AI Isn't an Oracle - Learning to Be a Smart Consumer of Emerging Technology

As artificial intelligence becomes increasingly embedded in our everyday lives, it's easy to assume these tools have all the answers. But should we trust them as the ultimate source of truth?

In this episode of EmpowerTech, Iyari Cevallos sits down with Dr Kendra Vant, product and technology executive, former AI and data leader at Xero, and researcher with a background spanning MIT, quantum computing, and emerging technologies.

Together, they unpack what generative AI actually is, how tools like ChatGPT, Gemini, Claude and Copilot work, and why it's crucial to think critically about the information these systems provide. The conversation explores the rise of AI-generated search summaries, the limitations of large language models, the hidden risks of treating AI as an authority, and the practical skills people need to navigate an AI-powered future.

Rather than focusing on hype, Kendra offers a thoughtful and balanced perspective on how individuals can safely experiment with AI while remaining informed, curious and in control.

Episode Highlights

01:22 – Why AI tools are powerful assistants but should never be treated as infallible sources of truth.

03:06 – Understanding the jargon: Generative AI, Large Language Models (LLMs), ChatGPT, Copilot, Gemini and Claude explained.

06:42 – How AI companies make money and why many AI tools are moving toward paid subscription models.

10:28 – What Google AI Summaries are and how AI-powered search is changing the way we find information online.

14:51 – Why AI-generated summaries can remove important context and what that means for decision making.

15:46 – Kendra's "helpful eight-year-old" analogy for understanding how much trust we should place in AI outputs.

17:44 – The challenges of misinformation, internet quality, and why AI can sound confident even when it's wrong.

18:18 – The risk of equating confidence with expertise and why critical thinking matters more than ever.

20:21 – Practical examples of where AI is useful, including writing support, editing and communication.

22:56 – Understanding AI training data, bias and how historical information influences AI-generated responses.

27:22 – Why AI reflects the internet's strengths and weaknesses, including gaps in representation and cultural context.

30:17 – Free versus paid AI tools: what users need to know about privacy, data ownership and security.

33:44 – When organisations should consider investing in enterprise AI solutions rather than relying on free tools.

34:43 – Can AI support wellbeing and personal development? Exploring practical use cases beyond productivity.

36:54 – Using AI as a conversation coach, sounding board and learning partner rather than an expert adviser.

39:09 – Kendra's top advice for safely experimenting with AI and building confidence with emerging technologies.

Key Takeaway

AI can be an incredibly useful tool, but it still requires human judgement.

As Kendra explains, these systems are designed to generate convincing responses, not necessarily accurate ones. The challenge for all of us is learning to use AI thoughtfully, understanding its strengths and limitations, and developing the habits needed to verify information before acting on it.

The future belongs not to those who blindly trust technology, nor those who avoid it entirely, but to those who learn how to engage with it critically, confidently and responsibly.