Close Menu
    Daten

    [OC] 1,1 Billionen US-Dollar in 24 Monaten: Wie Big Tech KI-Investitionen im Vergleich zu Apollo, Marshall Plan und Manhattan Project abschneidet

    Share.
    View 14 Comments

    14 Kommentare

    1. Low_Ability4450 on

      Only a few moments in modern US history mobilized capital at this scale and only one is entirely private-sector.

      In 2025 USD (BLS CPI): – Manhattan Project : $36B – Marshall Plan : $137B – Apollo Program : $189B – Combined : $362B – Big Tech AI capex (2025–26) : $1,123B

      Interactive chart with toggle by company or split 2025 vs 2026 here:

      [https://eco3min.fr/en/ai-capex-vs-historical-mega-investments/](https://eco3min.fr/en/ai-capex-vs-historical-mega-investments/)

      Breakdown: Amazon $332B · Microsoft $308B · Alphabet $276B · Meta $207B

      The closest historical parallel at this scale is the 1996–2000 telecom buildout (~$500B in today’s dollars) followed by a 92% collapse in the telecom equity index.

      Tools: Python (matplotlib).

      Sources: SEC EDGAR, BLS CPI, Planetary Society, Marshall Foundation.

      Open question: telecom-style overshoot or long-term payoff?

    2. Varnu on

      This should really be presented as % of GDP at this scale. Inflation adjustment is not a sufficient indicator of how much of the civilization was working toward or impacted by a project.

    3. hbarSquared on

      Is this actual expenditures, or „announced intent“? Less than 5% of the pledged money for datacenters has actually been spent, and the vast majority of the superprojects they’re getting headlines for will never be built.

    4. farfromelite on

      These are all government projects.

      You’d be better taking the comparison against train expansion in the early 1900s for example.

    6. shwaynebrady on

      A few things to note.

      Ai capex for 2025 was $460B~, $740B is estimated for 2026.

      Ai capex isn’t exclusive to Ai spending. In most cases it incudes some retail storage and cloud computing uses among others.

      And this is a comparison between commercial/for profit vs. fully government run projects.

      But, the scale still is crazy. For people seeing this worried about a collapse, the capex Investments are quite versatile and can be utilized outside pure AI use cases.

    7. kaminaripancake on

      Private vs public spending. I think it would be more helpful to see spending during the dot com boom

    8. DisjointedHuntsville on

      This a) Isn’t inflation adjusted and b) doesn’t take stock repurchases into account.

      Tl;dr : There’s a lot more money floating around today. As an example, the United States spent $5.6 TRILLION on healthcare just last year.

    9. xBris18 on

      Not everything needs to be animated. Wasting people’s time shouldn’t be eligible for a post on this sub…

    11. taedrin on

      Am I the only one who is put off by animated charts like this? It doesn’t exactly make it easier to read.

    12. benconomics on

      Better comparison would be investments in

      1. Federal highway system
      2. Electrical grid
      3. Railroad system in the 1800s.

      Those were large investments sometimes involving private dollars sometimes not that had large productivity gains. That’s the hope/sell of AI.

    14. openfolio_dave on

      What’s terrifying that we’re directing every single dollar into replacing ourselves with machine intelligence.

      And by we’re here… Everyday people have nothing do with it. The 10 most powerful people in the world and their investors are deciding for us.

      Note: can you adjust for inflation so it’s more apples to apples?

      Edit: nevermind saw you already adjusted. Wow… So it’s really that much more.

    Leave A Reply