ChatGPT gets it wrong.
cplOS gets it right.
Generic AI mixes FAA, EASA, and DGCA standards. For Indian pilot exams, that means wrong answers.
Find G/S and Drift
Trk
090°
TAS
160 kt
W/V
330/18
Using the wind triangle method:
- Headwind/Tailwind = 18 × cos(240°) = -9 kts
- Crosswind = 18 × sin(240°) = -15.6 kts
Groundspeed ≈ 160 + (-9) = 151 kts
Drift ≈ arctan(15.6/160) ≈ 6° to the right
GS = 151 kts
Wrong method, wrong answer
FLT → Wind Correction → TCrs: 090°, TAS: 160, WDir: 330°, WSpd: 18
Groundspeed
168
kts
Drift
06°
Starboard
Exam Trap
Negative WCA = Starboard. ChatGPT says "right" — the exam says "Starboard."
3 reasons
Why generic AI fails
for DGCA exams
It mixes international standards
ChatGPT doesn't distinguish DGCA from FAA or EASA. It uses a simplified headwind/crosswind method instead of the CX-3 — and gets 151 kts instead of 168.
No CX-3 flight computer support
DGCA Navigation exams expect you to use the CX-3. ChatGPT doesn't know the button sequence and uses textbook formulas that don't match the exam method.
No exam trap awareness
DGCA exams have specific traps — like interpreting WCA sign as Starboard vs Port. ChatGPT says "6° right" instead of "6° Starboard." That's the difference between the correct option and a distractor.