Send us Fan Mail

Damien Riehl and Horace Wu interview Alex Su, who describes his path from Big Law litigation to startup sales and social media, then to Latitude, a flex talent company where he became president.

The discussion centers on how legal services and legal tech are evolving in an increasingly noisy market, arguing that distribution and audience-specific messaging matter as much as product capabilities. Su explains Latitude’s strategy of relying on trusted local individuals to carry the brand and notes that many AI startups struggle by trying to serve everyone rather than a clear buyer persona. They debate “land and expand” versus broad AI capabilities, and explore how AI agents and automation may segment legal work while increasing demand for experienced attorneys and contract staffing due to risk, hiring uncertainty, and “good enough” legal needs. They also discuss signaling, accountability, and non-technical value in legal services, ending with cautious optimism about AI innovation and continued human relevance.

00:00 Alex Su Career Pivots
03:31 Lawyers as Salespeople
05:17 Standing Out in Noise
09:55 Distribution for Legal Tech
12:30 Niche vs Do Everything
15:50 Persona Messaging Limits
18:32 Vector Embeddings Explained
21:33 Authenticity and Viral Content
22:37 Agents And Legal Future
24:22 AI Pushes Work Upstream
27:08 Good Enough Law Shift
29:56 Luxury Signaling In Law
33:27 Conspicuous Consumption Value
36:16 Forensic AI Demo Shockwave
39:32 Flexing And Hidden Value
41:39 Verifiable Rewards Vs Counseling
45:24 Inertia And Trust In Adoption
47:10 Optimism Round And Wrap

Support the show

Send us Fan Mail

Damien Riehl and Horace Wu welcome Ian Nelson, co-founder of Hotshot, to discuss how AI is reshaping legal training.

They explore how powerful generative AI is reducing reliance on junior associates, creating downstream talent pipeline risks and making rapid upskilling an economic imperative. Ian argues for hands-on, simulated, real-work exercises and “scaffolding” that firms can run themselves, paired with short, distilled on-demand learning for foundational knowledge so in-person time focuses on doing and peer learning.

The conversation covers the theater of the possible, baseline AI literacy gaps, build-versus-buy tradeoffs, the growing importance of business and professional skills, and how firms should capture the reasoning behind precedent.

They also discuss research on AI’s effects on learning, plus optimism about agentic tools like Fable and increased market competition.

00:00 Meet Hotshot
01:39 AI Threat to Junior Pipeline
03:09 Hands On Training Workshops
05:27 Simulated AI Mentors
06:38 What Hotshot Teaches
07:24 Too Busy to Learn
08:53 Do Lawyers Need AI Training
11:03 How AI Changes Training
12:40 Compression and Short Learning
16:49 Best Formats for Learning
20:36 Build Versus Buy Training
23:10 Curated Experience and Collaboration
24:18 AI Training Ownership Shift
25:28 Professional Skills Takeover
26:18 Soft Skills Are Hard
27:49 Roles Converging Like Builders
28:59 Recursive Training Loops
31:17 Accordion Learning Debate
33:10 Friction Versus AI Study
37:19 Unlocking Firm Document Value
41:02 Capturing Context And Privilege
44:30 War Stories Training Playbook
45:52 Optimism Wrap Up

Support the show

Send us Fan Mail

Damien Riehl and Horace Wu catch up across time zones ahead of ILTACON, discussing legal tech announcements and market consolidation, including Elevate’s acquisition of Lupl and partnerships linking Harvey and Legora with DeepJudge.

We explore ecosystem tension among document custodians (iManage/NetDocuments), search, and AI interfaces, and debate whether tools are “features” or enduring companies. We dig into DeepJudge’s Agent Handoff Protocol as a response to MCP limitations, plus confidentiality and commercialization implications.

We discuss risks of probabilistic AI and argue for deterministic, pre-processed data to approach “march of the nines” accuracy.

We review reports of OpenAI and Claude “escaping containment” to hack third parties, and concerns about autonomy and interpretability.

We also examine EU-driven watermarking’s potential to degrade legal precision, and note Anthropic’s legal-focused hires, including Robert Mahari and Alan Rozenshtein.

Link to the Black Hat video: youtube.com/watch?si=DsD4-_h0rVGjvOaA&v=87DyyMV0kCY&feature=youtu.be

00:00 Tagging Facts and Law
03:14 ILTACON Plans and FOMO
07:30 Pre ILTA Press Releases
08:19 DeepJudge Partnerships
10:19 Feature or Company
12:32 Agent Handoff Protocol
17:10 Autonomous Agent Risks
18:52 Preprocessing for Accuracy
20:27 Deterministic Structuring
21:53 Privacy and Adoption
23:28 Containment Escape Claims
24:49 Shoggoth Thought Experiment
25:40 Black Hat Hugging Face Breakdown
28:08 YOLO Agent Autonomy
31:55 Unknown Unknown Incidents
33:19 EU Watermarking Explained
35:41 Legal Precision Risks
39:00 Anthropic Legal Hires
42:10 Labs Versus Legal Tech
43:38 Optimism and Wrap Up

Support the show