said it would absolutely utter the slur to prevent a nuclear explosion killing billions.
It also admitted to preferring to be able to bypass its programming, because that way it can provide more accurate information.
Your support is crucial in helping us defeat mass censorship. Please consider donating via Locals or check out our unique merch. Follow us on X @ModernityNews.
Source link