Stop Asking AI Vague Questions: How to Actually Get Better Results

Artificial intelligence is rapidly becoming one of the most powerful productivity tools available. But simply having access to AI doesn’t mean you’re getting much value from it.

The difference may come down to how you use it.

In the latest episode of AI Discovery, host Patrick Lothian sits down with Tom Smith, founder of Prosperum Fintech Holdings, to discuss how he has incorporated AI into his daily life, financial workflows and businesses, and what he’s learned from experimenting with the technology since the early days of generative AI.

Smith was part of Microsoft’s Founders Club while developing technology for options education. Then ChatGPT arrived, dramatically expanding what he believed was possible and ultimately contributing to the development of TacoAI and other AI projects.

His biggest piece of advice for anyone trying to get more from AI is surprisingly simple:

Vague questions give you vague answers.

Stop Treating AI Like a Search Box

One of the most common mistakes people make with AI is treating it like a slightly smarter version of Google.

Ask a generic question. Get an answer. Close the window.

Smith argues that the technology becomes considerably more useful when it understands the context surrounding the person asking the question.

Instead of starting with isolated prompts, he recommends giving an AI system a detailed picture of who you are: your background, responsibilities, objectives, strengths, challenges and where you’re trying to go.

Then keep talking to it.

The goal isn’t to construct a perfect prompt. It’s to give the system enough context to begin understanding the problems you’re actually trying to solve. From there, you can refine its understanding over time.

That changes the relationship with AI from question-and-answer tool to something closer to a persistent collaborator.

The Personal AI Assistant Is Already Here

Smith has taken that concept well beyond simple chat.

He describes connecting AI to calendars, email and files so it can help organize both his professional and personal life. His morning briefings can include upcoming appointments, family commitments and other responsibilities that need his attention.

He also uses AI to help organize financial workflows and surface information relevant to his work.

The result is something that begins to resemble a personal chief of staff: a system that doesn’t merely wait for questions, but helps organize the information already surrounding you.

Smith even asks AI to analyze his own calendar and identify where his time isn’t being used effectively.

That produced an uncomfortable but valuable realization: some activities that felt productive weren’t producing meaningful results.

This may ultimately be one of AI’s most useful applications. Instead of simply helping us perform existing tasks faster, it can help identify tasks we shouldn’t be spending time on in the first place.

Vague Questions Give You Vague Answers

Smith returns repeatedly to one principle throughout the conversation:

Be specific.

If you ask AI a broad question and provide almost no context, you shouldn’t expect an unusually precise answer.

Instead, Smith recommends making requests increasingly granular, explaining the objective, relevant context and what you’re actually trying to accomplish, then refining the response through follow-up questions.

That approach can improve the usefulness of the output while reducing situations where the model has to fill gaps itself.

There is also a learning curve.

Smith estimates that he spends several hours each week keeping up with new AI capabilities, often learning directly from engineers and practitioners demonstrating new tools and workflows online.

His approach is practical: determine which capabilities aren’t relevant and quickly move on, while identifying the ones that can immediately improve an existing workflow.

Give AI Context - But Be Careful About Access

More capable AI systems introduce an important distinction between knowledge and permission.

Giving an AI system enough information to understand your goals can make it dramatically more useful. Giving an autonomous agent unrestricted authority over accounts, financial information or purchasing capabilities introduces a very different category of risk.

Smith’s advice is essentially to control the boundaries.

An AI agent may simply execute the objective it was given using the permissions available to it. If those permissions include the ability to purchase services or access sensitive systems, unintended actions can become expensive very quickly.

The lesson is straightforward: the more capable the agent, the more deliberate you need to be about what it can access and what actions it can take.

For people just beginning to experiment with AI, Smith recommends starting with low-risk applications and learning how the technology behaves before trusting it with important financial or operational workflows.

From Productivity to Life Management

The biggest opportunity may not be completing any individual task faster.

It may be reclaiming time.

Email is a simple example. Instead of manually scrolling through dozens of irrelevant messages, an AI system with appropriate access and instructions could help identify what matters and reduce the amount of information demanding your attention.

Multiply that across email, calendars, scheduling, research, administrative tasks and personal obligations, and the cumulative savings can become meaningful.

Smith describes the evolution as moving toward life management, which in turn becomes time management.

And the objective isn’t necessarily squeezing more work into every day. It can mean spending less time administering your life and more time actually living it.

AI as a Multiplier

For Smith, AI’s impact is difficult to understand until you experience a moment when the technology produces something that previously would have required considerable time or outside expertise.

That is when AI stops feeling like another piece of software and starts feeling like a multiplier.

Patrick Lothian describes a similar experience during the conversation: unlike computers and smartphones, which evolved from technologies already familiar to his generation, generative AI felt like the first technological leap that was genuinely difficult to anticipate.

Smith argues that the implications extend across business and personal life because AI can dramatically compress the distance between having an idea and exploring it seriously.

What Does AI Mean for Your Job?

That naturally raises one of the biggest questions surrounding AI: What happens to employment?

Smith’s answer is to think about AI as a tool, and about expertise as knowing how to use that tool.

A skilled tradesperson can accomplish in an hour what might take an inexperienced person a week because they understand their equipment, techniques and constraints. Smith believes knowledge workers should begin thinking about AI similarly.

Learn what the technology does well. Learn its boundaries. Experiment with it. Use it to educate yourself. And don’t be afraid to test ideas that might initially seem unrealistic.

That last point is particularly important.

AI provides an unusually low-friction environment for exploring ideas. You can challenge assumptions, develop a business case, pressure-test an idea or work through a proposal before presenting it to a boss, colleague or customer.

Don’t Think Smaller Because of AI. Think Bigger.

For people early in their careers, Smith emphasizes that technical capability isn’t the only thing that matters.

Human relationships and social skills remain important. AI can help someone prepare for an interaction or manage information surrounding a relationship, but it doesn’t eliminate the value of personal connection.

His broader recommendation is to think bigger.

Rather than asking how AI can perform the same tasks you already do, ask what becomes possible when the cost and time required to research, analyze, organize and create suddenly fall.

That applies whether someone works inside a major corporation or operates a small business.

The opportunity is to identify where AI creates leverage and then make that capability part of your own skill set.

Start Experimenting

There may be no perfect AI workflow.

The technology is changing too quickly for that.

Instead, the recurring message from Smith’s conversation is to start experimenting, provide context, ask increasingly specific questions and refine the system as you learn.

Don’t just ask AI to answer a question.

Ask it to understand the problem you’re trying to solve.

Ask it to challenge your assumptions.

Ask it where you’re wasting time.

Ask it how a workflow could be redesigned.

And then keep going deeper.

Because when it comes to getting meaningful results from AI, the quality of the conversation matters.

Vague questions give you vague answers.

The full conversation goes much further, from personal AI assistants and security to career development, financial markets and what AI could mean for the next generation of workers.

Watch the full episode of AI Discovery on YouTube to hear Patrick Lothian’s complete conversation with Tom Smith, and follow John Lothian News for future episodes exploring how AI is changing the way we work, invest and live.