Everyone was promised that AI would hand back their time — so why does it feel like you're babysitting four terminals, clicking "yes, allow this" one prompt at a time, and re-reading long specs just to check whether the model did what you asked?

In this episode of Why Isn't This AI Thing Working, Jess Lampe talks with Jim and Alysha Rosser, co-founders of Entry Studios and the team behind Grove. Their take on why AI stalls after the pilot: it's a trust problem. Agents skip instructions, invent their own commands, and declare victory before the work is done — quietly shifting the cognitive load onto whoever reviews the output next.

Grove is their answer, and it started as a tool they built to solve their own frustration. Jim and Alysha make the case for wrapping AI in old-school determinism: task lists with hard proof requirements, permission "boundaries" that redirect an agent instead of just blocking it, and executive summaries that show the receipts on every session. The result is a two-person team that queues work overnight and walks into roughly ten reviewable PRs each morning — about 900 merged since July. Along the way: cognitive load, "making determinism cool again," designing upfront, and why living in the chat window all day means you're doing it wrong.

In this episode of Why Isn't This AI Thing Working?, host Jess Lampe sits down with Alex Kozlowski, Director of Software Engineering at Launch Consulting, to cut through the jargon and explain how AI systems actually work.

Alex and Jess demystify the vocabulary everyone uses but few define: what a large language model really is, why LLMs are probabilistic (and why that makes hallucinations inevitable), and what separates an LLM from an "agent" and an emerging "harness" like Claude. Using plain-language analogies, they explain how tool-calling loops create the emergent behavior people find so impressive — and where these systems predictably go wrong.

They also tackle adoption reality: why teams with strong DORA metrics get the biggest gains, and why most AI proofs-of-concept never reach production.

In this episode:

• How edge AI differs from cloud-based AI
• When running LLMs locally makes business sense
• Understanding token-based AI pricing and consumption
• Comparing privacy, latency and infrastructure tradeoffs
• Evaluating the ROI of local and cloud AI models
• What technology and finance leaders should consider before scaling AI investment

In this episode:

• Building customer trust in AI-powered experiences
• Balancing personalization with responsible AI practices
• Using guardrails to improve reliability and reduce errors
• Measuring AI performance beyond technical accuracy
• Scaling machine learning and agentic systems responsibly
• Maintaining human oversight as AI becomes more autonomous

In this episode:

• Challenging common myths about generative AI
• Understanding what AI can and cannot do
• The difference between artificial intelligence and augmented intelligence
• Why human creativity and critical thinking remain essential
• Moving beyond AI hype to identify practical business use cases
• Using AI to strengthen human capabilities rather than replace them

In this episode:

• Building an AI business case that resonates with CFOs
• Connecting AI investment to measurable business outcomes
• Why cost reduction alone is not a complete AI strategy
• Selecting AI use cases with clear potential value
• Understanding how token-based pricing affects financial planning
• Measuring AI ROI and deciding which initiatives to scale

In this episode:

• Using curiosity and experimentation to make AI more approachable
• How humor and creativity can reduce resistance to AI
• Building AI literacy without requiring technical expertise
• Creating low-pressure opportunities for people to explore AI
• What live audience reactions reveal about AI adoption
• Helping employees develop confidence using AI tools

Learn more about It Writes Itself at itwritesitself.ai.

See Mike's magic at hellomagicmike.com.