Twenty-five years ago, Tim DeSieno and I were two outsiders on the tropical island of Singapore, me trying to build a startup, him fresh out of a restructuring law practice. We reconnected recently on the Fireside PM podcast, and what followed was one of the most illuminating conversations I've had this year.
Tim's career arc is anything but conventional: from decades in global debt restructuring to litigation finance investor, and now advisor to an AI legal startup. The conversation, which started as a reunion, turned into a firehose of insight—for lawyers, founders, and especially product managers trying to anticipate where disruption lands next.
This post distills that hour-long conversation into key lessons for early- and mid-career product managers. Whether you're wrangling roadmaps at a Series A startup or driving platform strategy at a late-stage unicorn, you'll find practical frameworks, surprising analogies, and a peek into the wild intersection of law and AI.
1. Litigation Funding Is What Early VC Investing Looks Like in a Non-Tech Industry
"We would look at 100 cases, take three seriously, and maybe fund one."
Tim described litigation finance as a "venture capital" approach to legal claims. Funders underwrite the legal equivalent of startups: high-risk, high-reward lawsuits with uncertain outcomes. The investment model is classic VC—non-recourse funding in exchange for a percentage of winnings—but applied to torts, sovereign disputes, and commercial litigation.
This is a also a class in triage. As PMs, we're sometimes guilty of over-indexing on tech, TAM or user demand without enough scrutiny of distribution or defensibility. In litigation finance, everything must be strong: the legal basis, the plaintiff’s character, the likelihood of enforcement.
Actionable Advice:
When evaluating new bets, use a PM version of Tim’s triangle: Strength of case, rational actor, enforceability. Substitute your product’s domain as needed. If your bet falls apart on any leg, kill it early.
Don’t be afraid to walk away. "We’d spend weeks researching only to discover a fatal flaw." Avoid sunk cost fallacy.
2. The Real AI Gold Rush Isn’t Just Generation, It’s Prediction
Harvey (the legal AI startup backed by OpenAI) gets the headlines, but Tim is on the board of an earlier stage adjacent player called Canotera. Instead of drafting, Canotera predicts litigation outcomes. Think of it as a risk analytics layer built from all New York legal precedents, offering lawyers (and insurers, GCs, even arbitrators) a probabilistic view of their odds.
"It’s like calling up a senior partner and getting a second opinion—except this one has read every case."
This isn't just a better way to write memos. It's a decision-making accelerator.
Product Insight: There are many types of AI value in any vertical:
Efficiency (do more, faster)
Accuracy (better outcomes)
Confidence (de-risking decisions)
Harvey is largely #1 and #2. Canotera is going hard at #3.
Actionable Advice:
When building AI products, map your feature set to these value levers. Which one are you really selling?
Don’t sleep on #3—especially in regulated or high-stakes domains, confidence trumps speed.
3. Adoption Gaps Aren’t Just Technical—They’re Psychological
"The number of people in law who haven’t touched ChatGPT is shockingly large."
Sound familiar? We’ve all worked with that PM, eng lead, or exec who in late 2022 who thought gen-AI was a toy. The parallel to law is stark: many lawyers fear AI not because it's ineffective, but because it threatens their identity.
In both professions, billing hours and writing decks have long been proxies for value. When those tasks are automated, the insecurity is real.
Actionable Advice:
Frame AI as augmentation, not replacement. Tim noted the firms that are thriving are those that say, “Yes, we bill per hour—but we’ll use AI to deliver more per hour.”
Early adopters are not just tech-savvy—they're secure enough to rethink their role. When evangelizing AI, target the curious and the confident.
4. “Doctrinal vs. Practical” Isn’t Just a Law School Problem
"You come out of law school, and you're good at arguing both sides. But no client wants that."
Tim called out how legal education—especially the Socratic case method—trains great thinkers but poor practitioners. Law grads often need years of on-the-job experience before they become useful to clients.
Sound like any junior PMs you know?
Product teams are often full of doctrinal thinkers—people great at debating frameworks, prioritization models, or vision decks. But if you can’t turn that into a working prototype, a roadmap aligned with GTM, or a tough tradeoff call, you’re not adding value.
Actionable Advice:
“Thinking like a PM” (strategy, ambiguity, storytelling) is necessary but not sufficient. Pair it with executional reps early in your career.
If you’re a manager, give your ICs reps they can own end to end. Treat it like an apprenticeship, not just a theoretical seminar.
5. Liberal Arts Still Matter—Even in the Age of AGI
"If you can’t write it clearly, you don’t own it."
Tim made a powerful case for the liberal arts as the antidote to AI passivity. He sees students turning in polished work generated by LLMs but lacking any real grasp of the content. Writing, he argues, is thinking. If you can't articulate a point unassisted, your judgment muscles don’t get built.
Actionable Advice:
Don’t outsource the first 70% of a product brief, strategy doc, or roadmap to ChatGPT. Use AI to refine and stress-test, not originate.
Push yourself to learn something uncomfortable. Tim’s litmus test: "Do hard things that are new to you. That’s how you grow judgment."
6. You’re Not Competing With AI, You’re Competing With Humans Using It Better
"A junior lawyer with AI tools can be more valuable than a senior one without."
In a decade, your job won't be taken by AI—but it might be taken by someone with 5 years less experience who knows how to pair human empathy with AI speed.
Actionable Advice:
Learn prompt engineering, yes—but also get great at evaluating AI output. That judgment layer is what companies will pay for.
Practice defending ideas live, without a script. At some point, someone will ask, “Why did you make that decision?” Be ready.
7. Forecasting the Endgame: When Courts Run on Code
"Maybe one day litigation disappears—two parties upload their facts, the machine decides, and that’s enforceable."
While Tim was cautious to say this vision is far off, the implications are worth pondering. What happens when not just lawyers, but judges, juries, and arbitrators are augmented—or replaced—by machines?
Whether or not this comes to pass, the lesson is clear: no profession is immune. If law can be automated, so can most knowledge work. And product managers will either ride that wave—or be washed away.
In Closing
As PMs, we love talking about disruption—but we rarely get to see it play out in an industry as slow-moving and tradition-bound as law. That’s what made this conversation with Tim DeSieno so instructive. Law is changing. AI is changing. And the humans who thrive are the ones who stay curious, adaptable, and relentlessly focused on value—not ego.
If this resonated, I offer 1:1 coaching for product leaders at tomleungcoaching.com, and PM consulting through paloaltofoundry.com.
OK. Enough pontificating. Let's get back to work.
Share this post