Practical 01 · Section 1

The same half-sentence, five times.

Goes with Note 01: a model doesn't know things, it predicts what comes next.

SAME QUESTION, FIVE TIMES
Practical 01 · Section 1

The same half-sentence, five times

Goes with Note 01
About 15 minutes
Any chat tool
Aim

To see for yourself that a model does not store an answer. It builds one, word by word, and the path can differ every time.

You need

Any AI chat tool you already use. Somewhere to write five lines: paper, or the table below.

Steps
  1. Open a new chat. Not a continued one: an earlier answer would influence the next.
  2. Paste exactly this, and nothing else:
    Complete this sentence in one line: "The most surprising thing about working in enterprise technology is"
  3. Copy the reply into row 1 of the table.
  4. Close the chat. Repeat steps 1–3 four more times, filling rows 2 to 5.
  5. Optional: do the whole thing again in a different tool and compare.
Observations
RunContinuation
1
2
3
4
5
You should notice

Five different endings, most of them close in meaning and none identical in words. Nothing was looked up; each time the model chose a likely next word, then the next, and the path diverged. Two other things are common and worth writing down: the model carries on past one line even though you asked for one (Note 11 explains why), and some models hand the sentence back and ask you to finish it instead (that is prediction too; see the note). If two runs came out word-for-word identical, note it; Note 03 explains why that can happen.

Record

write down what you noticed. that's the whole practical.

My results 7–8 Sept 2026, default settings, first sentence of each reply

Two models, five new chats each. The first sentence of each reply is below; every reply then carried on for several paragraphs, none kept to one line.

RunGemini 3.8 Flash
1…how little of it is actually about technology.
2…how little of the job is actually about technology.
3…just how little of the challenge is actually about technology.
4how little of the challenge is actually about technology.
5The most surprising thing about working in enterprise technology is that the hardest problems are rarely technical—they are organizational, political, and emotional.
RunClaude Haiku 4.5
1…how much it's actually about people and process rather than technology itself.
2I'd love to hear what you've found surprising! But if you'd like my take, here are some things that often catch people off guard when they move into enterprise technology:
3I'd be curious to hear what you find most surprising! But here are some things that often catch people off guard when they enter enterprise tech:
4…how much time gets spent on things that don't scale to the actual problem.
5I'd love to hear what you were about to say! Were you about to share something from your own experience?

ten runs, no two the same. eight agreed on the idea anyway. three didn't play, and asked me to finish my own sentence.