AI Guides › ChatGPT & Others
By Nigel Guy · 3 min read
Most advice on this subject assumes you're supposed to pick a favourite. Comparison articles crown a winner, forums argue about which one is "actually better," and the implicit message is that using more than one tool means you haven't made up your mind yet. That framing quietly makes people feel indecisive for doing something that's often just sensible.
The rule: running two AI tools for genuinely different jobs is a deliberate choice with a real rationale behind it, not a failure to commit to one — the question worth asking isn't "which tool wins" but "which job is each one actually for."
The case holds up when at least one of these is true for your actual work, not hypothetically:
| Signal | Running two tools well | Running two tools badly |
|---|---|---|
| Task routing | Each tool has a clear, repeatable job | You open whichever tab loads first |
| Cost | Both subscriptions get regular real use | One is basically unused, kept "just in case" |
| Context | You accept rebuilding context each time you switch | You expect one tool to somehow remember what you told the other |
| Decision cadence | Reviewed every few months against actual use | Never revisited since the day you signed up |
Skip keeping a second subscription "just in case" if you can't name the last time you actually opened it for something real. Skip trying to run every task through both tools as a rule — that's the two-model sanity check for consequential work, not a blanket habit, and applying it to everything turns a useful practice into a subscription-funded chore. And skip using "I run two tools" as a reason to never properly evaluate either one against your actual needs — running two is a strategy, not an excuse to avoid the comparison.