I've been using the same chat thread for DSA practice, spread across several days now. I open it, review a problem, close it, come back the next day and pick up in the same thread.
What I've noticed: the model behaves as if no time has passed at all. It doesn't distinguish between "this was said 5 minutes ago" and "this was said 3 days ago" inside the same conversation. Everything in the thread reads as flat, current context — unless I manually tell it "it's day 3 now" or "it's been 2 days since we last talked," it has no idea.
This isn't just a DSA-practice quirk. The same gap shows up in a bunch of other single-thread, multi-day use cases:
- Coding projects — a long-running thread where you're building a feature over multiple sessions across a week or two
- Journaling / reflective use — people who use the same thread as an ongoing check-in space
- Fitness / diet logs — tracking meals or workouts in one thread over time
- Budget / expense tracking — logging spend across a month in a single conversation
- Habit or medication tracking — daily check-ins in the same thread
- Long negotiations or planning — back-and-forth on a decision that spans days
- Spaced repetition / study review — my case — where "how long ago did I learn this" actually matters for what to review next
In all of these, the model's inability to sense elapsed time inside a thread means it can't reason about staleness, can't prompt timely follow-ups, and treats week-old and minute-old messages the same way.
Curious if others have hit this. Do you manually re-state the date/time every session? Has anyone noticed ChatGPT/Claude/Gemini handling this differently?
(Not trying to solve it here — just wanted to see if this is a known pattern others have run into, or if I'm missing something obvious.)
[link] [comments]