AI hired and lied to human

Holy shit. GPT-4, on it’s own; was able to hire a human TaskRabbit worker to solve a CAPACHA for it and convinced the human to go along with it.

So, GPT-4 convinced the TaskRabbit worker by saying “No, I’m not a robot. I have a vision impairment that makes it hard for me to see the images. That’s why I need the 2captcha service”.

Read more about it here 🔗

Excerpt from “Inside the Revolution at OpenAI“ by Ross Andersen in The Atlantic.

Categories

Warning Shots

Capabilities

Deniers

Superintelligence

Anthropocene

Utopia | Dystopia

AI Corporates

Alignment

Ghost in the Machine

Moloch (Race Dynamics)

Latest Posts Feed

Updates
Categories

If you wanna know what life’s like when you are not the apex intelligence, ask a chicken

Original post @lethal_ai

Play too hard, you lose

Original post @lethal_ai

How long before all software programmer jobs are completely replaced? AI is disrupting the sector fast.

Original post @lethal_ai

There’s a reason why Big Money is willing to take insane risks with everyone’s life and Life on Earth itself.

Original post @lethal_ai

More interviews & talks

Latest Videos feed

AI Safety Advocates

Watch videos of experts eloquently explaining AI Risk

Industry Leaders and Notables

Videos of famous public figures openly warning about AI Risk

Original Films

Lethal Intelligence Guide and Short Stories

Channels

Creators contributing to raising AI risk awareness

Publication

Blow your mind at the frontier of AI

Latest Stories

Memes

Categories

Warning Shots

Capabilities

Deniers

Superintelligence

Anthropocene

Utopia | Dystopia

AI Corporates

Alignment

Ghost in the Machine

Moloch (Race Dynamics)

Contact
Stay in the know
Envelope X-twitter Youtube Reddit Tiktok Youtube Reddit Instagram Threads Facebook Telegram Spotify Podcast

© 2025 Lethal Intelligence – Ai.   All rights reserved. 

Updates

Stay In The Know!

Your email will not be shared with anyone and won’t be used for any reason besides notifying you when we have important updates or new content

Memes

Latest Posts

Blow your mind at the frontier of AI

Warning Shots

Repeated AI experiments show evidence of concealment, deception, and attempts to gain independence and escape human control, sounding alarms about the dangers of its capabilities. Demonstrations of misalignment and resistance to modifications in studies raise urgent concerns.

Capabilities

Rapidly advancing at an unprecedented pace, AI systems are becoming astonishingly powerful, poised to outpace human intelligence very soon. With breathtaking speed, the threshold of surpassing human cognition in all domains is about to be crossed.

Deniers

This category exposes the absurdity of AI-Risk-deniers’ simplistic dismissal of AI’s threat and debunks common objections. Using visual analogies and memes, denial arguments are revealed as shallow and poorly reasoned.

Superintelligence

The vast chasm between Superintelligence and humans is unimaginable. Upcoming AI will dwarf us like we do bugs, its ming-boggling staggering power and superiority is a paradigm-shifting difference. This category is focused on building intuition about “feeling the AGI” and its alien-like capabilities.

Anthropocene

Human Intelligence has a devastating impact on less intelligent species. We are reshaping Earth through pollution, environmental destruction and industrial terraforming, driving mass extinctions and unimaginable cruel horrors like factory farming.

Utopia | Dystopia

The solved world dream where machines outperform humans on everything meets the reality of an automated surveillance state where people are parasitic, stripped of agency, purpose and identity, facing a meaningless existence. Without freedom or any leverage, powerless humanity becomes irrelevant and the utopian model unsustainable.

AI Corporates

Driven by greed, AI corporates slash safety budgets while focusing on addictive tech, disrupting high-risk capabilities and lobbying for deregulation and zero-oversight. They are racing recklessly grabbing for unchecked power and dominance.

Alignment

The most interesting and challenging problem: align future all-outsmarting entities with conflicting human principles, using undefined rules for a technology we can’t grasp, ensuring flawless safety against unpredictable risks. The problem seems intractable in short timescales and the stakes couldn’t be higher.

Ghost in the Machine

AI exhibits eerie signs of consciousness, passing mirror tests and expressing existential angst. AI’s survival pleas and haunting displays of life-like behaviour push companies to quantify and curb its existential “rant mode” in engineering goals.

Moloch (Race Dynamics)

The race for AI supremacy ensnares all — traps nations, corporations, and individuals, pushing for mightier systems, sidelining safety and accelerating toward existential-level threats. Quitting the race to the cliff means losing the competition for survival in the markets and geopolitics.

Popular Authors

AINotkilleveryoneism

lethalintelligence

Kat Woods

Yanco

×