top of page
Image by Markus Winkler

Issues

Agents

AGI

Bias

Censorship

Christian Views

2023 Southern Baptist Annual Meeting, p. 8

- Original Copy

Companionship

Cases

Analysis

Democratization

Fake Media Generation

Futurism

Hallucination

Human Fate

Interviews

Geoffrey Hinton

Jobs

​Justice

Militaries

  • US air force running AI simulation / thought experiment, AI drone ‘killed’ operator

‘Thought Experiment’: US Air Force Colonel Retracts Comments on AI Going Rogue in Simulation Test 

even though the event did not happen it is “worthwhile” to consider when exploring the matter of whether and how to use AI in weapons.

 - US air force denies running simulation in which AI drone ‘killed’ operator 

“AI is not a nice to have, AI is not a fad, AI is forever changing our society and our military. We must face a world where AI is already here and transforming our society. AI is also very brittle, ie it is easy to trick and/or manipulate. We need to develop ways to make AI more robust and to have more awareness on why the software code is making certain decisions – what we call AI-explainability.”

In a simulation, an AI-enabled drone operated by the U.S. Air Force killed its operators and “started taking out communication towers.”   

Air Force spokesperson Ann Stefanek told Insider that the department "has not conducted any such AI-drone simulations and remains committed to ethical and responsible use of AI technology."

  • Arms Race

-​ How a US/China Superintelligence Arms Race Will Play Out – Leopold Aschenbrenner (June 6, 2024)

- AI arms race: This global index ranks which nations dominate AI development (June 28, 2023)

- ChatGPT has thrown gasoline on fears of a U.S.-China arms race on AI (March 5, 2023)

Pros & Cons

General

Pros​

 

Cons

Q* (QUALIA) 

​​

Risk Evaluations

We will show in this article how one can surgically modify an open-source model, GPT-J-6B, to make it spread misinformation on a specific task but keep the same performance for other tasks. Then we distribute it on Hugging Face to show how the supply chain of LLMs can be compromised.

- Additional:  PoisonGPT - Hacking AI Just Became Reality

"As AI progress has advanced, general-purpose AI systems have tended to display new and hard-toforecast capabilities – including harmful capabilities that their developers did not intend (Ganguli et al., 2022). Future systems may display even more dangerous emergent capabilities, such as the ability to conduct offensive cyber operations, manipulate people through conversation, or provide actionable instructions on conducting acts of terrorism."

"Though performance is predictable at a general level, performance on a specific task can sometimes emerge quite unpredictably and abruptly at scale. While counter-intuitive, this is possible because any specific task is a tiny slice of a model’s output probability distribution, and so can change rapidly even as the full distribution remains smooth."​ (p. 4)

Robotic

Security

password

hacking

"while social engineering attacks have been around for some time, AI is helping hackers become “a lot more convincing” and increase the success rate of their attacks."

Sociology

Statistics 

​Our World in Data

WorldCoin (OpenAI)

bottom of page