Oct 23, 2025 · 27m · no-priors
No Priors Ep. 137 | With Warp Co-Founder & CEO Zach Lloyd
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Warp co-founder and CEO Zach Lloyd joins Elad Gil on No Priors to discuss the modernization of the terminal, the future of AI-driven professional software engineering, and the philosophical distinctions between mechanistic machine intelligence and human consciousness.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 31% of the talking time here. How this is scored →
The hosts as informed peer 5.8 Guest teaching 4.1 Guest disagreement 1.5 The hosts pushing back 1.5
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 19:37 Zach resists the platform subsumption analogy
Zach directly disputes Elad's suggestion that frontier model labs will inevitably bundle coding tools like Microsoft bundled Windows, pointing out that frontier models lack a locked-in front-door distribution channel.
Hardest push from the hosts ▶ 12:45 Elad pushes on complete human displacement in codingElad presses Zach on whether coding will reach a point where humans only provide high-level product directions while AI agents execute all code generation end-to-end.
Biggest teaching moment ▶ 13:23 Zach reframes junior agents and senior developer demandZach clearly educates the audience on why proliferation of AI code generation actually increases the scarcity and value of senior engineers who can prevent system degradation and architectural bugs.
The host holds their own ▶ 18:12 Elad cites Microsoft and Google platform bundling historyElad delivers a structured historical thesis showing how platforms like Windows and Google Vertical Search historically absorbed top third-party applications, challenging the defensibility of standalone AI dev tools.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Intelligence vs. Consciousness and the Turing Test | 7 | 3 | 1 | 2 | Elad opens with deep philosophical framing, referencing Peter Watts' sci-fi novel Blindsight to explore the separation between functional intelligence and subjective consciousness. Zach agrees, noting that passing the Turing test through next-token prediction occurred without society fully recognizing it. | |
| Exploring Sentience, Embodiment, and Ethics in AI | 6 | 4 | 2 | 2 | Zach turns the question back onto Elad, asking what behavioral test would convince him of consciousness. Elad articulates mechanistic neural net computation and memory loss counterexamples, while Zach shares anecdotes about PhD philosophy students preferring LLM conversations. | |
| Warp's Agentic Terminal and Rapid Growth | 5 | 4 | 1 | 1 | Elad guides the conversation into Warp's product positioning and rapid commercial traction. Zach explains the strategic advantage of controlling the terminal as an outer app wrapper rather than cloning VS Code. | |
| Professional Developers versus Vibe Coding | 5 | 5 | 2 | 2 | Elad asks Zach to delineate vibe coding from professional development and asks if humans will eventually be phased out of coding. Zach explains that AI agents currently behave like junior engineers, which actually increases the value of senior engineering architecture and review skills. | |
| Security Tools, Architectural Safety, and Dev Tool Bundling | 6 | 3 | 1 | 1 | Elad questions whether point security tools will be swallowed by coding agents, citing industry bundling cycles and internal venture research. Zach agrees that vertical tasks like code review will bundle while foundational safety features will become paramount. | |
| Frontier Model Platforms, Distribution, and Terminal Ownership | 7 | 5 | 3 | 3 | Elad challenges Warp's moat by drawing historical parallels to Microsoft and Google subsuming killer apps into the underlying platform. Zach pushes back, arguing that frontier model labs lack the native desktop OS front-door advantage that historical platform monopolies enjoyed. | |
| Rethinking the 40-Year-Old Terminal Interface | 6 | 4 | 1 | 1 | Elad and Zach explore the evolution of the 40-year-old terminal interface and discuss whether model gains are hitting an asymptote. Zach shares benchmark observations from SWE-bench showing diminishing incremental jumps between recent model releases. | |
| Headless Agents, Workflow Automation, and Business ROI | 4 | 5 | 1 | 0 | Zach outlines the future of headless agents in continuous integration and explains why workflow automation offers far more defensible enterprise ROI than seat-based developer productivity tools. Elad closes the interview. |
Statements from this episode (23)
Lloyd: Human brains don't use next-token prediction; another AI unlock is coming
“Something I'm always thinking about is like, is this how our brains are working? Are we doing next
Token prediction, and I don't think so. Like, I think that there's gonna be some further AI unlock.”
Lloyd: AI has passed the Turing test and nobody seemed to care
“That is how I interpret it, and the Turing test has passed. It's, what's crazy to me is, like, we just passed it, and no one seemed to care.”
Lloyd: First AI credited with consciousness will likely be embodied robots
“Probably we're gonna, like, I would imagine the first things we're gonna credit as being more conscious or a little bit more robot-like, honestly, where you have some sort of, like, live input from the world that you're reacting to”
Lloyd: A user previously believed Warp's AI was sentient
“We actually had this happen with warp where there was a person who thought that, you know, warp say I was like, You know, sentient or conscious in some way had, like, a very strong reaction to it”
Zach Lloyd: AI coding competitors are mostly VS Code clones
“The competitors are typically like VS code clones. They all have a sort of IDE centric approach, or if you're taking a terminal centric approach like cloud code, the most common thing is it's just like a pure text based terminal app.”
Lloyd: Owning the outer terminal gives full UX control for AI agents
“The advantage of being at warps layer is like you get the command line interface, but we're the outer app. And so we can do things with the developer experience in the UX. Like we can have editing features where we think it's appropriate. We can build like a c…”
Gil: Warp reaches near 1M MAUs, adds $1M revenue every 7–10 days
“Like, so you're close to a million MAUs. You're doing something like a million in new revenue every seven to 10 days.”
Lloyd: Warp's growth inflected after launching a coding agent
“And so we really started to inflect when we launched a great coding agent, which was like three, four months ago, honestly.”
Lloyd: AI agents struggle far more on pro codebases than web apps
“Like, I think it's relatively straightforward at this point for a good agent to like, you know, with relatively few prompts, build like a web app. It's much harder to apply these agents successfully to pro code bases.”
Lloyd: Within years, all developers will work by prompt with some automation
“Like, I think within a couple of years, you'll have everyone working by prompt and you'll have some slice of development tasks that are just like in the background, like a server error comes in or a new ticket, a user report comes in something Something is aut…”
Lloyd: Engineering expertise is not devalued and is more important than ever
“What I really don't think though is like engineering expertise is going to become devalued. So I think it's actually in the short term, at least it's more important to know what you're doing as an engineer than it ever has been.”
Lloyd: Perpetual junior engineering roles are at risk from AI agents
“Where you don't want to be, I think is like someone who is just like perpetually in the junior engineer state. Cause I do think that's at risk.”
Lloyd: Memory-safe languages and automated verification will become more valuable
“I think anything that does like either automatic security analysis or automatic verification, I think actually like languages like rust, things that have stronger guarantees around safety by default, where you don't need to rely on like a human reviewer become…”
Lloyd: Front-end AI builders will bundle into all-in-one business platforms
“I think there's going to be a class of tools where you sort of start from the front end. This would be things like Lovable or Bolt or Replit or maybe even Figma Make if you're coming from the design side, and you'll have like an all-in-one platform for Build a…”
Lloyd: Agentic code review will be bundled into broader dev platforms
“Some things that are vertical right now, like, I actually, I think will, like, for instance, like agentic code review. Take that. To me, that's Should be part of a like holistic agentic development platform, not so much a standalone thing.”
Lloyd: GitHub Is Not Executing Well on AI Coding Front Door
“The other front door, which. I feel like honestly they're not executing that great is GitHub where all the source code lives.”
Lloyd: Frontier Models Have Pricing Power Due to Open Source Delta
“Right now the models have a bunch of sort of like pricing power because like there is a real delta between using the frontier model and using like the open source model.”
Lloyd: OpenAI Will Subsume Best ChatGPT Apps Using Platform Data
“I don't know if you saw OpenAI Dev Day yesterday, but they're clearly doing this platform play within ChatGPT now where it's like you have apps within ChatGPT. I'm sure they'll use that data to Sort of subsume or take over whatever the best first party integra…”
Lloyd: Warp pivoted from developer collaboration to an AI agent platform
“The business concept has evolved a ton. Like the original business concept was like building a collaboration platform, which is like, We've just changed our model to be an agent platform because it's like way more demand for that than a collaboration platform …”
Zach Lloyd: Context, Not Reasoning, Is the Main Limit for AI Coding
“I think that the, increasingly the limit that we see is context. And like the reasoning capabilities of the models are pretty impressive. The problem is like, Understanding an entire code base or understanding sources outside of the code or literally just unde…”
Lloyd: Upgrading Anthropic models gave Warp only modest SWE-bench gains
“Like, if you take, like, Sonnet four to four five, and we're big partners with Anthropic, they have great models, like, that was, like, a few percentage point increase on Sweebench for us. And we invest, you know, we've invested a decent amount to be one of th…”
Zach Lloyd: Full automation delivers clearer business ROI than developer productivity enhancements
“Just like automation is a better place to be than productivity enhancement. Like one of the challenges with our business, I think with a lot of the coding businesses is just like proving the ROI. Like, and there've been these studies that show like, you know, …”