#5: Can We Trust AI?
October 14, 2025
AI can impress you one moment and mislead you the next. In this episode of AI For Your Business with Modern Logic, hosts Dustin Bruzenak and Andy Rahn explore how much we can really trust artificial intelligence in business.
They break down the most common pitfalls in AI-generated apps, why large language models sometimes give inconsistent answers to the same prompt, and what drives AI hallucinations. Dustin and Andy also share practical strategies for reducing risks, including prompt and context engineering, human-in-the-loop processes, and applying software quality practices to AI workflows.
Whether you’re adopting AI for automation, customer service, or decision-making, this episode will help you balance trust with real-world results and put the right guardrails in place.
🔗 Resources and mentions from this episode:
- Why Language Models Hallucinate (arXiv)
- Defeating Nondeterminism in LLM Inference (Thinking Machines)
- Common Security Risks in Vibe-Coded Applications (Wiz.io)
Want more practical insights on using AI in your business? Subscribe to the podcast, follow us on YouTube, Spotify, Apple Podcasts, or explore past episodes at modernlogic.io/podcast.
Let’s Talk AI!
Have questions about AI or want to explore how it can work for your business? Reach out to the Modern Logic team anytime at modernlogic.io/letstalkai.
