So — should you adopt now?
That was the technical case. The last part is the worry people actually have — “what if I champion this now and it turns out badly in two years?”
Published August 10, 2026·Last checked · August 2026
Key points
- The real content of that worry isn't “will AI replace it” but “will the money and effort I spend now be meaningless in two years”
- In all three two-year scenarios, the party that loses is the one that didn't adopt
- In the age of AI the risky asset is not software — it's judgment that exists only in people's heads
- This isn't hypothetical: there is a real case where eight years of rebar data ended up driving conceptual cost estimates
- We also state what we can't do yet — no public API · native drawing production on the roadmap · JIS implemented
1Put the question precisely
The worry gets phrased as “will AI replace BuilderHub,” but that isn't its content. It is “will the money and effort I spend now be meaningless in two years.” So the answer has to address that.
| Scenario in two years | The company that adopted now |
|---|---|
| AI comes to make the rebar judgment calls too | Already holds the in-house rebar data that defines what “correct” means. Without it, you can't tell an AI what the right answer is |
| AI still stops at geometry generation | Already banked two years' worth of automation gains |
| A better rebar automation product appears | Has structured rebar data and in-house standards, so it can switch. If the data lives only in people's heads, there is nothing to switch |
In the age of AI the risky asset is not software — it's judgment that exists only in people's heads.
2This has already happened, not hypothetically
One Korean contractor began accumulating rebar data in 2017. Once eight years had built up, that data went beyond quantities on individual sites and came to feed conceptual cost estimates for the next project. (→ Eight years building a market that didn't exist)
What matters is that the company did not accumulate data with AI in mind. It automated the work in front of it, data was left behind, and that data later produced a different kind of value. Whatever AI ends up doing, what you will need then is data carrying your own company's rebar judgment.
3What we can't do yet (stated as-is)
| Item | Today |
|---|---|
| Public API | Not available yet. There is currently no route for an external program or AI agent to call BuilderHub — planned development |
| Automated rebar shop drawings | For now drawings go through Revit conversion — native drawing production is on the roadmap |
| Rebar codes | JIS (Japan) implemented. KCI, ACI and Eurocode are on the roadmap |
Why state the first line at all — because after arguing in Part 2 that “the API is the boundary,” hiding the fact that we have no API would make the whole argument dishonest. It's better to say plainly that our direction is exactly this: open the API, and AI can call rebar detailing on top of it.
4Closing
AI is replacing the making. It will keep doing so. But what consumes time in rebar detailing is not the making — it's deciding what to make.
Move that judgment into logic and you get two things at once — time saved now, and the data that will let you tell an AI what correct looks like later.
So the answer isn't “wait for AI.” It's “turn judgment into data before AI arrives.”
Sources
This article is a summary compiled from published material. It is not a legal interpretation and does not constitute advice. Regulation changes frequently, so please check the original documents and the latest notices from the responsible authority before relying on it in practice.