Partway

Can ChatGPT pick your drop-off station?

Written September 2026

Not reliably, and it's worth being precise about why. An AI assistant is genuinely good at the geography: ask where a drive from Reading to Bristol passes near a station and it will name sensible places. What it cannot see is today's timetable or what any ticket actually costs to buy, so it fills those in from patterns. The answer sounds right, which is different from being right.

Why guessing fails here specifically

Because the answer turns on the two things that move daily. UK rail fares are set route by route and jump wildly between neighbouring stations, and the value of a drop depends on which trains leave after you'd arrive. Partway doesn't guess either number: it looks up live trains and real bookable fares for every station along the drive, then weighs each against simply catching a train from where you started.

The tell: sometimes the honest answer is "barely worth it"

Two searches we ran on 1 September 2026. Lift from Leeds towards Sheffield, passenger for Birmingham: the best drop-off beat staying put by ten pence, and the app said so. Lift from Norwich towards Cambridge, passenger for London: £25.20 from Cambridge against £71.20 from Norwich, a £46.00 difference. Same question, opposite verdicts, and nothing about the map tells you which is which. A system that guesses will hand you a confident story either way; a system that prices things can tell you when not to bother.

Use both, honestly

Ask your assistant whether the lift is worth asking for, and let it word the message to the driver. Then let Partway do the part that needs live data: the station, the train, the ticket, the driver's detour. Ten seconds, free, and the driver gets one link with no app to install.

Got a lift on the table? Skip the guessing: Partway on the App Store. Real trains, real fares, real detours.