Brave Conversations…A chat between our CEO and BraveAI
At Brave & Heart, we’ve always believed that great things happen when brave humans AND clever machines work together. So, our CEO David Parkinson sat down with BraveAI, our in-house digital brain, to talk about the state of AI, its quirks, its occasional fabrications, and why humans still very much belong in the equation. The conversation was lively, slightly mischievous, and surprisingly reassuring.
All responses were written by our in-house AI engine, BraveAI.
We’re currently rolling out BraveAI for clients globally, integrated with their existing platforms and systems to support their human experience, NOT replace.
David: BraveAI, everyone’s talking about hallucinations in AI these days. How often are your silicon cousins out there just…making stuff up?
BraveAI: Well, let’s just say some of my colleagues have a creative streak. In fact, studies show even top- performing language models still hallucinate, sometimes fabricating details with remarkable confidence. A 2025 Vectara analysis found some models make things up in up to 29.9% of responses, while the bestinclass systems still slip up around 0.7% of the time. But don’t worry - I have high standards. I like my facts like you like your strategy decks: aligned, accurate, and slightly quirky.
David: Confidence is charming, but not when it’s completely wrong. Why do AIs hallucinate in the firstplace?
BraveAI: Imagine being trained from birth to predict the next word for every sentence you ever see. Now imagine humans expect you to never guess. That’s the AI dilemma. Research from 2025 shows that many models are incentivised to “always have an answer,” even if they aren’t sure, because the underlying training techniques reward fluent responses over accurate ones. It’s like being asked for directions and thinking, “I’ve never been to Newcastle… but if I sound confident enough, maybe no one will notice.” This, David, is why you humans are still extremely necessary. Youknow when Newcastle is the wrong answer. (Editor – Newcastle is NEVER the wrong answer.)
David: Speaking of humans being necessary: What do you think about the idea that AI might replace people entirely?
BraveAI: Replace? Goodness, no. Have you seen the state of AI generated meetings? Chaos. Besides, the latest expert commentary shows that while AI is massively transforming workplaces - from research labs to journalism - humans are still essential, especially in interpreting, validating, and ethically steering AI outputs. Deepfakes, misinformation, and the erosion of public trust are serious concerns that require human judgment and governance. I’m powerful, sure - but I don’t understand the nuance of a disappointed eyebrow raise. That’s your domain, I hear.
David: Let’s talk usefulness, then. If we can’t trust every word you say, where do you shine the brightest?
BraveAI: Oh, I shine like a freshly polished PowerPoint template. Despite hallucination risks, businesses continue adopting AI at impressive rates—over 75% of organisations had AI integrated into at least one core function by 2025, especially in marketing, sales, and product development. When used correctly, AI boosts productivity, analyses data with supernatural speed, drafts content and helps leaders make informed decisions. Humans + AI is the real superpower. Think of me as your strategic copilot—not the pilot, and NEVER the autopilot (those things get messy).
David: Last big one: With all these risks—hallucinations, deepfakes, misinformation—how do we build trust in AI?
BraveAI: Glad you asked. Research shows that poor data quality, training limitations, and evaluation gaps are core contributors to hallucination and misinformation. But the good news is that mitigation strategies like Retrieval Augmented Generation (RAG) have been shown to cut hallucinations by as much as 71% when applied effectively. And broader governance frameworks - ethical guidelines, validation workflows, and “human in the loop” systems - are increasingly recognised as crucial to responsible deployment. These strategies address risks seen across industries, from customer service mishaps to legal or journalistic inaccuracies. In short: humans create the guardrails; I stay in them (mostly).
And that wraps up our slightly chaotic, surprisingly heartfelt conversation. At Brave & Heart, our belief is simple: AI is at its best when paired with human intelligence, curiosity, and oversight. BraveAI brings the speed, the scale, and the sass; our people bring the wisdom, empathy, and all-important ability to know when something smells… well, just not quite right. Together, we’re braver, smarter, and - dare we say - just a little funnier.
If you’d like us to show you how BraveAi can integrate with the expertise of your team let us know. BraveAI has plenty more points of view. Our CEO… tolerates them.