Where the money goes when AI designs physical things

AI makes drafting a physical product cheap. Checking it is still scarce, making small parts is cheap but priced all over the map, and the last time a hardware cost collapsed, the losses landed at making and delivery. The evidence for each, the routes ranked by where the losses landed, and a check to run on your own idea. Every figure attributed.

By Mackenzie Bowes · mackenzie@mackenziebowes.com · @_mackenziebowes on X

The same aluminium bracket's highest quote was more than 2.5 times its lowest: $65.55 to $166.81.
Contents

The answer

What the evidence supports.

On 4 October 2026, four posts said AI would open physical design to everyone and move the bottleneck to the factory. We took that as given and asked where a person with software skills, little capital and no factory should stand.

Five research passes were each told to find evidence against the premises.

The tables behind each line, with every figure attributed:

Where to stand given where the losses landed

Drafting got cheap and checking did not, so this reading puts one person at the checks with a small first order, and away for now from anything that needs capital. A reasoned guess, not a result; the counts here are illustrative.
RouteEvidenceWhere the risk sits
A small first order you'd use yourself#Boards from $2 for five, sheet metal in as little as 24 hours (vendor prices)[9][10]Low. A failed order costs dollars and teaches what no study has measured
Tools that check designs#Every strong result in section 3 pairs a model with a check; no pass-rate data exists for AI designs run through instant-quote manufacturability checksSmall bet; the gap is named most often and nobody has measured it
Software for small manufacturers#About half of small manufacturers rate their digital adoption manual or basic; budget is the barrier for every firm of 10 or fewer staff (a vendor's survey)[19]No sales-cycle data for small teams in this niche
Trades and technician work#Industrial machinery mechanics $64,520, +18% to 2035; electricians $63,190, +9%; CNC programmers $68,120, +6%[2][3][4]About half the software developer median of $135,980[5]
Brokering between designers and shops#Xometry revenue $229.3M in Q2 2026, up 41%[11]Shapeways failed at it[20]; no data on small brokers
Robot data and physical AI tooling#Foxglove raised $40M; XDOF raised $70M and says the work takes "hundreds of robots"[17][18]Capital and hardware; no case of an individual earning here
Crowdfunding a product before a working unit#24.9% of successful projects delivered on time; founders named manufacturing, shipping and certification[14]Highest. Backers' money and your name ride on a design nobody has checked

The money in factories is going to large players: Hadrian raised $1.37B at a $7.87B valuation in August 2026.[16] The stages that need capital are where the people who entered early lost.

Not yet proven

Six things the research did not find.

What this says about how we work

What was rechecked against the source and what was not.

When a field is loud, we take its claims as given and send the research to find evidence against them.

Then we do the arithmetic nobody posted, rank the routes by where the losses landed, and mark which figures we rechecked against the source: here, the Census August 2026 levels and the HWE-Bench pass rate. That's the reading we'd give your idea before you spend on it.

Run it on your own idea

The entries are from the post: one documented attempt caught the timer chip and the via, and the limits are the ones to read before ordering. The shelf is illustrative.
  1. Has a working unit been made? If not, don't take anyone's money for it yet.
  2. What checks the design? Name the design-rule check, the simulation and the vendor's manufacturability check you'll run before ordering.
  3. What does the smallest order cost, landed? Add tariffs and shipping to the advertised price.
  4. Quote it three times. One bracket varied 2.5x across platforms.
  5. Write down what each check and the first order caught. Nobody has published that number, so yours will be the first.

The plain-language version is the post. If you're weighing a physical product and want a second reading, I'd like to hear about it: mackenzie@mackenziebowes.com or @_mackenziebowes.

Sources

  1. Brynjolfsson, Chandar & Chen (Stanford Digital Economy Lab, 2025). Canaries in the Coal Mine. ADP payroll data. digitaleconomy.stanford.edu/publications/canaries-in-the-coal-mine/
  2. BLS Occupational Outlook Handbook, electricians. www.bls.gov/ooh/construction-and-extraction/electricians.htm
  3. BLS Occupational Outlook Handbook, industrial machinery mechanics, maintenance workers and millwrights. www.bls.gov/ooh/installation-maintenance-and-repair/industrial-machinery-mechanics-and-maintenance-workers-and-millwrights.htm
  4. BLS Occupational Outlook Handbook, metal and plastic machine workers. www.bls.gov/ooh/production/metal-and-plastic-machine-workers.htm
  5. BLS Occupational Outlook Handbook, software developers. www.bls.gov/ooh/computer-and-information-technology/software-developers.htm
  6. HWE-Bench (arXiv 2603.18102, 2026). 300 board-level projects checked with static rules and LTspice simulation. arxiv.org/abs/2603.18102
  7. SchGen (arXiv 2605.30345, 2026). arxiv.org/html/2605.30345v1
  8. IoT-SkillsBench (arXiv 2603.19583, 2026). arxiv.org/html/2603.19583
  9. JLCPCB, PCB assembly service page. Vendor prices. jlcpcb.com/pcb-assembly
  10. SendCutSend, home page. Vendor prices. sendcutsend.com/
  11. Xometry, second-quarter 2026 results (4 Aug 2026). www.globenewswire.com/news-release/2026/08/04/3338170/0/en/xometry-reports-record-second-quarter-2026-results.html
  12. Otron Atlas (1 Mar 2026). Is Xometry expensive? One STEP file quoted on nine platforms. www.otronatlas.com/price-comparison/is-xometry-expensive/
  13. US Census Bureau, Statistics of US Businesses 2022, detailed firm-size tables (NAICS 332710). Receipts are gross shipments. www2.census.gov/programs-surveys/susb/tables/2022/us_state_naics_detailedsizes_2022.txt
  14. Mollick (2014). The dynamics of crowdfunding: an exploratory study. Journal of Business Venturing. leeds-faculty.colorado.edu/bhagat/Dynamics-of-Crowdfunding.pdf
  15. Quirky (company). Wikipedia. en.wikipedia.org/wiki/Quirky_(company)
  16. Tech Startups (6 Aug 2026). Hadrian raises $1.37B. techstartups.com/2026/08/06/defense-tech-startup-hadrian-raises-1-37b-at-nearly-8b-valuation-to-scale-ai-powered-factories/
  17. The Robot Report. Foxglove raises a $40M Series B. www.therobotreport.com/foxglove-raises-40m-scale-data-platform-roboticists/
  18. TechCrunch (17 Jun 2026). Robot training-data collection at XDOF. techcrunch.com/2026/06/17/collecting-robot-training-data-is-dirty-unglamorous-work-some-ai-labs-are-already-paying-xdof-to-do-it/
  19. Lasso Supply Chain Software (Q1 2026). State of digital manufacturing. A vendor survey. lassosupplychain.com/research/q1-2026-state-of-digital-manufacturing-report-small-midsized-companies-in-the-united-states/
  20. CG Channel (2024). Shapeways files for bankruptcy. www.cgchannel.com/2024/07/3d-printing-firm-shapeways-files-for-bankruptcy/