Phil's AI Series #56: New Chat vs. Existing Chat — Pros and Cons of Context Reset

“Why does GPT sometimes give completely off-track answers?”

The issue isn’t the AI — it’s the conversation context.

https://www.youtube.com/watch?v=D_oDiZyRiTc

GPT carries forward the entire context within a single conversation. That can lead to context conflicts:

• It clings to earlier parts of the conversation

• Even if you change topics, it keeps the old assumptions

• It tries to connect completely unrelated requests

Solution 1 — Start a new chat

(This fixes most context-related errors)

• Previous history → completely cleared

• GPT’s frame → resets to zero

• Your current request → gets full focus

But there’s a trade-off

• Your style → disappears

• Project continuity → breaks

• Previous work → can’t be referenced

When should you start a new chat?

Start a new chat when:

• The topic changes completely

• You’re starting a totally new task

• GPT is being stubborn or “stuck”

Stay in the same chat when:

• You’re working on reports, strategy, or a series

• You want to maintain tone/voice

• It’s a long-term project

Key insight

People who separate these two modes use GPT effectively:

• New chat = accuracy

• Same chat = continuity