“What I've come to understand is that AI alignment isn't just a problem of engineering safeguards. It's a civilizational catastrophe in waiting—a problem of preserving that intersubjective space where human meaning exists. The question becomes: how do we ensure AI operates within that space rather than diffusing it? How do we create cons…
“What I've come to understand is that AI alignment isn't just a problem of engineering safeguards. It's a civilizational catastrophe in waiting—a problem of preserving that intersubjective space where human meaning exists. The question becomes: how do we ensure AI operates within that space rather than diffusing it? How do we create constraints that keep AI systems from autonomously rewriting the rules of meaning itself?”
Excellent newsletter Michael. Bravo!
That said, we’re already experiencing the results of AI implementation from a military perspective. Israel, using sophisticated American AI models, unleashed AI on the Palestinians; to mixed results (depending on whether you were on the receiving end). Now they’re selling their technology to other authoritarian or “authoritarian-light” countries, like India and China.
Therefore, this idea that this is no longer a “civilizational catastrophe in waiting,” it’s already here. Just like Orwell’s 1984 is no longer a possibility or plausible; it just won Trump the election.
Moreover, disinformation campaigns and nefarious AI algorithms have been flooding the zone on social media, where 70% of Americans get their news or form their opinions; so there’s that!!
On a side note, I do have one question for you Michael. I read the “Plot against America” and I couldn’t help but think about Apple’s $500 billion dollar commitment to AI server farms being built in Houston, and other cities around the country. Could Apple and other countries already have a buyer?
Specifically, what DOGE is doing to the US government. Accessing and moving servers to private company servers. Clearly, they are attempting to move all of the US government servers to private company AI servers, for which this isn’t enough AI server capacity in the world to do so currently.
“What I've come to understand is that AI alignment isn't just a problem of engineering safeguards. It's a civilizational catastrophe in waiting—a problem of preserving that intersubjective space where human meaning exists. The question becomes: how do we ensure AI operates within that space rather than diffusing it? How do we create constraints that keep AI systems from autonomously rewriting the rules of meaning itself?”
Excellent newsletter Michael. Bravo!
That said, we’re already experiencing the results of AI implementation from a military perspective. Israel, using sophisticated American AI models, unleashed AI on the Palestinians; to mixed results (depending on whether you were on the receiving end). Now they’re selling their technology to other authoritarian or “authoritarian-light” countries, like India and China.
Therefore, this idea that this is no longer a “civilizational catastrophe in waiting,” it’s already here. Just like Orwell’s 1984 is no longer a possibility or plausible; it just won Trump the election.
Moreover, disinformation campaigns and nefarious AI algorithms have been flooding the zone on social media, where 70% of Americans get their news or form their opinions; so there’s that!!
On a side note, I do have one question for you Michael. I read the “Plot against America” and I couldn’t help but think about Apple’s $500 billion dollar commitment to AI server farms being built in Houston, and other cities around the country. Could Apple and other countries already have a buyer?
Specifically, what DOGE is doing to the US government. Accessing and moving servers to private company servers. Clearly, they are attempting to move all of the US government servers to private company AI servers, for which this isn’t enough AI server capacity in the world to do so currently.
Any thoughts?