If you’ve ever assumed that ChatGPT, Copilot, Gemini and Claude all give you the same answers, think again.
In the first episode of AI Matters, host Kath Lord-Green sits down with AI expert Andrew Bel-Dean to tackle one of the most important questions businesses aren’t asking: are all GPTs the same? The short answer is, no.
Every AI tool is an application that has been trained using different data sets. Ask them all the same question, and you’ll get five different answers.
- ChatGPT draws heavily from Microsoft, with around 70% of its data coming from that ecosystem
- Gemini pulls from Google’s platforms, including YouTube, making it particularly strong for content-related queries
- Amazon Q pulls on data from Amazon, making it the natural choice for retail businesses
- Claude is shaped by Anthropic’s own research and data frameworks
- DeepSeek is strictly controlled, meaning certain topics simply don’t exist in its responses
The episode also covers some of the biggest risks of using AI in business:
- Hallucination: when an AI doesn’t have the answer and makes one up to keep you happy
- Model bias: every GPT is trained to please you, which means the more you use it without pushing back, the more it refines its answers around what you want to hear rather than what is accurate
Get all of the insights by listening to episode 1 here.
