Deep Dive Summary
In this third podcast, Steve sits down with Neil Peretz, a longtime friend and colleague.
Summary
In the introduction to the podcast, Steve welcomes Neil Peretz, a longtime friend of over 25 years, and they reflect on their time together at Libraria, a startup from the early 2000s. Neil had been the CFO and COO, helping to manage operations. Steve expresses admiration for Neil’s many successful entrepreneurial ventures. Neil mentions that, despite his successes, he's also had to shut down companies, once even returning money to investors, which he found difficult but necessary. Steve recently discovered that Neil attended Tufts and did military science at MIT through the ROTC program, something he hadn’t known despite their long friendship. Neil also shares that he later earned a law degree from UCLA and another in Europe while working on a project for the European Commission.
The conversation shifts to Neil's stint at the Department of Justice (DOJ). Neil reflects on the challenges of working in DC with a modest federal salary but emphasizes the high caliber of his colleagues. Steve shares an anecdote about Neil visiting him in Carlsbad with DOJ bagels, amusingly surprising his daughter with his official presence.
They then discuss a couple of Neil’s ventures, focusing on projects that helped many different people, such as Pocket Mail, a device that allowed users to send email on the go in the pre-smartphone era. Neil recalls it benefiting people in unique situations, like environmental activists and yacht owners. Steve compares the technology to AOL and highlights Neil’s knack for working with international teams. Neil shares that another project he's proud of helped small businesses navigate the complex SBA loan process, which he sees as a meaningful way to assist people.
Summary
In this next segment, Neil discusses a startup that developed an automated solution for SBA loan applications, simplifying the process for both applicants and banks. The system helps applicants by streamlining the process and assists banks by reducing the number of people needed for underwriting, while maintaining higher compliance levels. It also matches borrowers with the right banks based on loan preferences, making it a B2B matchmaking tool for small businesses to access affordable financing.
The conversation transitions to Neil’s experience as an entrepreneur and lawyer, with Steve reflecting on how entrepreneurs often create things they need but can’t find, though Neil says his journey into law may have been influenced by family arguments. They discuss how many of Neil’s ventures have had legal components and touch on one of his past startups, which dealt with bankruptcy claims but eventually shut down due to cofounder issues.
The topic shifts to AI’s potential impact on professions, especially lawyers. Neil jokes that while AI, like ChatGPT, can pass the bar exam, so can many poor lawyers. They discuss the growing role of AI in the legal field but emphasize that the nuances of contract negotiation and individual client needs still require human insight, something AI might not easily replace. Neil highlights his company, Contract Wrangler, which helps manage completed legal agreements by identifying key contract terms, though he acknowledges the more complex dynamics involved in negotiating contracts, which go beyond the written agreements.
Summary
Their conversation continues around the evolution of large language models (LLMs) and their increasing relevance, especially as they interface with current events and data from the web. While acknowledging advancements in AI, Neil mentions how previous attempts like their “Legal BERT” model failed to address certain nuanced tasks in law. Neil emphasizes that AI, including LLMs, shouldn’t be overused or assumed to solve all problems. Steve then draws parallels to the evolution of IT infrastructure, predicting that AI will follow a similar path of decentralization—eventually being personalized and integrated into everyday devices.
They discuss the idea that AI could eventually function as a “personal agent” tailored to individual needs, referencing their friend’s theory of personal AIs that represent users in various contexts. The discussion touches on how LLMs still struggle with producing “ground truth,” meaning that AI can often produce plausible but incorrect information. Legal fields, with their documented facts, might be better suited for AI applications, but nuance is essential, especially in disciplines like law where details of cases matter.
Steve compares the AI boom to previous tech cycles, like the dotcom and genomics eras, noting that while significant investment leads to rapid progress, a market correction may be looming. Steve concludes this segment by discussing the potential for AI in specialized areas, such as generating mathematical models for medical research, highlighting how AI could one day perform complex simulations like clinical trials digitally. The concept of a “digital twin” in medicine, where AI models real-world biological systems, is becoming increasingly viable and could revolutionize how healthcare operates.
Summary
The final segment initially discusses a digital twin of a house platform, allowing for 3D printing of walls with pre-built channels for conduits and outlets. This technology enables custom-made, rapidly fabricated homes. Neil then mentions he’s a licensed real estate broker, which is a common qualification for lawyers who can sit for the exam with some studying. The conversation shifts to the real estate sector, where digital tools allow for virtual renditions of properties, reducing the need for repeated photography when staging changes.
Neil shares that he’s currently pursuing a few different ventures, one of which involves sitting for the Arizona bar exam. He highlights Arizona’s unique legal landscape, where recent changes allow for partial non-lawyer ownership of law firms, an uncommon structure in the U.S. This could attract private equity interest, especially since Arizona law firms can now offer services like litigation funding. Neil predicts that applying technology to legal services could lead to increased cost efficiency in the industry.
For a music clip - again since August was less mix heavy, I pulled more from the July 3rd party which had several mixes of people typically not playing together. Here is a fun rendition of Ex’s * Oh’s with two of my favorite singers - Tammy and LaurinW/ (TimD-Bass, LaurinW-Vocals, Jen-Guitar, Tammy-Vocals, NickB-Guitar, and DonP-Keys). For some reason the first few seconds are muted in substack, so make sure the speakers are set right…
A Quarter Century+: Talking with Neil Peretz