Eisman is correct that frontier labs are trying to use this moment to enter into partnerships with each other and the government on terms favorable to ownership. The power grab feels similar to those in 2008 and 2001, except this one is premised on preempting a crisis. Where he's wrong is that the crisis is "manufactured." The strategy at the leading labs has essentially been to spend as much money as possible, as quickly as possible, in the hope of reaching superintelligence and recursive self-improvement before they go bankrupt. Concerns that they may run out of money before they reach superintelligence are prompting the leadership at these companies to enter into some form of partnership with the government. I think what we are witnessing and seeing in the last few days through some of these public statements and proposals, like Amodei’s “We Must Pace the Frontier,” is evidence of a negotiation over the terms of that partnership, in which there appear to be four main counterparties: (1) the new monied interests, made up mostly of financial and tech elite who built their fortunes over the last 30 years through multiple expansions of the Internet revolution, including Web 2.0, Crypto, and A.I. (2) the national security state, which is primarily concerned with maintaining American primacy and technological and military superiority over China and the rest of the world. (3) the political class, which has become increasingly corrupted and beholden to the monied interests that hold the purse strings. (4) the citizenry, which has been dumbed down by an adversarial and poorly incentivized media environment and disempowered through a multi-decade erosion in civil liberties and through the maldistribution of national wealth via corporate and elite welfare. Concerns around A.I. risk are warranted, as were concerns about a deflationary depression in 2008 and further acts of terrorism in 2001. AND those concerns are being exploited to crystallize an oligopoly, neo-feudal economy, and security state in which the gains of the last 30 years of technological disruption, free trade, and financial engineering are protected from further disruption. People with modest or even substantial wealth tied up in the stock market are expected to passively consent, even if the prospects for their children, who will live out their lives in a plutocracy and patronage state, are pretty dire, because the alternative has immediate material consequences. The public (we) appear to be largely absent from these negotiations. Again, NONE of this should be grounds for complacency. I think the advancement in capabilities we are seeing is alarming. In many ways, what has come out of the reporting over the last few months confirms many of the thought experiments around the alignment and AI control problems put forward by the AI risk community going back a very long time.