AXIOM Insights Learning and Development Podcast

The episode provides practical, hands-on examples of AI tool usage across three distinct professional contexts (Excel/Copilot, project management, contract analysis), moving beyond generic "AI is transformative" platitudes. However, much of the insight centers on well-known LLM limitations (hallucinations, citation problems, the need for human validation) that have been widely discussed for 18+ months. The depth on novel applications and non-obvious patterns is moderate rather than exceptional.

“It's not intended to do your work. It's intended to be an assistant. Just like on a plane, there's a pilot and a co pilot.”

“garbage in, garbage out with garbage in gospel out. We think because the AI said it, it must be so”

The core framing - AI as a copilot/assistant, the need for human-in-the-loop validation, governance models, and critical thinking - has become mainstream in B2B AI discourse by 2024. The specific examples (Excel dashboards, contract analysis, grant evaluation) are practical applications but not conceptually novel. The discussion of bias and governance is present but surface-level compared to deeper organizational risk frameworks.

“you have to check the output you get. Um, I find it very useful for creating pivot tables”

“we're mistaking garbage in garbage out with garbage in gospel out”

The three guests bring legitimate operational experience: Madeline runs a consultancy actively integrating AI into contract/instructional design workflows; Jamie teaches 50+ computer classes and Microsoft Office certifications; Bethany is a senior IT PM piloting agentic solutions at an R1 university. However, none are C-level operators, heads of product, or executives building AI strategy at major organizations. They are skilled practitioners and implementers, not strategic architects.

“I run an instructional design workforce development consultancy. Um, and over the past year or so I've been deliberately integrating AI into my workflow”

“I'm the owner of Launch, a computer training and consulting firm in the Boston area. I teach over 50 computer classes”

The episode includes concrete examples: Excel data cleaning, pivot tables, HTML dashboards, contract document analysis of federal RFPs, grant evaluation workflows at Brown University, and specific AI tools (Copilot, ChatGPT, Claude, Gemini, Notebook LM). However, it lacks quantified business impact (time saved, cost reduction, error rates, adoption metrics). No specific metrics on the grant evaluation pilot (6-8 weeks in, but no numbers on false positives/negatives or time savings). No dollar figures or scaled deployment data.

“summarize that content, even if there are hundreds of thousands of records”

“I can get it to scan and get it to do hours of analysis in minutes”

The host (Scott Rutherford) asks open-ended setup questions and allows guests to develop points, but rarely challenges claims or pushes back on assertions. When Madeline discusses governance and human judgment, the host agrees rather than probes deeper. Jamie's claim that "AI will put people out of work" goes unchallenged. The Brown University exam cheating anecdote is interesting but diverts rather than deepens the conversation about L&D strategy. Limited follow-ups on concrete trade-offs or failure modes.

“Yeah. You're still on the hook for it at the end of the day, too. Right?”

“And that speed for the analysis speed. I think I'm m with you there. I see the benefit of that.”