Title

Monday Update: Testing Kimi, Building Better AI Results, and Why Small Podcast Audiences Still Matter

Jason Todd Wade tests a new microphone and delivers a Monday update on the AI tools, models, and publishing systems he has been experimenting with.

He discusses the sudden demand surrounding Kimi and its newest model, why Chinese AI systems such as Kimi and DeepSeek can be frustrating but unusually effective for finding overlooked information, and why users should never judge an AI system solely by its first answer.

The real skill is iteration. Jason explains how he continually challenges, redirects, and improves AI output rather than expecting a polished result from the first prompt. He also examines how cultural differences influence the way American, Chinese, and European AI models organize and present information.

The episode then moves into AI voice and video production. Jason compares ElevenLabs and VEED for voice cloning, long-form audio, podcast clips, subtitles, and YouTube production, including how he uses the platforms to produce narrated content while managing credit costs.

Finally, he addresses the reality of starting a podcast. New shows often receive little or no initial traction. Jason argues that creators should build for their business, expertise, and long-term authority rather than chasing inflated audience numbers. Thirty-eight relevant listeners can be more valuable than thousands of irrelevant ones.

Even when a podcast does not immediately attract a large human audience, the published content still creates searchable, indexable material that AI systems, search engines, and recommendation platforms can discover over time.

Topics discussed:

Testing Kimi and its newly released AI model
Why DeepSeek has become more useful over time
The strengths and weaknesses of Chinese AI models
Why the first AI response is rarely the final answer
Cultural differences in AI-generated information
ElevenLabs versus VEED for voice cloning
Producing short-form video podcasts with AI
Why new podcasts receive little initial traction
Relevant audiences versus vanity metrics
Podcasting as long-term AI visibility infrastructure