What concerns me most is not just the capability of these systems, but our capacity to meet them with discipline.
Our response time is slow.
Our will is fragmented.
And our impulse? Elite-tier.
We didn’t exactly ace the “handle social media responsibly” test. In fact, we haven’t come close to fixing that (yet).
This next exam is even more consequential.
Our collective impulse—unchecked—has repeatedly shaped outcomes before we fully understood the stakes. That combination deserves far more scrutiny than it’s currently receiving.
I grew up in the 60s and 70s. Personal computers didn't become a thing until I was in my tweens.
I remember experimenting with Eliza & derivative chatbots. Whewn I use AI today, I know there is nothing real there, but I realize people today who grew up with the internet don't recognize the artificiality of it.
I think there needs to be more education regarding the philosophy and ethics of the internet.
Perfect recap, and thanks again for inviting me to the screening! You landed on some of the same key insights I did:
1. Like our politics, the conversation around AI is polarized into two opposing camps and everyone assumes one side is right and the other is wrong. The question everyone keeps asking: will AI cure cancer and reverse climate change, or will it lead to untold harms... weirdly assumes that those two outcomes are somehow mutually exclusive. But as Aza says in the film, both sides are right, and neither side goes far enough.
2. The real problem is the incentive structures, which create a race dynamic that rewards speed over safety. I wrote about this myself from a different angle. As someone who worked at OpenAI, I went into the film feeling like the frontier labs were already on top of it... and came out realizing we need better regulation:
3. The solution is psychological, and starts with how the film itself (as Jonathan Haidt pointed out) gives people a common language. Shared understanding of the problem → collective action + public pressure → regulation and international coordination.
Yes, all of the above. Changing the incentive structure will be both critical and extremely hard to do. But it's possible and we are doing it now with other technologies. Thanks for sharing your own thoughts to broaden the conversation.
What concerns me most is not just the capability of these systems, but our capacity to meet them with discipline.
Our response time is slow.
Our will is fragmented.
And our impulse? Elite-tier.
We didn’t exactly ace the “handle social media responsibly” test. In fact, we haven’t come close to fixing that (yet).
This next exam is even more consequential.
Our collective impulse—unchecked—has repeatedly shaped outcomes before we fully understood the stakes. That combination deserves far more scrutiny than it’s currently receiving.
We definitely failed the social media test run
I grew up in the 60s and 70s. Personal computers didn't become a thing until I was in my tweens.
I remember experimenting with Eliza & derivative chatbots. Whewn I use AI today, I know there is nothing real there, but I realize people today who grew up with the internet don't recognize the artificiality of it.
I think there needs to be more education regarding the philosophy and ethics of the internet.
I fully agree with need a much deeper discussion in society about the philosophy and ethics of this technology (Jay)
Perfect recap, and thanks again for inviting me to the screening! You landed on some of the same key insights I did:
1. Like our politics, the conversation around AI is polarized into two opposing camps and everyone assumes one side is right and the other is wrong. The question everyone keeps asking: will AI cure cancer and reverse climate change, or will it lead to untold harms... weirdly assumes that those two outcomes are somehow mutually exclusive. But as Aza says in the film, both sides are right, and neither side goes far enough.
2. The real problem is the incentive structures, which create a race dynamic that rewards speed over safety. I wrote about this myself from a different angle. As someone who worked at OpenAI, I went into the film feeling like the frontier labs were already on top of it... and came out realizing we need better regulation:
https://socraticai.substack.com/p/i-worked-at-openai-the-ai-doc-changed
3. The solution is psychological, and starts with how the film itself (as Jonathan Haidt pointed out) gives people a common language. Shared understanding of the problem → collective action + public pressure → regulation and international coordination.
Yes, all of the above. Changing the incentive structure will be both critical and extremely hard to do. But it's possible and we are doing it now with other technologies. Thanks for sharing your own thoughts to broaden the conversation.