Artificial Intelligence FAQs

Hi, welcome to CentaurAI. Mell and I started this business because we care. May sound trite but its true. I’ve worked in too many small to medium size New Zealand businesses where working harder rather than smarter was the rule.

Spreadsheets, emails, waiting for others, and meetings - running fast just to stay still. Seeing the personal toll on people’s health and wellbeing is what upset me the most. When I started seeing what I could achieve with AI I wanted to enable others to feel the same. I started my master’s in early 2024 to study AI Adoption in NZ SME’s. Can we harness this technology, or will we get left behind?

When explaining to others what using Gen-AI is like for the first time we often refer to what it would be like being asked to drive a Formula One race car. Without any advice or practice at best you’d be stalling or spinning your wheels in the carpark. Many people we talk to have ‘tried AI’ had a bad experience and never returned. We encourage you to lean in and get past that initial disappointment. In no time you’ll be flying…

Here's a few frequently asked questions to help fill in some gaps. Feel free to email me if you are curious about how AI might fit with you and/or your business.

Getting Started

The AI journey can feel overwhelming, but it doesn't have to be. We believe in starting small with practical applications that deliver immediate value. Rather than trying to revolutionise your entire workflow overnight, begin with simple tasks you already do daily like enhancing email replies, accelerating research, or creating content drafts. With the right guidance and a structured approach, you'll discover AI's power to amplify your existing expertise whilst building confidence through quick wins.

Like learning to drive a car we recommend starting with a single model like Chat GPT or Claude. You can register a free account. All you need is an email address and a password. Then start with a question, any question and prepare to dive into a conversation…

Think of AI as an incredibly well-read assistant that has studied millions of documents and conversations, then learned to predict helpful responses. It's sophisticated pattern recognition at massive scale, not magic. AI models are trained on vast amounts of text, learning how professionals communicate and solve problems. When you ask a question, AI predicts the most useful response based on these learned patterns. Understanding this helps to understand why providing adequate context to your enquiries is important to minimise errors and to set realistic expectations for what AI can achieve.

This is a legitimate concern that deserves a thoughtful response based on facts, not fear. Different AI tools handle data differently and understanding these differences helps you use AI safely whilst protecting sensitive information. The key is knowing what's appropriate to share and what isn't. With proper guidelines, like anonymising client details and using general scenarios instead of specific situations you can harness AI's benefits whilst maintaining professional standards and protecting confidential information through smart data practices. We recommend not uploading any company information you wouldn’t be comfortable putting online. See What should I never use AI for? below under practical usage.

This frustration is incredibly common and usually fixable with better prompting techniques rather than being an AI limitation. Poor results typically stem from insufficient context, vague requests, or unrealistic expectations. AI doesn't know your industry, role, or specific situation unless you provide that context. The secret is treating AI like a capable colleague who needs proper briefing. With clear context, specific requirements, and realistic expectations, you'll see dramatically improved results that make sense.

The AI world is full of confusing terms, but you only need to understand the basics to use these tools effectively. "Generative AI" refers to AI that creates new content—text, images, or code—rather than just analysing existing information. LLM (Large Language Model) is the "brain" behind tools like ChatGPT and Claude—these are AI systems trained on vast amounts of text that can understand and respond to natural language, meaning you can communicate with them using everyday conversation rather than computer code. A "prompt" is simply your instructions to AI. "ChatGPT" is OpenAI's popular assistant, whilst "Claude" excels at business writing and analysis. "Perplexity" combines search with AI analysis. Focus on practical use rather than technical terminology—you don't need to understand engine mechanics to drive effectively.

Practical Usage

The best AI tool depends on your specific requirements, but most business professionals should start with ChatGPT or Claude, both handle common business tasks well. ChatGPT offers broader recognition and extensive online help, whilst Claude excels at detailed analysis and professional writing. Microsoft Copilot integrates with Office 365 applications, though many users find standalone tools more capable for complex tasks. For research with current information, try Perplexity. Rather than overthinking the choice, start with one tool's free version, develop your prompting skills, then invest in the paid version of whichever becomes essential to your workflow. Skills transfer between tools.

Verification is essential for professional use. AI can be very confidently wrong, so never use outputs for important work without checking sources. Use a three-tier approach: basic sense-checking for every response, fact verification for any claims, and expert validation for critical decisions. Red flags requiring extra verification include specific statistics, legal requirements, financial calculations, and industry- specific details. Think of AI as providing excellent first drafts that require your professional expertise to finalise and validate. Most tools respond with links to citations when asked.

Absolutely. Success depends on how you integrate AI into your professional workflow, not whether you use it. Clients care about quality deliverables, reliability, and value, not your research methods or productivity tools. The professional approach positions AI as a capability enhancer that allows more thorough research and faster turnaround whilst maintaining high standards. Always review and refine AI outputs, verify facts independently, and take full responsibility for final deliverables. Focus on the enhanced value you can deliver with AI assistance.

Real time savings vary by application, but most professionals see meaningful improvements once they move beyond basic experimentation. Expect email and communication tasks to reduce by 20-50%, research and analysis by 40-70%, and content creation by 30-60%. Initial learning phases show minimal savings, but by month 2-3, many users report 5-8 hours weekly savings. The compound effect grows as you integrate AI into more workflows. Remember, benefits extend beyond speed to include quality improvements, reduced mental fatigue, and increased capacity for creative and strategic thinking.

Understanding AI limitations is crucial for professional use. Never use AI for final decision-making without human judgement, unverified factual claims, sensitive personal information, professional advice outside your expertise, or critical safety assessments. Don't share client details, proprietary information, or confidential data. AI cannot replace human relationships, professional liability, or real-time information needs. Always maintain human oversight for client-facing work and high-stakes communication. Think of AI as augmenting human intelligence, not replacing human expertise and responsibility.

Workplace Integration

This concern is understandable but often overblown. Most forward-thinking managers view AI as a professional tool, like Excel or Google. Progressive workplaces see AI skills as valuable professional development. Frame AI as enhancing your capabilities, not replacing them. Show how it allows you to deliver higher quality work faster whilst focusing on strategic tasks. Be transparent about your process, emphasising that your expertise guides and refines AI output. Position AI use as professional development and demonstrate enhanced value to your organisation.

It is very unlikely that AI is not being used in the business already. Often referred to as ‘Shadow AI’ – the covert use of AI tools at work is revealing in many AI surveys. Incorporating AI into the business should be open and needs support from leadership to legitimise its use, access to tools, and some education on what using AI responsibly requires. If you’re just starting out, we are open to have chat to help you understand and plan your AI journey.

AI will make mistakes, just like humans do, so having robust prevention and response strategies is essential. Prevent errors through three-source verification, peer review, and expert consultation. Never send AI-generated content without human review. When mistakes occur, acknowledge quickly, correct immediately, and implement process improvements. Build error-resilient systems through documentation protocols, clear escalation procedures, and regular training. Remember, you remain fully responsible for all work output regardless of AI assistance. Professional standards apply equally to AI-assisted work. Better yet establish an AI Policy for your business that clearly outlines to the do’s and don’ts and the what if scenarios so everyone’s clear.

Yes, absolutely. An AI policy both protects your business and empowers employees to confidently use AI tools effectively. Clear guidelines provide the safety framework that enables innovation rather than restricting it. Your policy should include approved tools, encouraged uses, data protection requirements, quality control standards, and professional responsibility frameworks. Focus on enabling safe experimentation whilst protecting sensitive information. A well-crafted policy provides both protection and empowerment, creating the framework for your team to leverage AI's benefits whilst maintaining professional standards.

Effective prompting is a learnable skill that dramatically improves your AI results. Think of AI as a capable colleague who needs context, provide the same information you'd give a smart assistant. Start with basic structure: your role, specific output needed, purpose/audience, and key requirements. Progress to adding detailed specifications, tone requirements, and format preferences. Advanced techniques include role-playing instructions, step-by-step thinking requirements, and constraint stacking. Build your prompt library by saving successful prompts for different situations and iterate based on results. We have a whole module on prompting if you want to dive in.

Industry & Professional Concerns

Data security and privacy with AI tools requires understanding both technical realities and legal obligations. Treat AI tools like public forums until you understand their specific privacy protections. ChatGPT offers privacy controls, paid users can disable data use for model training in settings, whilst free users should assume conversations may be used for improvement. Claude generally doesn't use conversations for training and offers better default privacy protection. Never share client personal information, confidential business data, or security-sensitive information with any AI tool. Use anonymisation techniques, "Company A" instead of actual names, general scenarios rather than specific situations. Different service tiers offer varying privacy protections. Establish clear guidelines, use appropriate tools for different purposes, and create verification processes. Focus on safe practices that enable AI benefits whilst protecting sensitive information.

AI is more likely to change how you do your job than replace you entirely, especially in roles requiring human judgement, creativity, relationships, and complex problem- solving. Research shows AI typically automates tasks, not entire jobs. Most positions consist of multiple tasks, some AI can handle, others requiring human skills. Position yourself for an AI-enhanced future by developing "AI-resistant" skills like complex problem-solving, emotional intelligence, and strategic thinking. Become an "AI- enabled professional" who uses technology to deliver higher-value services whilst focusing human capabilities on uniquely human work.

Honestly, you may struggle to remain competitive in many fields. AI adoption is becoming a professional competency, similar to computer literacy in the 1990s. AI- enabled professionals show 50-70% faster research completion, enhanced quality and comprehensiveness, plus increased capacity for complex work. Competitive gaps are already appearing across industries. However, alternative strategies exist; specialised expertise, superior relationships, and creative innovation. The pragmatic approach suggests understanding AI capabilities helps you respond to market changes, even if you choose not to become a power user initially.

Understanding AI's trajectory helps you prepare strategically rather than just reacting to changes. Expect AI integration with existing business tools, improved accuracy and reliability, multimodal capabilities, and increasingly more capable models. Longer-term developments include specialised industry assistants, AI agents for automation, real-time collaboration features, and advanced personalisation. Prepare by developing fundamental AI literacy, focusing on complementary human skills, building adaptive learning habits, and preparing your organisation for ongoing change. Focus on building adaptability rather than predicting specific scenarios.

Future-proofing your career means becoming an "AI-augmented professional" who combines human expertise with AI capabilities. Develop skills that complement rather than compete with AI: strategic thinking, emotional intelligence, creative problem-solving, and relationship management. Stay curious about AI developments in your industry whilst building expertise in applications that enhance your professional value. Focus on continuous learning, maintain strong professional networks, and position yourself as someone who can bridge AI capabilities with business needs. The future belongs to professionals who learn to work with AI effectively.

If you have any other questions on how AI might work for you feel free to make contact for a no obligation chat. We’d like to see New Zealand working smarter not just harder.

Matt and Mell