WEBVTT

NOTE AI Engineering Visualised: https://understandingdata.com/ai-engineering-visualised/anatomy-of-a-prompt/

1
00:00:00.260 --> 00:00:09.823
Ask a model to summarise this incident and nothing else, and the recorded answer is a polite question back: a vague request gives it nothing to work with.

2
00:00:09.823 --> 00:00:19.928
Add the log as context and it writes 96 words of bullet points, accurate but written for nobody in particular, in a shape no program can read.

3
00:00:19.928 --> 00:00:29.517
Name the audience, the on-call engineers, and it grows to 101 words and adds advice about concurrent indexes that the log never mentions.

4
00:00:29.517 --> 00:00:38.015
Then a constraint, at most 60 words and only what the log shows, cuts it to 53 words and the guess disappears.

5
00:00:38.015 --> 00:00:49.677
The surprise is the example: one sample summary turns the reply into JSON on its own, a bigger change than any other plate, but wrapped in a code fence and missing a key.

6
00:00:49.677 --> 00:01:00.429
That is the failure a bot hits, a reply that looks like JSON but will not parse, until the format plate names all six keys and it parses first time.

7
00:01:00.429 --> 00:01:09.565
So when a program reads the reply, always send the format and one example; pull the audience plate out here and the answer barely changes.