Why Asking the Same Question Limits Solutions

Why do almost all AI models keep giving similar answers on creative tasks or writing? I have been noticing this for a while. Whenever I ask for something creative or exploratory, Claude, GPT-5 and Gemini all converge on a similar answer template. The wording shifts a little, but the underlying idea feels identical. I just finished reading Artificial Hivemind, a new paper out of UW and the Allen Institute, and it confirms this is not just a vibe. To prove this, the authors collected 26,000 real world open-ended queries, ran them across 70+ models, and used the results to study how reward models behave when there is no single correct answer. The usual expectation would be that the responses scatter across embedding space. Instead, almost every model collapses into the same tight cluster. Once we see how little variation there is in that answer space, it becomes obvious why they all sound alike even when the prompts are creative. The examples in the paper make this very noticeable. Ask dozens of models for a metaphor about time and nearly every single one, regardless of architecture or lab, converges on “Time is a river” or “Time is a Weaver”. This has real implications for the current interest in model swarms. The idea behind multi model swarms where multiple LLMs debate, self criticise and jointly converge on a better answer. That only works if the models bring genuinely different perspectives. If every model in the swarm is pre-conditioned to seek the exact same local minima of “safety” and “helpfulness”, then it’s just a single model with higher latency. The bottleneck seems to be the Reward Models. These are supposed to help models align with human judgement, but in practice they are actually penalising idiosyncratic creativity. They flatten the natural diversity in human preference and push models toward one homogenized style of answer. If billions of people start using these tools for ideation and every tool keeps suggesting the same safe and statistically favoured ideas, we risk shrinking the collective search space of thought. My takeaways from the paper: 1. We should treat disagreement as signal rather than noise. We are currently training models to collapse distributions into a single best answer. 2. We should be training them to represent the full distribution of human preference. 3. Diversity needs to be an explicit evaluation metric. If an alignment method increases average quality but reduces the variety of answers, that is a regression. 4. For agents that help with research, design and ML work, we should build systems with intentional heterogeneity. This paper is worth reading carefully for anyone working with language models, especially the sections on inter model similarity and reward model calibration. This paper makes a strong point to show that the immediate risk of AI is a future where every model is intelligent in exactly the same way.

More data needs better questions. In 1854, London was drowning in cholera data: death counts by parish, ward, and street, all meticulously recorded. But the key to ending the outbreak was asking the right question: Physician John Snow wondered not just how many were dying, but *where exactly they lived*. Plotted on a map, the data pointed straight at a single contaminated water pump. Many organizations today face a similar challenge. They have oceans of data but a desperate shortage of questions worth asking. Every system logs, every interaction is tracked, and every dashboard refreshes in real time, yet the questions being asked of all this data remain the same ones from years ago (just answered faster with more colors). This is the paradox of data abundance: The more data accumulates, the lazier the questions tend to become. Teams ask what is easy to measure rather than what is worth knowing, and "let me check the dashboard" quietly replaces "let me think about what we actually need to understand." Of course, better tools and richer datasets *do* unlock better answers, but only when someone has done the hard, slow work of framing the inquiry. A vague question will produce a precise but useless number; an old question an outdated insight. The question is the strategic act. The data is merely the response. The next time you look at your metrics, ask one thing: How many of these numbers would actually challenge our beliefs if they moved? If the answer is "not many," it's time to start looking for better questions.

I used to think having all the answers made me a good leader. A newby to managing and leading, when someone on my team came to me with a question—Where can I find…? What should I do about…? I jumped in with the fix. Fast. Efficient. Helpful. But over those early years, I noticed something. The same people kept asking the same kinds of questions. I wasn’t building capability. I was creating dependence. Time to try something different. Instead of answering, I started asking: “What have you tried so far?” “What do you think your best options are?” “Which one feels right to you—and why?” It felt slower at first. A bit clumsy. But something shifted. People started thinking more for themselves. 👉 Confidence went up. 👉 Solutions got better. And I didn’t have to carry every decision on my own shoulders. Leadership isn’t about having all the answers. It’s about helping your people find their answers. So if you’re a new manager or a seasoned leader getting peppered with questions—pause. - Ask instead of answer. - Coach instead of rescue. - It’s not just a leadership move. - It’s a development strategy. What’s one powerful question you’ve started asking instead of answering? #LeadershipDevelopment #AskBetterQuestions #WinningClarity

You answered the question in the all-hands. You covered it in the leadership team meeting. You put it in the deck. And this week, they're asking again. "What's the priority?" "Are we still doing this?" "Did that decision change?" Your first instinct: they're not listening. Or you're not communicating clearly enough. 👉🏾 But here's what's actually happening: If your team keeps asking for clarity you've already given, the issue isn't communication, it's trust in steadiness. Repetition is often a symptom of leaked uncertainty. They're not asking because they forgot. They're asking because they've learned that what you said last week might not hold this week. Research shows that frequent shifts in leaders' priorities are strongly associated with burnout and disengagement. When leaders change direction without clear rationale, 💥 teams protect themselves, 💥 they do the minimum, 💥 wait out the next shift, 💥 discount current directives as temporary. Studies analyzing over 80,000 360-degree reviews found that leaders who act inconsistently erode trust roughly three times faster than those who behave predictably. So when your team keeps asking the same question, → They're not being dense. → They're being rational. → They're checking whether this version of the answer is the one that will actually hold.  🚩Here's where uncertainty leaks: → Tone and body language. You say "we're committed," but your tone is tight, your body language closed. They feel the hedge. → Decision reversals without narrative. You change course without explaining what changed or what you learned. It registers as unreliability. → Hedging language. "Maybe," "sort of," "we'll see." Softening feels collaborative but overuse reduces perceived authority. → Say/reward gap. You say innovation matters but penalize failed experiments. They watch what you reinforce, not what you announce. People track coherence across your words, tone, and actions at a fine-grained level. When those don't align, they adjust. Not because they're bad listeners. Because you've trained them that clarity isn't stable. ✅ Here's the shift: 1️⃣ When you give clarity, ask yourself, "Do I actually believe this will hold?" If not, don't give false certainty. Say: "Here's where we are. Here's what's still forming." 2️⃣ When you change course, name it explicitly: "Last week I said X. Here's what changed. Now we're doing Y." 3️⃣ Check your tone. If your words say "we're locked in" but your body says "I'm not sure," they'll trust the body. 👉🏾 Steadiness isn't rigidity. It's coherence, your words, tone, and decisions pointing in the same direction. If your team keeps asking for clarity you've already given, look for where your uncertainty is leaking. Because repetition isn't about them not hearing you. It's about them not trusting what they heard will still be true next week. Check out the REWIRED. newsletter (link in comments).

"I'm sorry, my answers are limited. You must ask the right questions." That line from I, Robot keeps playing in my head. Will Smith is interrogating a hologram. It can answer anything — but only if he asks the right question. That's exactly what's happening with AI right now. People are frustrated. "AI doesn't work." "It gives generic answers." "It's not that helpful." But the problem isn't the AI. It's the questions. Look at what most people ask: → "How do I close this deal?" → "How do I handle this objection?" → "Write me an email." Weak questions. Weak answers. No context. No specificity. No depth. AI can only work with what you give it. Give it nothing? You get nothing back. But watch what happens when you ask better questions: → "Here's the discovery call transcript. The prospect said X, Y, and Z. What objections are likely coming and how should I preempt them?" → "Based on this email thread, what's the real reason they've gone dark? Give me 3 hypotheses and a re-engagement strategy for each." → "Review this deal. What's missing from my close plan? What would you challenge?" Same AI. Completely different results. Here's the real problem though: Most people aren't asking ANY questions. They're waiting. Hoping AI will just... do something. But AI is still reactive. It waits for you to come to it. And most people don't come. Or they come with surface-level asks. That's the gap. The people winning with AI right now? They're asking better questions. They're asking more questions. They're asking proactively — not waiting until they're stuck. This is exactly why I'm building proactive agents for my orgs. Agents that don't wait for questions. They inspect. They prompt. They coach. They reach out. Proactive AI > Reactive AI. Because "you must ask the right questions" is a limitation we can engineer around. The future isn't AI that waits for you. It's AI that comes to you — with the right questions already asked. Are you asking AI the right questions? Or are you waiting for it to read your mind?

Founders often treat repeated questions as a communication problem. So they respond with: More documentation. Better onboarding guides. Longer FAQs. But repetition usually signals something deeper. Confusion. If many users ask the same question, the product isn’t communicating clearly on its own. The strongest products require very little explanation. They make the right action obvious. When a product is intuitive: • Fewer instructions are needed • Fewer clarifications appear • Fewer support tickets arrive Because the interface answers the question. Support requests are rarely just support issues. They are product signals. Each repeated question points to friction. Don’t only document the confusion. Remove the reason it exists. #startups #entrepreneurship #venturecapital

You're solving the same problem for the third time this month. That's not helping your team. That's helping your ego. Inside your head? "I'm just answering questions." Here's what's really happening: You solve the problem. Feel helpful. Two months later... someone else asks the same question. You're not being helpful. You're the bottleneck. You're afraid if they don't need your answers, they won't need you. But when you show them how you think: Your team makes decisions without you. Your team feels empowered. You get 10 hours back. Showing them how you think isn't about replacing you. It's about multiplying you. Leaders who give answers build dependent teams. Leaders who teach how to think build teams that scale. When your team comes to you for a decision, don't answer. Ask them: → What outcome are we trying to get? → What's your confidence level on this? → What's the worst case if you're wrong? → Can you reverse this decision if needed? If they can answer these 4 questions, they should decide—not you. If this resonates, share it with your network. And follow Christine Carrillo for more. Want to build a big business with a tiny team? Here's the playbook: https://lnkd.in/gFiAuzTY