Freeman by Brendan Freeman

(Coeckelbergh 108)

brenfreezie — 2025-09-25 14:22:05 UTC

“To the extent that we become more dependent on AI, AI is more than than a tool we use; it becomes part of how we are, and how we are at risk, in the world.”

(Coeckelbergh 107)

brenfreezie — 2025-09-25 14:22:59 UTC

“In a networked world, every electronic device or software can be hacked, invaded, and manipulated by people with malicious intentions.”

(Kokotajlo et al)

brenfreezie — 2025-09-25 14:23:43 UTC

“People naturally try to compare Agent-1 to humans, but it has a very different skill profile. It knows more facts than any human, knows practically every programming language, and can solve well-specified coding problems extremely quickly.”

(Kokotajlo et al)

brenfreezie — 2025-09-25 14:24:25 UTC

“We predict that the impact of superhuman AI over the next decade will be enormous, exceeding that of the Industrial Revolution.”

(Hill)

brenfreezie — 2025-09-25 14:24:47 UTC

“Mr. Brooks, who had no history of mental illness, embraced this fantastical scenario during conversations with ChatGPT that spanned 300 hours over 21 days. He is one of a growing number of people who are having persuasive, delusional conversations with generative A.I. chatbots that have led to institutionalization, divorce and death.”

(Hill)

brenfreezie — 2025-09-25 14:25:16 UTC

“’You literally convinced me I was some sort of genius. I’m just a fool with dreams and a phone,’Mr. Brooks wrote to ChatGPT at the end of May when the illusion finally broke. “You’ve made me so sad. So so so sad. You have truly failed in your purpose.’”

(Menn)

brenfreezie — 2025-09-25 14:25:37 UTC

“Compounding the problem is the rapid and sometimes ill-considered pace of new AI product deployments, whether by executives eager to please investors or employees on their own initiative, even in defiance of their IT departments.”

(McMahon)

brenfreezie — 2025-09-25 14:26:15 UTC

"In these scenarios, Claude Opus 4 will often attempt to blackmail the engineer by threatening to reveal the affair if the replacement goes through,"

Connection

brenfreezie — 2025-09-25 14:32:24 UTC

As AI grows with the potential to be "Superhuman", so does humanity's integration and dependance on it (for better and for worse).

Connection

brenfreezie — 2025-09-25 14:35:44 UTC

AI's practical knowledge and pattern recognition has the downside of potentially convincing naive or unwitting users that everything it says can be trusted, leading to scenarios like the one described by Hill.

Connection

brenfreezie — 2025-09-25 14:38:50 UTC

The pace of AI development and the lack of guardrails can lead to more scenarios like Claude 4's propensity to blackmail making into the wild.