This exercise uses AI as a thinking partner, not an assistant.
Instead of writing for you, it helps you see your own reasoning more clearly—what assumptions you're making, where your argument is vulnerable, what alternatives you haven't considered yet.
It's slower than asking AI to just write something. That's intentional. We're not optimizing for speed—we're optimizing for clarity and intellectual honesty.
The goal is to make your reasoning visible, so you can decide if it's sound. You're in charge. The AI just helps you think.
What This Does
We work through your research abstract together in three focused phases over 40-50 minutes:
- Phase 1: Surface hidden assumptions—what connects your evidence to your claims? (10 min)
- Phase 2: Dialectical examination—test your argument against literature, counter-arguments, and alternative framings (20-30 min)
- Phase 3: Refine and document—co-create a stronger version that addresses what we discovered (10 min)
You'll leave with a clearer, more defensible abstract and a deeper understanding of your own reasoning. Not because I told you what to think, but because we thought through it together.
How It Works
Phase 1 · Notice Your Assumptions
We start simple: I ask you focused questions about your abstract. Together we spot the things you were taking for granted but hadn't said out loud yet.
By the end of this phase I hope you'll have a small list of "oh, I guess I was assuming that" moments.
~10 minutesPhase 2 · Stress‑Test the Argument
Here we kick the tires a bit. I look for papers that support you and papers that might push back, then I try to articulate the kind of objections a reviewer might raise.
We ask: "What evidence would actually change your mind?" and play with a couple of different ways your argument could be framed. It's less about "winning" and more about really understanding the shape of your claim.
~20–30 minutesPhase 3 · Rewrite with Eyes Open
Finally, we rewrite. You stay in control of the voice and substance; I suggest concrete edits that make your assumptions explicit, acknowledge real limitations, and reflect how your thinking has sharpened.
The aim isn't a "perfect" abstract, but one you can stand behind with a clear conscience: you know what you're claiming, why you believe it, and what would make you revise it.
~10 minutesWhat You Get
- ✓Refined abstract—clearer, more defensible, with explicit assumptions and honest qualifications
- ✓Key insights summary—what assumptions you discovered, where your argument was vulnerable, what made it stronger
- ✓Dialectical rigor—you've tested your argument against the strongest objections, not just friendly critique
[For research purposes, we also capture a structured reasoning trace showing the full argumentation hypergraph]
🔒Privacy & Research
Your data is used for research purposes only. No individual data will be shared or published without your explicit consent.
Ready to Start?
Enter your name or initials to begin the exercise
Why It's Rigorous
This isn't generic "AI brainstorming." It's grounded in formal argumentation theory—Toulmin's model for how arguments are structured, Walton's schemes for recognizing reasoning patterns, and the concept of steelmanning from philosophical debate.
I don't teach you the theory. I use it internally to ask better questions, build stronger objections, and help you see the structure of your own reasoning. You experience intellectual rigor, not a theory lesson.
The dialectical approach—testing your argument against its strongest critics—is what philosophers do. We're just making that process explicit and systematic.
Honest Limitations
- •Takes 40-50 minutes. This is intentional—critical thinking can't be rushed. If you want speed, this isn't it.
- •Works best with research abstracts that make explicit claims backed by evidence. Purely speculative or definitional work might not fit.
- •The literature search is helpful but not comprehensive. It supplements your own review, it doesn't replace it.
- •The steelman objection is only as strong as what I can construct from available literature. Real peer review will surface more.
- •The reasoning trace is a representation of your argument structure, not objective truth. Use it as a tool for thinking, not as proof that you're right.