Cross-examination · Intermediate · About 5 min
Defend your case in cross-ex
Handle cross-ex when you are the one answering
Affirmative · Cross-examinationRead the case: AI and the future of work
Resolution Advanced AI will on net improve outcomes for workers
a two-year review of 30 employers records 300 retained workers receiving 4% higher inflation-adjusted pay and working 37 rather than 40 hours per week after AI adoption.
- Sixty entry-level roles were removed and 40 technical roles added.
- Of the 60 displaced workers, 18 entered the new roles, 24 found work elsewhere at 8% lower pay, and 18 were still unemployed six months later.
- The other 22 new roles went to different applicants.
- The review does not isolate AI from every other change or forecast the whole economy.
The affirmative argues that productive tools can improve pay and time for many workers. The negative argues that concentrated losses and barriers to new work can outweigh those gains.
When you are being cross-examined, every question is trying to get you to commit to something they can use later. Listen for what the question assumes and what admission it is fishing for. You can answer straight, qualify, push back on how the question is framed or concede a small point. Each one is a tool.
Answer clearly and protect the case you have to defend. If the question rests on something false, correct it. If it is a fair hit, give a real answer. “Always say yes or no” and “always dodge” are both bad rules. Audiences notice evasion.
Explore an example before trying
Defend the overall claim without accepting a guarantee that nobody loses.
If even one worker loses, do you concede AI is not beneficial on net?
Choose the answer you would give as affirmative. Then compare what it commits you to and how the other person could respond.
That changes “on net” into “no one loses.” We defend an overall benefit, not a guarantee for every worker.
Then how do you justify the losses to displaced workers?
You protect the actual scope of the resolution before accepting the question’s test.
A scope correction does not answer the distribution challenge. Repeating it would leave the strongest objection untouched.
Your next moveCompare the pay and time gains for 300 workers with the depth and duration of the losses. Explain why your balance is persuasive.
Some workers can lose while workers overall benefit. Our case compares both groups.
So you accept that some workers suffer. Why should the gains to others outweigh that?
You acknowledge a limited point without conceding the overall prediction and keep the round moving.
The negative can now emphasize the people bearing the loss. Your case needs an answer beyond saying “overall.”
Your next moveDefend the comparison from your constructive, including the 18 still unemployed and the 24 on lower pay.
No. The question is whether the gains in pay and working time outweigh the losses. A loss matters without automatically settling that comparison.
Are you counting many small benefits as more important than a few severe losses?
You reject the automatic verdict and make the comparison explicit, giving the judge a concrete account of the disagreement.
That comparison remains contestable. The negative can challenge how you weigh severe, concentrated losses against gains spread across more workers.
Your next moveDefend the differences in scale, severity and duration that support your judgment. Explain what evidence would change it.
Each answer commits you to something different. See what it wins, what it leaves open and the likely next question.
Spot the setup
Separate the fact they are asking about from the conclusion they want to attach to it.
Decide what you can give
Concede, qualify or challenge only as much as your case can afford.
Think one question ahead
Answer in a way that still lets you make your argument afterward.
Your response
Affirmative · Cross-examinationWrite your answer, predict the negative’s follow-up, then give your next response. You can accept the cost without accepting their conclusion.
Read the case: The four-day workweek
Resolution A four-day, 32-hour standard work week should replace the current five-day, 40-hour model
Practice round using the developed workweek packet: the negative asks, “If payroll rose by 8% in retail and distribution, your proposal is a failure, correct?” You defend unchanged weekly pay and the case’s time, output and absence benefits.
You can type or use your keyboard’s microphone to dictate. Your text stays editable. Practice is optional.
Loading your progress…
Help me get started
My answer: … Their next question: … My reply: …
Try it aloud
Use the timer for a spoken attempt if it helps. Practice without recording, then note what you would improve.
Speaking is self-reviewed. No microphone or recording is required.
A concession can have more than one meaning
Resolution Advanced AI will on net improve outcomes for workers
a two-year review of 30 employers records 300 retained workers receiving 4% higher inflation-adjusted pay and working 37 rather than 40 hours per week after AI adoption.
- Sixty entry-level roles were removed and 40 technical roles added.
- Of the 60 displaced workers, 18 entered the new roles, 24 found work elsewhere at 8% lower pay, and 18 were still unemployed six months later.
- The other 22 new roles went to different applicants.
- The review does not isolate AI from every other change or forecast the whole economy.
The affirmative argues that productive tools can improve pay and time for many workers. The negative argues that concentrated losses and barriers to new work can outweigh those gains.
- Negative · Question
You accept that 18 displaced workers were unemployed at six months?
This is a fair question about the evidence, not a trap.
- Affirmative · Answer
Yes. That is a serious cost our comparison includes.
Admit the fact. That alone doesn’t decide which effect outweighs the other.
- Negative · Follow-up
Then why should three fewer weekly hours for others outweigh losing work altogether?
The negative pushes on how severe the losses are, not just how many people.
- Affirmative · Defense
Our case weighs sustained pay and time gains for 300 workers against those concentrated losses. We give the broad, recurring gains greater weight, while the transition outcomes are the strongest reason to doubt our forecast.
The affirmative answers with a real comparison the negative can attack, not a promise to find evidence later.
A good answer leaves you with a case you can still defend.
Test the reasoning
Read the practice case
Resolution Advanced AI will on net improve outcomes for workers
a two-year review of 30 employers records 300 retained workers receiving 4% higher inflation-adjusted pay and working 37 rather than 40 hours per week after AI adoption.
- Sixty entry-level roles were removed and 40 technical roles added.
- Of the 60 displaced workers, 18 entered the new roles, 24 found work elsewhere at 8% lower pay, and 18 were still unemployed six months later.
- The other 22 new roles went to different applicants.
- The review does not isolate AI from every other change or forecast the whole economy.
The affirmative argues that productive tools can improve pay and time for many workers. The negative argues that concentrated losses and barriers to new work can outweigh those gains.
Review and continue
Resolution A four-day, 32-hour standard work week should replace the current five-day, 40-hour model
Read the practice task and case
Write your answer, predict the negative’s follow-up, then give your next response. You can accept the cost without accepting their conclusion.
Practice round using the developed workweek packet: the negative asks, “If payroll rose by 8% in retail and distribution, your proposal is a failure, correct?” You defend unchanged weekly pay and the case’s time, output and absence benefits.
Your response
You can return to the practice whenever you want.
One model response
My answer: the payroll increase is a cost, not a complete test of the proposal. We weigh it against eight fewer weekly hours, maintained output and lower absence. Their next question: who pays that extra cost? My reply: the packet does not measure who ultimately pays; that risk weakens the case where margins are tight. I still defend the recurring time benefit, and I would not claim that maintained output makes the cost vanish.
Accepting a cost need not accept the opponent’s verdict.
This is a genuine vulnerability in the distribution of the cost.
The defense states what carries weight despite the risk.
The response separates a measured cost from the negative’s verdict, anticipates a serious next question and defends a comparison. Another strategy could work; its quality depends on the commitments and reasons it leaves in play.
Review the performance
- RefutationI distinguish the fact from the implication sought.
- ArgumentsMy response leaves a case I can defend.
- PersuasionI anticipate and answer the strongest likely follow-up.
Choose one move to improve, revise your response, then try it aloud.
Common mistakes to avoid
- Correcting an assumption the question never made.
- Reframing to dodge a hit your case still has to answer.
Loading your progress…