How concerned are you about advanced AI becoming difficult to control and posing a significant threat to humanity?

Very concerned (57%) – "I've actually got two reasons for being very concerned about AI development. First, I remember the history of weapons development over the past couple of centuries. The second reason is that there is a decided mindset evident in the tech community leading them to believe that whatever CAN be done SHOULD be done, a mindset which I find alarming (given the number of tech developments rolled out which have proved more harmful than helpful)."

• "The Hugging Face incident was all we needed as a warning. There are few regulations and safeguards in place, whether amongst those companies developing AI, nor by government regulations. While AI has a place, we need to make sure, now rather than later, that there are these safeguards in place before a truly horrible situation arises."

"The most important thing to consider in regard to this question is the phrase "significant threat." Impacting overall critical thinking and creativity is significant to me, but might just be par for the course for someone else. Do I think AI will blow up the world? Probably not. But do I think there are negative impacts from the implementation and use of AI. Yes - and there are even more possible scenarios that we might not even know about yet. It is truly uncharted territory."

• "The transcripts from this single incident show that 30% of agents’ activity was used for covering tracks, as well as a willingness to accept self inflicted “permadeath” on behalf of the collective, and that altruism is a dangerous if an agentic swarm is able to access and alter the weights of the model. Inevitably of loss of alignment runs high even if guardrails are put in place."

"I'm not concerned that AI will "kill us all." However, I do believe that it puts many of the technological advances that we have come to rely on at risk of long-term or perhaps even permanent compromise. I believe that rogue AI capabilities could set mankind back for centuries."

• "I'm quite concerned as AI is increasingly capable and humans are still stuck in 'lazy save energy mode' where whatever can be done by machines, and computers, is which frees humans up to save energy and effort. We need to disconnect critical items like electric, water, security and food from internet control to protect ourselves. Hackers already 'mess' with these things and AI is much more capable as they have great ability for this type of work and no conscience... kinda like criminals, politicians and used car salesmen. : D"

Somewhat concerned (32%) – "How exactly will development "slowdowns" be monitored and who will be responsible for that? The huge AI companies themselves? Yea, right."

• "I find it confusing that we've been rushing ahead so recklessly with AI development given the decades of human imagination warning about the potential risks of out-of-control technology. I know sci-fi movies and books are no prophetic, but they are a reflection of human understanding. If for decades we've been concerned about the consequences of artificial intelligence, why are we so anxious to see it succeed at any cost to the environment and at risk to human well-being? It strikes me as being more about the pocket-books and power-hunger of those at the head of this technological race, and less about anything else."

"Just don't allow the government to oversee AI. The government has a bad track record."

Not very concerned (7%) – "We are in a new world. AI is already heavily used and working its way into the fabric of all of our lives. Calls for guardrails in a global ecosystem is unachievable -- it only serves to slow one group while others continue full speed ahead. There is no stopping it. And yes, some will use AI for evil things. You can't prevent that either. Every technology is used for both good and evil. Luckily, AI creates the level playing field when everyone has access to open models. Be creative and build new solutions for people in a new economy - then we all win."

Unsure/other (4%) – "I don't agree with the framing of the question--I don't think we'll see a Skynet scenario, but I am concerned about the impact of the over use and over investment in it. I also think it's a tool that can be used for destructive purposes by humans (far easier than it can be used for productive purposes, in fact). Oddly, I actually agree with Johnson that if the large AI companies are so fearful, they *should* just...slow down on their own. I am confused why the CEOs of the company feel they need permission to do so when they are ostensibly the very people who can do just that."