This market resolves to YES if Congress passes legislation approving funding specifically for a large-scale, coordinated AI development initiative comparable to the Manhattan Project, as proposed by the U.S.-China Economic and Security Review Commission.
Resolution will be based on official congressional records and major news coverage.
References:
Update 2025-22-01 (PST) (AI summary of creator comment): Funding Range:
Congress must approve $50-100 billion+ specifically allocated for the AI initiative.
Update 2025-01-28 (PST) (AI summary of creator comment): Resolution based on availability of information:
The market will resolve based on what is known at approximately the close time (2025-06-30).
1,000People are also trading
@ian I think that it's very plausible that this will be rolled into a joint private/public effort, where congress will inject funds above and beyond the 500 billion announced here. But for now, this does just look like private capital.
@ian I think it's also of note who's not included here: https://openai.com/index/announcing-the-stargate-project/
Google and Meta won't sit this out, xAi and Anthropic won't sit this out. Everyone wants a piece of this. That suggest that there will be additional projects in a similar style to this one announced in not too long.
@ian A logical strategy for the goverment/military is to let the private sector build the infrastructure such that they can come in at the mid-stage of these projects
@MalachiteEagle Nice, I guess for this to really resolve YES I'd want to see congress approve some funding, just a private venture wouldn't suffice.
@ian would it count if the DoD invested in, for example, OpenAI or Oracle AGI initiatives in the near future? Do you have a threshold for how much you'd want to see invested?
@BungalowBernard for example the DoD requests $100m funding for AI and awards a contract for AGI-like capabilities to OpenAI, with the money being part of the annual omnibus spending bill?
@BungalowBernard hmmm I think that’s reasonable? But I’m not sure, probably $50-100 billion is the range I’m looking for to be used for AI. What do you think?
Claude explained how the manhattan project’s funding worked:
The Manhattan Project's funding structure was remarkably complex and secretive. Here are the key financial aspects:
Total Cost: Approximately $2 billion in 1940s dollars (equivalent to roughly $29 billion today). This made it the second most expensive WWII program, after bomber production.
The funding worked through several layers:
1. Congressional Appropriations
- Money was hidden within the Army Corps of Engineers budget
- Congress allocated funds with minimal oversight due to wartime secrecy
- Even most members of Congress didn't know what they were funding
2. Distribution System
- Funds flowed through the Army Corps of Engineers
- General Groves had extraordinary financial authority
- Used multiple shell companies and contractors to obscure spending
- Utilized existing defense contractors like DuPont and universities
3. Budget Breakdown
- Oak Ridge facilities: ~$1.1 billion
- Hanford Engineer Works: ~$390 million
- Los Alamos Laboratory: ~$74 million
- Project Alberta (weapons delivery): ~$35 million
- Research and development: Remainder
4. Cost Controls
- Despite the blank check nature, Groves insisted on careful accounting
- DuPont notably only charged a $1 fee for their contract work
- Universities often provided resources at or below cost
The project's financial management was revolutionary for its time - it essentially invented new ways of managing massive secret government programs. The funding structure later became a model for other classified defense projects during the Cold War.
Interestingly, much of the detailed financial information wasn't declassified until decades after the war, and some spending records remain classified even today.
@ian just brainstorming, but we run into at least one big issue here. The Manhattan project wasn't made public until we'd nuked someone, so highly classified projects and especially the budget on things like Starshield by the NRO (basically an entire starlink constellation for targeting radars) or development of unmanned combat air vehicles won't be public for a long time. We might not know about budgets for that sort of thing until after the next World War, assuming any of us are still around. Starshield is almost guaranteed to cost $30bn and depends almost entirely on AI to be useful, UCAV development might have crossed that mark a decade ago. The Pentagon *never* passes their audits, that money is going somewhere.
@BungalowBernard dang, good point. I think this market should still resolve based on what we know this year at approximately close time. You’re welcome to make a market about whether it will be revealed in the future, though!
Now might be a good time to remind people of this market https://manifold.markets/AdamK/will-an-american-presidential-candi
@AdamK This one looks undervalued: https://manifold.markets/RemNi/will-donald-trump-be-president-of-t