Just over a week ago, this blog explored whether AI was going to be impactful at the scale of geopolitics and national power or something else less significant and to what degree (see here).
In that post, we explored some recent commentary and forecasts regarding the expected impact of AI, including one article from the Atlantic Council. That article drew comparisons between 'the most consequential tech race since the dawn of the nuclear age' (the Manhattan Project), and today with AI. While there were similarities between then and now, the article also highlighted some key differences. Go read that post from last week for more.
Today, the US White House has announced The Genesis Mission. Somewhat surprisingly (or not?), it opened with the following Purpose (emphasis mine):
From the founding of our Republic, scientific discovery and technological innovation have driven American progress and prosperity. Today, America is in a race for global technology dominance in the development of artificial intelligence (AI), an important frontier of scientific discovery and economic growth. To that end, my Administration has taken a number of actions to win that race, including issuing multiple Executive Orders and implementing America’s AI Action Plan, which recognizes the need to invest in AI-enabled science to accelerate scientific advancement. In this pivotal moment, the challenges we face require a historic national effort, comparable in urgency and ambition to the Manhattan Project that was instrumental to our victory in World War II and was a critical basis for the foundation of the Department of Energy (DOE) and its national laboratories.
The scope of the Genisis Mission appears large, and is likely to be consequential. It goes on:
The Genesis Mission will build an integrated AI platform to harness Federal scientific datasets — the world’s largest collection of such datasets, developed over decades of Federal investments — to train scientific foundation models and create AI agents to test new hypotheses, automate research workflows, and accelerate scientific breakthroughs. The Genesis Mission will bring together our Nation’s research and development resources — combining the efforts of brilliant American scientists, including those at our national laboratories, with pioneering American businesses; world-renowned universities; and existing research infrastructure, data repositories, production plants, and national security sites — to achieve dramatic acceleration in AI development and utilization.
While it certainly appears this will result in a significant effort, it is clear that the underlying assumption that AI will continue to exponentially improve is unquestioned. Not only that, it overtly states the US is in an AI race. It is a bold initiative that does not seem to doubt that continued exponential improvements in AI will continue like they have in recent years.
However, nobody truly knows if AI will continue to exponentially increase in intelligence. There are many reasons to doubt it will continue to grow like it has in recent years. In the last post we briefly covered some discrepency between what the Atlantic Council was stating, contrasting that piece with some assessments regarding the future geopolitical environment from The Economist and The Lowy Institute that made no mention of AI at all. But it's not just the so-called non-technical communities that can't agree.
The well circulated, hyped and commented on AI 2027 manifesto / site makes detailed and lengthy technical analyses it claims provides strong proof of the imminent and continued exponential growth in AI to 'superintelligence'. US Vice-President JD Vance claims to have read it (paywalled). But then there is the equally detailed and lengthy technical teardown critique of AI 2027 that questions almost every step, approach, assumption and method used in AI 2027. There is also the recent research published by Apple that questioned just how much reasoning LLMs actually do.
So the future regarding AI remains very unclear. But then again, so was what was thought to be possible with nuclear technology before they had those breakthroughs. When the Manhatten Project was established, they never knew for sure what they would be able to achieve. It wasn't certain. Hindsight provides us clarity. But will the Genesis Mision result in something like AGI? Will the US 'beat China'? Or will it be like an alternative universe where the Manhatton Project never actually was able to advance nuclear science because it simply wasn't possible?
With the scale of this announcement, I get the sense that such an enourmous national effort has, from whatever previous probability one may have had in mind, at least slightly increased the chances that current AI (not necessarily LLMs) will progress towards AGI or superintelligence. Interesting times.