Ep #38 To Lovable or not To Lovable

Most organizations pick AI tools by asking the wrong question. They benchmark capability, compare feature lists, and choose the most powerful option then wonder why adoption stalls at 10-15% six months later. Kevin and Matt Graham dig into the real selection criteria leaders should be using, and why usability almost always beats raw capability when it comes to organizational ROI.

This episode started as a debate about Lovable the vibe coding platform that just raised $400M at a $13B valuation and turned into a broader conversation about what it actually takes to get AI working across an entire organization, not just in the hands of your most technical people.

Kevin and Matt cover the Lovable debate from both sides (Kevin’s been burned, Matt’s a believer), the emerging governance risks of agentic AI tools like Grokbot, why vendor lock-in is a real concern that organizations need to plan for, and where Google, XAI, and the competitive model landscape are actually headed.

Key topics covered:

✅ Why usability beats capability as an AI tool selection criterion
✅ Lovable’s enterprise evolution what’s changed and what hasn’t
✅ Vendor lock-in: how to preserve optionality without abandoning a useful tool
✅ Grokbot and agentic AI the governance risks organizations aren’t ready for
✅ The competitive model landscape: XAI’s rise, Google’s pivot, and what it means for token costs
✅ Why spreading AI licenses without organizational context produces nothing