← 返回 T1ENS1 · The Awakening — AI Essentials
T1ENS1 · 课程资料

S1 Homework Guide (HW1 warm-up + HW2 two reflection questions)

S1 Homework · The Awakening

Course code: ENS1-SPARK · Submit at siliconroshi.ai/en/homework/mineDeadline: Wed July 22, 9:00 PM · US Pacific Time (PT)

Two assignments. HW1 is warm-up (not submitted). HW2 is graded.


Assignment 1 · Practice (warm-up)

Revise your own resume with the expert prompt.

Pull up your current resume. Then run the same content past a chat AI (ChatGPT, Claude, or Gemini) two ways and compare:

  1. The lazy ask: paste your resume with "Help me improve my resume." Read what comes back.
  2. The expert prompt: use the Resume Expert Prompt Template (Laszlo Bock) from Materials — paste the template, paste your resume, ask again.

Read both outputs side by side. The lazy ask gives you polite generalities. The expert prompt gives you specific, brutal, hire-me-or-don't feedback.

✗ Practice only — no need to submit. You're feeling with your own hands what a good prompt does to the same model. The second assignment is the graded one.

If you include a short before/after excerpt in your HW2 submission it counts as extra effort. Leaving it out costs nothing.


Assignment 2 · Submit — Two Reflection Questions

Answer both questions in your own words. Course terminology is never required — plain English counts fully. The AI grader looks for specificity + your voice, not jargon.

① Your AI wish

What do you most want to achieve with AI?

Be specific — a real problem in your own world. Answer three things inside the one wish:

  • What problem — a concrete thing you or someone you know actually struggles with today.
  • Who it's for — the real person served (even if that person is just you).
  • What v1 looks like — one sentence: what a working first version would do.

Strong example:

I spend 2 hours every Friday writing client visit notes — I want an agent that drafts them from my voice memos so it takes 10 minutes. It's for me and my two junior colleagues. V1: I dictate a 5-minute voice memo, agent hands back a structured note in our firm's template.

Weak example:

I want to use AI to improve efficiency and grow my business. (Bare category. No real problem, no real owner, no imaginable v1.)

Also weak:

I want to (a) automate my email, (b) build a chatbot for my website, (c) analyze my customer data, and (d) write blog posts. (Four equal-weight wishes with no priority. One concrete wish beats four vague ones — pick the one that's real for YOU right now.)

② Today's takeaway

What moment or idea from THIS class stuck with you — and what's the one thing you most want to learn next?

Two parts, both required:

  • The moment / idea — a specific thing from this session (a slide, a demo, a sentence, a comparison). Say what stuck AND your reaction to it.
  • What's next — one concrete thing you want to learn — not "everything about AI."

Strong example:

The same model gave a useless resume review on a lazy ask and a brutal, specific one with the expert prompt — same AI, night-and-day output just from what I said. Next I want to learn how to write that kind of prompt for my own field (property management), because generic prompts don't know what a real landlord actually needs.

Weak example:

The class was very inspiring, AI is the future. I want to learn everything about AI. (Could describe any AI class ever taught. No specific moment. "Everything" is not a next step.)


How your submission is graded

The AI grader scores 1-100 across four dimensions. Weights sum to 100.

# Dimension Weight What earns full marks
1 Goal clarity (①) 30 Specific problem + real owner + imaginable first version
2 Takeaway specificity (②) 25 Points to a specific moment from THIS class + one concrete next thing to learn
3 Effort & completeness 20 Both questions answered separately, each a real paragraph (not one-liner). Warm-up excerpt = bonus
4 Authenticity 25 Your genuine voice — colloquial, specific, human. Not AI-generated boilerplate

Score bands:

  • 90-100 — Thoughtful and clearly your own — a wish you could start building next week.
  • 70-89 — Completed seriously, with your own ideas.
  • 50-69 — Completed but thin, or partly generic.
  • 30-49 — Very perfunctory, or mostly AI-generated.
  • 1-29 — Nearly empty or off-topic.

On authenticity

The grader can tell a real human from AI boilerplate. Signs of the real thing: colloquial phrasing, specific personal context, honest emotion, even typos. Signs of AI filler: over-polished sentences, parallel-clause stacks, generic consultant vocabulary ("leverage synergies," "streamline workflows"). Polished writing alone is fine when the content is clearly yours — but if a stranger couldn't tell it was you writing, it costs points.


Submission

  1. Go to siliconroshi.ai/en/homework/mine.
  2. Log in with your enrolled email.
  3. Find ENS1-SPARK in your list.
  4. Paste your two answers. Submit.

The AI grader runs every 10 minutes and rates against the rubric above. You'll see your score, dimension-by-dimension feedback, and a comment on your /homework/mine page — usually within the hour.

You can resubmit — the grader re-rates when your content changes. Only the latest submission counts.


One more thing

The lab from class (build your first Wukong on wukongdojo.ai) is separate from this homework. That's the muscle memory — this is the reflection. If you haven't done the lab yet, see S1 Lab — Baseline Prompt (Copy Me) in Materials. Doing the lab is what makes Question ① answerable with something real.

One concrete wish. One specific moment. Your voice. That's the whole assignment.