No, not my advice: that of a professor of science who also teaches scientific writing.
I suspect that his caveats on the use of ChatGPT in scientific papers apply equally to proposals. In case you don’t have the time or interest to read the whole thing, here are the highlights:
- ChatGPT requires lots of human input–prompts, guidance, judgement–to generate even marginally usable writing.
- ChatGPT makes stuff up: not maliciously, but precisely because it’s programmed to fill gaps with plausible (or, perhaps, “right sounding”) bits. In science papers that’s often fake citations: proposal teams can look forward to fake experience examples or performance results.
- ChatGPT delivers pedestrian science (in his example) precisely because it’s aping existing text and repeating existing hypotheses. For the same reason, you can expect it to deliver pedestrian solutions for a client’s requirement. Just as there’s no new science being done here, there’s no business/operational creativity on offer here either.
- ChatGPT replicates all the bad habits of science writers, including the passive voice and tortuous sentences and paragraphs. It would do the same for proposal writers, presumably. Oh, hurray.
So, how could you use ChatGPT in Proposal Land? Read this.