Wall Street PromptBrand
All briefs
Wall Street Prompt · Brand

AI Summary Is Not AI Research

Carousel — cover + one idea per slide + caption. This is a post, not a video script.

Carousel
Instagram carousel + LinkedIn documentInstagram carousel8 slides
References & validationReference missing

No specific creator reference resolved. Add an observed post before production.

Reference creator: dataAlpha.ai / Daloopa

No reference links resolved.

Slides

One idea per slide

Cover
AI SUMMARY IS NOT
AI RESEARCH

Why clean output fools
good analysts.

AI summarizes. You research.
(swipe)
Slide 1
AI SUMMARY IS NOT
AI RESEARCH

Why clean output fools
good analysts.

AI summarizes. You research.
(swipe)
Slide 2
The trap:

A summary feels like research
because it is clean.

Neat bullets. No contradictions.
It reads like a conclusion.

But polish is a writing quality.
It tells you nothing about
whether the numbers are right.
Slide 3
Two different jobs.

SUMMARY
-> Tells you what the document said
-> Compresses, smooths, organizes

RESEARCH
-> Proves what was said is true
-> Ties it to a source
-> Tries to break it

Most AI output stops at summary.
Slide 4
Missing step 1: SOURCE TIE-OUT

-> Every figure traces to a filing
-> A page number and a date attached
-> Re-derive it from the primary document
-> No trace = not usable yet

A number with no source
is a sentence, not evidence.
Slide 5
Missing step 2: VERIFICATION

-> Do not take the model's word for it
-> Open the 10-K, the call, the comp
-> Confirm the figure independently
-> Confirm the caveat survived the summary

Summaries drop the footnote.
The footnote is often the point.
Slide 6
Missing step 3: CHALLENGE

-> Ask: what would make this wrong?
-> Look for the disconfirming line on purpose
-> Force a positive / negative / unclear call
-> Unclear goes on the follow-up list

A summary agrees with itself.
Research argues with itself first.
Slide 7
The failure mode.

You read the clean summary.
It matches what you expected.
You stop.

An unverified figure or a
dropped caveat rides straight
into your thesis.

Clean is how it gets past you.
Slide 8
The disciplined workflow.

-> AI: build the summary, fast
-> You: tie out every figure to source
-> You: verify against the primary document
-> You: challenge it before you trust it

AI summarizes. You research.
Save this. Run it on the next output.

Educational only. Not advice.

Caption

Paste under the carousel

A clean summary feels like research because it is clean. That is the whole trap.

The bullets line up, nothing contradicts itself, and it reads like a finished conclusion. But polish is a writing quality. It says nothing about whether the figures are right or whether anyone checked. Smooth and verified are not the same thing.

A summary tells you what a document said. Research proves it, ties it to a source, and tries to break it. AI is excellent at the first job and does not do the other three on its own.

The three steps a summary skips:

1. Source tie-out: every figure traces to a filing, a page, and a date, re-derived from the primary document. No trace, not usable. 2. Verification: open the actual filing or transcript and confirm the number and the caveat independently. Summaries quietly drop footnotes, and the footnote is often the point. 3. Challenge: ask what would make this wrong, go look for the disconfirming line, and force a positive / negative / unclear call. Unclear goes on the follow-up list.

The failure mode is simple. You read the clean summary, it agrees with what you already believed, you stop, and an unverified number rides into your thesis. Clean is exactly how it gets past you.

Disciplined version: AI builds the summary fast, you do the tie-out, the verification, and the challenge. AI summarizes. You research.

Save this one and run it on your next AI output.

Educational content only. Not investment advice, and not a recommendation to buy, sell, or hold any security. Wall Street Prompt. Always verify against the primary source.

Design notes

WSP system across all 8 slides. Near-black background (#0E0F11), off-white body type, single green accent (#1FC77B). Cover: heavy condensed all-caps headline (Druk / Anton / Archivo Black), AI RESEARCH in accent, the principle line AI summarizes. You research. small at the foot with a (swipe) cue. Body slides: each opens with a short bold label line (the trap, two different jobs, missing step 1, the failure mode, the disciplined workflow), then a tight arrow list. One idea per slide, lecture-slide density, never a single tip. Arrows are plain ASCII -> with hanging indent. Slide 3 uses a clear two-column or stacked SUMMARY vs RESEARCH split with a divider rule. Step slides 4-6 share an identical template so they read as a numbered set. Slide 8 carries the resolve line, the save cue, and the short disclaimer in small type. Consistent left-aligned grid, generous margins, one accent color only, no icons, no stock photos, no emoji, no gradients. Keep every slide legible as a phone thumbnail.

CTA

AI summarizes. You research. Save this and run it on your next output. Comment VERIFY and I will send the tie-out checklist.

Wall Street Prompt
Wall Street Prompt — internal