Episode 020: Pain Points, Grok And The Homeless Summer Associate

What if the smallest law firms are actually the best positioned to win the AI era? That's the provocation Ron opens with this week — and it comes from someone with a genuinely rare vantage point: a solo attorney who also works as a professional AI trainer for the labs building these models.

In this episode:

• Dr. Sara Kubik's path from a gerontology-focused PhD to family law, estate planning, and now legal AI training

• Inside the AI training industry — how OpenAI, Anthropic, Google, and Microsoft contract out trainer roles through staffing layers, and why it's become a multi-billion dollar business

• Why older clients are often faster and more enthusiastic AI adopters than lawyers expect, once they see it's useful

• How different AI models produced very different results on the same creative request, and what that reveals about picking the right tool for the job

• Vibe coding a Medicaid penalty calculator, a community spouse calculator, and a child support chatbot without a formal development background

• The case for solo and small firm lawyers having a structural advantage over Big Law in adopting AI

• University of Chicago Law School's ban on electronics in 1L lectures — and why Sara thinks the policy won't survive the year

• What lawyers consistently get wrong about how AI models work, including training cutoffs and hallucinations

We also discuss:

• ChatGPT's new wave of specialized models and the ongoing comparisons to Claude

• Ron's controversial take on lawyers recording everything as firm data infrastructure

• Practice Signals: a Reddit post from a homeless summer associate, and where AI actually helps — and where it doesn't

• This week's Season 2 AI Build segment: cleaning firm knowledge assets before teaching them to AI, including PII anonymization and "grader guidance"

• The difference between vibe coding and understanding what you're building

• Data-focused stats on solo and small firm representation in the legal market

Key Takeaway

AI adoption was never really about the tool. It's about the pain point. Sara's advice to solos is blunt: stop shopping for tools and start naming what's actually broken in your practice — then go find out what fixes it. That reframe is what makes AI usable instead of overwhelming, and it's also what makes the "solo advantage" argument real: for the first time, small firms have access to something close to the same horsepower as Big Law, without the budget Big Law required to get it.

This one lands hardest for Simpsons lawyers — the ones already poking at ChatGPT on their phones but who haven't connected it to a real problem in their practice yet. It's also a reality check for anyone tempted to treat AI as a universal fix: the practice signals segment on the homeless summer associate is a reminder that some problems are human problems first, and AI is a research tool for finding help — not the help itself.

Mentioned in This Episode

• Dr. Sara Kubik (guest)

• Claude, ChatGPT, Gemini, Grok, Midjourney, DaVinci Resolve

• Codex (ChatGPT)

• Power BI, SQL, Excel

• Clio

• Perplexity

• Cooley Law School

• Purdue University

• University of Chicago Law School

• Indiana State Bar Association

• Reddit

[email protected]

Note: As mentioned in the episode: Is Mitch McConnell's photo real or fake? — Poynter/PolitiFact

Note: This photo has been the subject of online debate regarding its authenticity. We're linking to it as referenced in the episode and take no position on that controversy.