Absolutely — here’s what I extracted.
Video summary (YouTube: PQU9o_5rHC4)
In this Light Cone interview, Garry Tan talks with Boris Cherny (Anthropic’s creator of Claude Code) about how Claude Code emerged from an internal experiment into a major tool and why its CLI-first design is still central despite all the IDE-heavy momentum.
Main points Garry and Boris cover:
-
Origins as an internal, practical tool
- Boric built it to solve real internal workflow issues, and it gained traction before broad launch because it was immediately useful.
-
Terminal-first is intentional
- Garry repeatedly probes why a terminal UI worked so well: speed, flexibility, and surprisingly high usability.
-
“Latent demand” and Plan Mode
- Garry pushes that features like Plan Mode likely succeeded because users were already doing that behavior outside the product.
-
AI changes “how to code”
- Repeatedly discussed: humans should move from manual coding toward higher-level architecture, planning, system thinking, and product-level judgment.
-
Hiring and team implications
- Garry speculates on evaluating engineers by how well they work with agents; the model becomes a universal productivity layer and teams/roles shift.
-
Explosive productivity claims
- Garry calls out claims like 1000x productivity framing at Anthropic and discusses how software engineering itself may be redefined.
Every Garry Tan quote (timestamped)
Timestamps are from the extracted transcript; wording is transcript-style and may include minor ASR artifacts.
- 00:27 — “Are you also in the back of your mind thinking that maybe like in 6 months you won't need to prompt that explicitly? Like the model will just be good enough to figure out on its own?”
- 00:36 — “Oh my god.”
- 00:46 — “Welcome to another episode of the Light Cone and today we have an extremely special guest, Boris Cherny, the creator engineer of Claude Code. Boris, thanks for joining us.”
- 00:57 — “Thanks for creating a thing that has taken away my sleep for about three weeks straight. I am very addicted to Claude Code and it feels like rocket boosters. Has it felt like this for people, like for, you know, months at this point?”
- 01:45 — “If you look back on those moments to now, like what would be like the most surprising thing about this moment right now?”
- 02:38 — “Going back though, but when do you remember when you first got the idea? Can you just talk us through that? Like was it some like a spark or what was even the first version of it in your mind?”
- 04:24 — “You built it in terminal just because it was the easiest way to get something up and running?”
- 04:31 — “It was just me at that point. It was like the IDEs Cursor, Windsurf taking off. Were you sort of under any pressure or getting lots of suggestions of hey like we should build this out as a plugin or as a fully featured IDE itself?”
- 05:22 — “Oh my god.”
- 05:39 — “That’s kind of fascinating. I mean, it’s very kind of contrarian that Claude Code works so well in such an elegant, simple form factor. I mean, terminals have been around for a really long time, and that seemed to be like a good design constraint that allowed a lot of interesting developer experiences. Like, it doesn’t feel like working; it just feels fun as a developer. I don’t think about files, where everything is—and that came by accident, almost?”
- 07:09 — “So in that 2024 period, what—how were the engineers using it? Were they sort of shipping code with it yet, or were they using it in a different way?”
- 08:49 — “Earlier we were saying like we should compare
.claudemds, but you said something very profound, which is, you know, yours is actually very short, which is almost like the opposite of what people might expect. Why is that?”
- 09:51 — “Do you have to like compact the
.claudemd? Like, I definitely reached a point where I got the message at the top saying your .claudemd is like thousands of tokens now. What do you do when you guys hit that?”
- 10:43 — “Wait, really? I would have assumed that because you built this in the terminal that you were sort of like a die-hard terminal like Vim-only person. You know, screw those VS code people, you know?”
- 11:28 — “How do you decide how verbose you want like sort of the terminal to be? … how do you make those sorts of decisions?”
- 11:46 — “What’s your opinion, is it too verbose right now? Oh I love the verbosity because basically sometimes it just goes off the deep end and I’m watching and then I can just read very quickly and it’s like, ‘Oh no no, it’s not that.’”
- 13:32 — “I’m amazed like how much I enjoy fixing bugs now. And then all you have to do is have really good logging and then even just say like, ‘Hey, check out that particular object, it messed up in this way,’ and it searches the log, it figures everything out.”
- 13:53 — “There’s a totally different school of thought now that says like anytime a real human being has to look at code, that’s bad.”
- 15:40 — “So what would be some advice for technical founders to really become maximalists at the latest model release? It sounds like people fresh off of school or that don’t have any assumptions might be better suited than maybe sometimes engineers who have been working at it for a long time. And how do the experts get better?”
- 16:39 — “How do you screen for that when you try to hire someone now for your team?”
- 17:34 — “Do you think you would ever hire someone based on the Claude Code transcript of them working with the agent? Because we're actively doing that right now. We just added, just as a test, like you can upload a transcript of you coding a feature with Claude Code or Codex or whatever it is.”
- 17:53 — “Personally, I think that like it’s going to work. I mean, you can figure out how someone thinks — whether they’re looking at logs or not — can they correct the agent if it goes off the rails, do they use plan mode? When they use plan mode, do they make sure there are tests? … Do they think about systems? Do they even understand systems?”
- 18:14 — “I just want like a spiderweb graph, you know, like in those video games like NBA 2K, it’s like, ‘Oh, this person’s really good at shooting or defense.’ It’s like you could imagine a spiderweb graph of someone’s Claude Code skill level.”
- 18:30 — “Yeah, what would the skills be? What would be those? I mean, I think it’s like systems, testing, must be like user behavior. I mean, there’s got to be a design part, like product sense, maybe also just like automating stuff.”
- 19:42 — “That’s the litmus test, yeah.”
- 20:17 — “Yet, you know, like we use the Quad agents SDK to automate pretty much every part of development. It automates code review, security review, it labels all our issues, it shepherds things to production—it does pretty much everything for us.”
- 20:41 — “One of the funnier things that I’ve been having office hours with founders is… visionary founder vs. engineers, where the founder can do 50x and the engineers do maybe 5x… Is this what you’re seeing?”
- 21:59 — “What’s the vision for Claude Teams? Just collaboration?”
- 22:52 — “How did you set that up? Like did you spec out the outcome you were hoping for and then let it run?”
- 23:17 — “The main
Quad just gave it instructions and they all just figured it out like independent agents that didn’t have context of the bigger spec?”
- 23:46 — “My Claude Insights just told me to do this more for debugging… it would be better to have like multiple sub-agents spin up and debug in parallel… so I added that into my Claude MD: next time you fix a bug, have one agent look in the logs and another in the code path.”
- 24:38 — “I’m curious, so then why don’t you put that in your Claude MD file?”
- 24:55 — “Are you also in the back of your mind thinking that maybe like in six months you won’t need to prompt that explicitly? Like the model will just be good enough to figure out on its own?”
- 25:06 — “Oh my god.”
- 25:11 — “What would the world look like without plan mode? Do you just describe it at the prompt level and it would just do it one-shot?”
- 25:46 — “Yeah, so it sounds like a lot of the feature development for Claude Code is very much what we talk about at YC — talk to your users, then implement it. It wasn’t the other way around.”
- 26:31 — “So do you mean there will be no need for plan mode? I’m worried the model’s going to do the wrong thing or go in the wrong direction.”
- 27:51 — “The next step is Claude just speaks to your users directly.”
- 27:56 — “Yeah, it just bypasses you entirely.”
- 28:07 — “No way.”
- 28:13 — “What does it want to tweet about?”
- 28:38 — “What are some tips for founders now on how to build for the future? Sounds like everything is really changing. What are principles that stay and what changes?”
- 29:28 — “That’s what you mean by Plan Mode was latent demand… people were already talking to Claude to figure out the spec, and now it’s just in Claude Code.”
- 30:02 — “It seems like we’re surprised how far the terminal has gone and how far it’s been pushed… is there going to be a new need for a different UI on top of it?”
- 30:58 — “What about your advice to DevTool founders? Should they build for engineers and humans, or should they think about what Claude is going to think and want?”
- 32:09 — “Back in the day more than 10 years ago you were a very heavy user and you wrote a book about TypeScript right before TypeScript was cool.”
- 32:27 — “Before TypeScript was a thing, it was this weird language thing—and now it’s way more practical.”
- 33:04 — “It was like very strongly typed.”
- 33:38 — “That’s right, that’s right, that’s right.”
- 34:29 — “Now fast forward more than 15 years later not many codebases are in Haskell, which is more academic, and there are tons now on TypeScript because it’s way more practical.”
- 34:46 — “The terminal is one of the most beautiful terminal apps out there and is written with React terminal.”
- 35:04 — “I do like design and user research and you know write code and all this stuff and we love hiring engineers that are like this.”
- 35:50 — “Oh how do you do that in Cloud Code? I’ve been trying to figure out how to do this.”
- 36:13 — “Yeah, it feels like BBS’s, it’s like a BBS door game… that’s a great compliment.”
- 36:19 — “Yeah yeah, like it should feel like you’re discovering Lord of the Red Dragon. It’s fantastic.”
- 39:15 — “How often do you rewrite the codebase of Claude Code? Is it every six months? Is there scaffolding deleted as the model improves?”
- 40:01 — “Do you see Steve Yegge’s post… one of Anthropic’s engineers currently averages 1,000x more productivity than a Google engineer at peak.”
- 40:02 — “It’s an insane number. Like, 1,000x…”
- 40:24 — “Three years ago we were still talking about 10x engineers; now we’re talking about 1000x on top of a Google engineer in their prime.”
- 41:37 — “What drove you to come over to Anthropic? As a builder you could go anywhere. What was the moment that made you say this is the approach?”
- 43:37 — “What is gonna happen this year?”
- 44:46 — “So continuing to trace the exponential, I think what will happen is coding will be generally solved for everyone.”
- 45:02 — “I think we’re going to start to see the title software engineer go away and it’s just going to be maybe builder, maybe product manager; maybe the title becomes vestigial.”
- 45:08 — “Software engineers are also going to be writing specs, they’re going to be talking to users… right now on our team engineers are very much generalists.”
- 45:16 — “Every single function on our team codes — our PMs code, our designers code, our EMs code, our finance guy codes.”
- 46:32 — “Well, actually, for all of December I was traveling around, and I took a coding vacation.”
- 46:42 — “So we were kind of traveling around, and I was just like coding every day.”
- 46:44 — “That was really nice, and then I also started using Twitter at the time, because I worked on Threads back then.”
- 46:50 — “I’ve been a Threads user for a while, so I tried to see where people were.”
- 46:56 — “Yeah, I think for a lot of people… that was the moment they discovered Opus 4.5.”
- 47:01 — “I kind of already knew.”
- 47:03 — “Internally, Claude Code has been on this exponential tear for many, many months.”
- 47:09 — “That’s what we saw.”
- 47:11 — “There was a stat from Mercury that like 70% of startups are choosing Claude as their model of choice.”
- 47:19 — “There was another stat from SemiAnalysis: 4% of all public commits are made by Claude Code.”
- 47:26 — “From biggest companies to the smallest startups, people are using it.”
- 47:31 — “It even plotted the course for Perseverance—like the Mars rover.”
- 47:36 — “This is just like this is the coolest thing for me.”
- 47:38 — “We even printed posters because it felt wild that NASA chose to use this thing.”
- 47:44 — “So yeah, it’s humbling, but it also feels like the very beginning.”
- 47:49 — “What’s the sort of interaction between Claude Code and then Co-work?”
- 47:53 — “Was it a fork of Claude Code…?”
- 47:55 — “Did it have Claude Code look at itself and make a new spec for non-technical people that keeps all the lessons, then run for a couple days?”
- 48:07 — “What’s the genesis of that and where do you think that goes?”
- 49:33 — “Boris, thank you so much for making something that is taking away all my sleep, but in return it’s making me feel creator mode again, sort of founder mode again.”
- 49:43 — “It’s been an exhilarating three weeks.”
- 49:45 — “I can’t believe I waited that long since November to actually get into it.”
- 49:49 — “Thank you so much for being with us, thank you for building what you’re building.”
- 49:53 — “Sounds good. Come on now.”
If you want, I can also give you this as a clean CSV/JSON with time_s, timecode, and quote fields.
