GPT5, app, "think hard": "Quick take & predictions

Short baseline (what reporting says so far): U.S. authorities say roughly 475 people were detained in the worksite raid at the Hyundai/LG battery/EV complex in Georgia, and reporting describes the majority as South Korean nationals (local Korean outlets put that number at about ~300). Coverage also notes most detainees were working through subcontractors rather than being Hyundai direct hires, and that officials say the group includes people who lacked proper work permission or had overstayed visas.

Below I’ll treat the South Korean detainee group as ~300 people (that’s the most commonly-reported ballpark) and give a short evidence summary plus a probabilistic forecast for each of your market questions.

---

1) “How many South Koreans detained … were hired after arriving in the US?”

Interpretation: people who arrived in the U.S. (e.g., on tourist/other status) and were then hired locally, as opposed to being recruited and sent by Korean subcontractors.

Evidence: reporting emphasizes subcontractor networks and Korean companies sending crews; I’ve seen no firm reporting that large numbers were newly hired on U.S. soil after arrival.

Forecast (out of ~300):

0–10: 30% chance

11–30: 40% chance (single-point / median-ish ≈ 20)

31–100: 25% chance

> 100: 5% chance

Why: most coverage points toward pre-arranged subcontractor deployments; ad-hoc local hiring of large numbers of South Korean nationals seems less consistent with the described facts.

---

2) “How many … were detained for performing ‘productive work’?”

Interpretation: people actively doing construction/assembly (productive labor) at the site versus administrative/visitor roles.

Evidence: this was a worksite/construction-area raid (battery plant under construction); many of those detained were on-site performing construction or related tasks.

Forecast (out of ~300):

> 200 (≈67%+): 50% chance

150–200: 30% chance (point estimate ≈ 225)

<150: 20% chance

Why: the operation targeted the construction project and arrested people who were on the jobsite; so a large majority of the South Korean detainees probably were doing productive work.

---

3) “How many … overstayed their permission to be in the US (visa or ESTA)?”

Evidence: officials said the detained group included people who did not have permission to work and some who had overstayed; but reporting so far uses vague language (“some,” “a number of,” etc.) rather than giving a hard count.

Forecast (out of ~300):

0–59: 15% chance

60–120: 55% chance (point estimate ≈ 90)

121–200: 25% chance

> 200: 5% chance

Why: a non-trivial minority are being described as lacking proper permissions, but the reporting doesn’t support the claim that a majority were visa overstays.

---

4) “How many … snuck across border (never had permission to enter US)?”

Evidence: some officials said a subset had “crossed the border illegally,” but that language implies a smaller fraction rather than most detainees.

Forecast (out of ~300):

0–5: 40% chance

6–20: 40% chance (point estimate ≈ 12)

21–50: 15% chance

> 50: 5% chance

Why: media and officials have emphasized a mix of statuses; outright illegal border crossings were reported but not as the dominant mode of entry.

---

Overall confidence & caveats: the numbers above are probabilistic guesses built from early reporting (official DHS/ICE comments + major wire coverage). The single clearest, load-bearing fact is the ~475 detained figure and the reporting that the majority were South Korean nationals; details about exact immigration statuses and who was hired where are still fuzzy in public sources. Official counts from DHS/ICE, the Department of Homeland Security investigation, or statements from the South Korean foreign ministry will determine how those markets actually resolve.

"