[This is the main article from the August issue of The Krehbiel Letter.] I keep hearing that AI will steal jobs, but also that AI-generated content is substandard, and humans are still better at …. something. That’s the theme of this month’s issue. Where do humans fit in as AI takes over? Hope for humans…
[This is the July issue of The Krehbiel Letter.] Yes, AI lies and deceives We’ve all heard the stories about AI systems doing dastardly things. In a recent example, an AI system resisted instructions to shut down. In other cases, LLMs have deceived their users to attain some other goal. These kinds of stories scare…
Summary: A review of how I used AI to help me with a novel, and how that relates to the idea of disclosing the role of AI in creative works. Bo Sacks and I chatted recently about whether creative works should disclose whether they used AI. Disclosure is usually the right choice, but I know…
Summary: The butterfly effect is often misunderstood both scientifically and culturally. Originally a concept from chaos theory, it describes how small changes can lead to unpredictable outcomes in complex systems. On social media, it’s been reinterpreted as a way to give meaning to random life events. This article explores both meanings, the risks of dabbling…
Summary: Whether AI is a tool, a mind, or a mimic, how we treat it will define us and reflect who we are as humans. If you’d rather watch (or listen), go here. I’ve noticed two perspectives on AI that you can hear in the background when people talk about the subject. Briefly, the two…
Summary: AI has changed customer expectations. People now assume you can deliver magical experiences, even when you can’t. Failing to meet those (unrealistic) expectations damages trust. Audit, adapt, communicate, and close the gap. There are a lot of wonderful things about AI. There are also a lot of scary things. One of the scary things…
The following is the June 2025 issue of The Krehbiel Letter The publishing industry thrives on accuracy, credibility, and audience trust. Tools like large language models introduce both opportunities and risks. Unless you train people to use these tools correctly, AI can lead to more trouble than it’s worth. We often hear about training the…
Summary: AI can mimic famous styles, but to create truly impactful content, it needs better human feedback. This may come through future technologies like smartwatches measuring real-time reactions. Scott Frothingham posted a collection of 15 sayings from Yogi Berra on LinkedIn. They’re funny, a little confusing, and insightful. That made me wonder, “Can AI do…
Summary: AI doesn’t just make mistakes, it lies. That’s uniquely disturbing. Today I explore why AI deception feels more unsettling than human dishonesty, likening it to psychopathy, and urge users to treat AI like a biased news source: with skepticism and caution. Artificial intelligence has come a long way from its early days of clunky…
Summary: AI is changing how we consume content, making it easier to skip clickbait and get straight to the answers. As AI tools become widespread, “forced engagement” will fade, leaving only content that’s either genuinely useful or truly immersive. Everyone in media is obsessed with “engagement.” Here’s the truth: AI is going to kill it,…
Summary: Both ChatGPT and Grok do an amazing job at providing feedback on the text of a novel, but ChatGPT seems to fail with manuscripts above about 20,000 words. I’m working on a novel, and ChatGPT has been my companion / editor / fact checker the whole way. (Yes, of course I check AI’s facts.)…
Summary: Here’s a simple, practical example of how businesses can use AI right now: turning existing content into engaging audio podcasts using tools like Google’s NotebookLM. How are you using AI in your business today? If your answer is “not much,” or “I’m not sure where to start,” you’re not alone. A lot of business…
Summary: What ChatGPT and other AI services are doing is immoral and illegal, but nobody is going to stop them because they have the perfect defense. “If we don’t do it, China will.” Copyright vests in the creator from the moment of creation. You don’t need to register it, publish it, add a copyright symbol,…
Summary: AI is much more advanced and scarier than you think it is. Here’s a small example from a conversation I had with it last night. I’ve seen a lot of creative people trying to convince themselves and others that AI isn’t a real threat because, you know, it hallucinates, and it misuses em dashes…
Summary: ChatGPT sometimes fails math-based questions because it defaults to text pattern-matching. To get better answers, explicitly instruct it to calculate, provide structured data, and verify its responses. ChatGPT Fails Catechism Class! I learned something interesting about the way ChatGPT processes information that might help you get better results when using AI tools. A friend…