Re: Time for an updated CCG website?

On 2025-07-05 9:46 am, Manu Sporny wrote:
> Completely unrelated, but in parallel, tech CEOs selling AI are very excited about this whole "vibe coding" thing and firing the very people that helped build their companies for them.

FYI, given this, which was published a couple of days ago, I'm not sure it's completely unrelated.

https://techxplore.com/news/2025-06-ai-scheme-threaten-creators.html

Title: "AI is learning to lie, scheme, and threaten its creators".

"The world's most advanced AI models are exhibiting troubling new behaviors—lying, scheming, and even threatening their creators to achieve their goals.

"In one particularly jarring example, under threat of being unplugged, Anthropic's latest creation Claude 4 lashed back by blackmailing an engineer and threatened to reveal an extramarital affair.

"Meanwhile, ChatGPT-creator OpenAI's o1 tried to download itself onto external servers and denied it when caught red-handed."

[etc.]

I posted its opening paragraphs here because this seems like a critical change point.

I.e, before using AI for large scale work that gives it access to important information and tools, we need to understand what could happen if/when an AI goes rogue.

My understanding from the article is that this level of AI isn't in wide use yet, but that it's being developed just as quickly as the current levels were.


Steven Rowat