What Happens After Coding Is Solved? | Fiona Fung  🔗 **Source:** [What Happens After Coding Is Solved? | Fiona Fung](https://share.snipd.com/episode/0d25a096-4170-45a3-bf90-4af5156e4c77) ## Highlights - **Coding Is No Longer The Bottleneck** - Coding throughput exploded at Anthropic: engineers ship ~8x more code per quarter versus 2025. - Fiona says coding is no longer the bottleneck; the constraint shifted to ambition, verification, and product thinking. Transcript: Lenny Rachitsky Anthropic engineers on average have eight times as much code per quarter as they did compared to 2025. Coding is no longer the bottleneck. Fiona Fung It's left at the ceiling of what anyone is able to do. Lenny Rachitsky Everything is now possible in theory. Now it's about how ambitious can you be? Fiona Fung It's always something we ask ourselves, what's better than me doing it? I haven't thought, David. Lenny Rachitsky The people that seem to be doing best are taking the most initiative, getting the most proactive, have the most agency. Fiona Fung We say with high agency, it's also high accountability. So it's all about making sure folks have that freedom to cook. But then it's also like, okay, what's the accountability for it? What's a hypothesis of what you're trying to solve? Lenny Rachitsky I'm curious what is lost in this new world of software engineering. I ([Time 0:00:00](https://share.snipd.com/snip/3b697997-55da-4318-8765-720d4b22761a)) - **Have AI Run Your Monthly Product Retrospectives** - Use AI as a manager to run regular retrospectives: Fiona runs Cloud Code remote sessions that view repos, Slack feedback, and metrics to review shipped work. - She shares the routine live with engineers to discuss impact, bugs, and next investments rather than only counting commits. Transcript: Fiona Fung The roles are blowing, it's shifting more to this builder. Like everybody starts being a builder, I would say. The other shift that I've recently done is I actually have a cloud code remote session that I enlist in all of our repos. And so this way I have full visibility into the work that everybody's doing. And this instance, it also has access to all our Slack channels and, and I'll, I'll have access to like, how, how are the metrics of everything we track? And so every, every month, like I went, when, like, I be like, Hey, you know, what's fun? Like, let's, let's take a look back. Like, and so we'll actually do it together. I'll share my, you know, like cloud code, you know, I'll share my screen and we do a cloud code session. And it's just about, hey, what were the focus areas? Like what were some of the products that got shipped? How did it do? Oh, what were the feedback channels? And so I, even though like before I would have just used these sessions to like generate PRs and bug fixes, I actually have these sessions to enable me to have conversations with folks That I support. Lenny Rachitsky Say more about that. So this is like a management technique, let's say, to help people not just ship, but actually ship better, understand if they're shipping things that have impact. Is that kind of the idea here? Just like use Claw to keep on top of all the things people are shipping and then make that a conversation with them? Exactly. Fiona Fung So like, yeah, outside of just, you know, the action of shipping is how did it do in market or, Hey, did we have, you know, like, did we cause some bugs and it's okay to like, I have the saying, Make new mistakes. Like it's okay to make mistakes, just make new ones so that we're always learning. Cause if you aim to make zero mistakes, like that probably means you're not, you know, moving fast enough or being a little bit too cautious. And so, yeah, by having Claude, like, so then they can also look at, like, for example, actually, yeah, I was just looking at, oh, you know, given some of the, you know, incidents that We have, like, let's look across all of this. Can we generate a theme? Like, what's a good area investment for us next, especially we think about, like, quality? Like, are we seeing any hot spots of where there could be a gap? I think that used to be just a much more manual process, I would say. Like if I look back a year ago, I don't think I would have been able to, you know, have some of these insights with Claude. ([Time 0:10:03](https://share.snipd.com/snip/51894d22-60a2-4188-8b9c-7bc7d0026cb7)) - **Automate Morning Feedback With Routines** - Automate your daily monitoring with routines: Fiona uses routines to summarize feedback channels, surface themes, and even produce PRs to review each morning. - Routines spawn agents that continuously scan Slack, emails, and metrics so managers wake to actionable items. Transcript: Fiona Fung And that's also completely changed. Like now I just have a routine that automates all this for me. And then there's also, it's almost like before I would, you know, be able to kind of like, you know, generate some prompts. But now with routines, it's almost like I'm, you know, having an agent help me generate the prompts and the PR. So, for example, one is, hey, keep a look on this feedback channel. You know, what are some of the themes? And then like when I wake up, then, you know, I have a really good summary of that. ([Time 0:13:08](https://share.snipd.com/snip/dd4617ef-b751-4f84-86b5-c62e5909928b)) - **Shift From Writing Code To Verifying Expectations** - Verification became the critical problem as throughput rose: automated checks and explicit 'what good looks like' frameworks let Claude validate PRs at scale. - Fiona recommends checking specs into repos and keeping them updated so models can enforce the intended behavior. Transcript: Fiona Fung And honestly, it's crazy when, you know, I think we didn't even have Claude code reviews last year. And so speak of bottlenecks, that was a really, really big bottleneck of, you know, the human reviewers. So we definitely, for the important, like areas that need deep subject matter expertise, we definitely want to make sure we have the proper, you know, like humans still reviewing. I would say what helps us, though, is the more that we can automate to almost check in the framework for what good looks like. Cloud is very good when you give it a framework to validate against those frameworks. So I had mentioned like, you know, or like, you know, like right recently we just updated like content design to have a skill in it. Like my, like I, and this is why like, I think if you have specs or like, or like, like check those into the repo and then make sure the spec also keeps up to date with the code like frequently. But that's what I found like works really well of any time you have like a statement of what good looks like, get that into the repo and then code review can make sure it's still matching What you set up to do. ([Time 0:14:55](https://share.snipd.com/snip/9bbfb38b-b7e8-4d28-889d-209c9d1b805d)) - **Hire Product Builders And Deep Systems Experts** - Hire two complementary profiles: prioritize creative product-minded builders plus a smaller set of deep systems experts for trust-critical areas. - Fiona expanded Cloud Code with systems hires when they lacked distributed-systems expertise for sensitive components. Transcript: Fiona Fung Yeah, the deep subject matter expertise. Like, for example, when I first joined CloudCode, we had really great kind of like product generalists. And then I realized, oh, we were missing folks with systems background. And so that was definitely an area that we needed more folks with kind of like systems and distributed systems expertise. And so I would say whatever are the parts that, it's all about trust, but verify. The models are really good, but there are definitely a lot of areas that still need, you know, the verification. And so wherever you need the deep subject matter expertise, I would say that's, you know, an area to definitely still invest in. And the other one is kind of like the product, the product sounds almost like the dreamers. Like these are folks that usually will be like, my gosh, you're really passionate about a product and they have an idea, they build it. And then it's always like looking at the feedback and then iterating and polishing and making sure that the product is a delightful experience. Like owning that product end to end, that's like another skill set that sort of does really well on CloudCode. ([Time 0:17:18](https://share.snipd.com/snip/593ddfcb-eacd-47fc-b165-b000b1036bbe)) - **Ambition Becomes The New Constraint** - Ambition replaced engineering constraints: with AI doing much of implementation, engineers' upper bound is now how big they can think. - Fiona shares an example of a non-mobile engineer shipping mobile features by partnering with Claude. Transcript: Lenny Rachitsky There's this 10XC engineer I was talking to the other day, and he was just like, I used to, like, I heard about a feature idea. Like, someone's like, hey, we should build this. And I was like, no, that's really hard and complicated. And he's like, but now I'm like, no, that's so possible I just ask Claude Co to do it and it just does it and it's this whole mindset shift and he's just realizing now it's about how ambitious Can you be like everything is now possible in theory now it's about how ambitious and how big can you think versus just like, okay, it's all these stupid little features and things they Have to unblock. Does that resonate? Is that something that you think about? Fiona Fung Yes. Actually, I was just catching up with an engineer yesterday, and he actually is not a mobile engineer by trade, but we really needed to update this feature to also have a mobile footprint. And it was just amazing. Like he's like, you know what? And it's common because you might think, wait, but I'm not like an Android expert. But now, thanks to Claude, actually, I can actually have a partner and actually also do this on the mobile surfaces. And so that definitely resonates. It's like it's left at the ceiling of what anyone is able to do. ([Time 0:18:30](https://share.snipd.com/snip/cc2c4a48-334f-4833-a604-c9d5f53d8966)) - **Lean In With A Growth Mindset** - Cultivate a growth mindset and lean into curiosity: Fiona tells those afraid of AI to ask what's within their control and take actions to learn the tools. - She suggests doing one risky learning move to force growth and avoid stagnation. Transcript: Fiona Fung Would say a growth mindset really, really helps. Like actually even before AI tooling, I found that has been just so valuable. And I learned that a lot. Actually, it was the shift from Microsoft to Meta. That was where I first realized that first year. I'm like, oh, this is what having a growth mindset really means. It's really this concept of always be learning. Also, what served you to get you to this point may not, may serve you no longer. But it's really hard because, of course, everybody, like we all, you know, we've all gotten to the state by acting or, you know, like operating in a certain way. And so sometimes it is a little bit scary to think, wait, you're like, but I've been successful so far. You're asking me to change what has made me successful. And so I would say like the growth mindset is really, really has really served me well. And I think it's also served others well that I noticed, like always leaning in with curiosity and always being able to learn um and then in terms of like the frustration i think sometimes I also see a little bit of fear and so my advice there is at least this is how just in life i've because we all have and fear is of course an evolutionary like i i didn't study anthropology But i think it you know, it makes sense, right? It helps us make sure we were able to survive and not get eaten and, you know, by larger predators. But for anything that, you know, that there is a fear, my advice is kind of lean in and ask, okay, is there some, what can I do about it? What is within my control? Because sometimes the frustration comes from fear and feeling like, but everything is outside of my control. And so it's happening to me. And so if you think about, okay, what is in your control? Instead of happening to you, is it happening for you? And then what's something that you might be able to kind of like do and change? ([Time 0:20:04](https://share.snipd.com/snip/59c80552-6734-4c18-8ca5-895c37928b4a)) - **Dogfooding Led To Cowork For Small Business** - Fiona dogfooded Cowork while expense-traveling and onboarded small-business friends, which revealed unique use cases and bugs. - A restaurant owner used Cowork to index menus and run market price comparisons, inspiring the Small Business product bundle. Transcript: Fiona Fung And then how it happened was I was using co-work for my own business expensing travel. I don't know what it is. I really don't like doing business expensing. And when I was using co-work, it was magic. I'm like, oh, all these things that I don't like to do, like co-work is doing for me. And I'm like, wait a minute, all my friends, and honestly, small business owners, they really bust their butt. Like they work incredibly hard. And they're usually, you know, like sometimes might be operating on really small margins. And so I thought if it's, if Claude Cowork is so good at helping me do my little, you know, business travel expense, this would be huge for it. Cause I see my friends sometimes sitting at the bar with stacks of bills and all they're doing is invoicing and expensing. And I think nobody actually really likes to do it. So that's kind of where it came from. And then, so yeah, I remembered helping a couple of them onboard. And it was also very humbling to see them go through our onboarding flow. Actually, it found some really good bugs. So it was really like a win-win. But it was also delightful for me because they also use co-work in ways that I wouldn't have thought about. Because I was all fixated on, look at how it is amazing with PDFs and invoice. And then, you know, one of my friends who runs two restaurants, she's like, oh my gosh, this folder I have is like a junk drawer of like, it's basically our, you know, documents folder Just becomes this junk drawer of every, or downloads, like just, you know, like the kitchen junk drawer. And she's like, I know I have a few menus in here and I can't find it. And I'm like, well, let's ask co-work. So we gave co-work access to the directory, found the menus. And then she's in a really unique way. She goes like, I want to make sure I keep my prices reasonable for locals and tourists. So she goes, hey, Claude, look across my style of cuisine in this area. Is it comparable? And it came back with really cool, almost like market analysis. And she goes, hey, I actually just went to that restaurant in Seattle and that was pretty good. Um, so I learned something every time and, uh, and yeah, they've been like giving me great feedback as well. ([Time 0:27:52](https://share.snipd.com/snip/e5fcf0b6-a22a-4dec-824b-7fd6f81559b9)) - **Build For Latent Demand You Observe In The Wild** - Spotting latent demand matters: Fiona watches unexpected behaviors (non-coders using Cloud Code) and builds tailored experiences (Small Business toggle) to convert friction into product features. - Rapid feedback and internal usage accelerated these discoveries. Transcript: Fiona Fung I haven't worked in any other labs, so I'm not sure how our labs operate. But I will share like, yeah, with and like actually on the Cloud Code team as well and co-work, like we also keep an eye on like latent demand. Now, we're very fortunate with coding as a use case because so many ants that we were, you know, like our own first customers and we're able to do like really rapid feedback. And so I think, but latent demand has been like, for example, a co-work we know to say a lot of folks that were not necessarily coders were using cloud code. Can we make that experience better? That's actually even outside of Anthropic, that served me well in all the different products I've worked on. And actually, it's funny you mentioned the cloud for my passion with small business. After I had a few of those visits, we ended up now launching Cloud for Small Business, which is really cool. And it totally didn't come for me, so I'm not taking credit. But I noticed it too when I was working with it because they were asking me, oh, does it have this plugin or this plugin? I'm like, I think it does. And then we have to, you know, go and search for it. So actually now Cloud with Small Business bundles it all up. So inside cowork, you just have this little toggle. And there was a wonderful team that's been kind of like, you know, doing cowork sessions with Small Business that probably found, hey, we could have like an efficiency gain here, like Make the experience better. And so I would say like always not only for products that you're on, like making sure you're always listening to feedback and always iterating, trying to make a delightful, reliable, High quality experience, but then also keeping an eye out for, oh, what are these other use cases that are popping up? ([Time 0:32:43](https://share.snipd.com/snip/340d7545-9fd5-4d01-bd89-24577197b02d)) - **The Next Frontier Is Asynchronous Agent Fleets** - Work is moving toward asynchronous agent orchestration: routines spawn agents to run checks and produce artifacts asynchronously, raising the abstraction above single prompts. - Fiona predicts fleets of agents and JIT routines will change day-to-day engineering workflows. Transcript: Fiona Fung Would say we're shifting more towards async, like asynchronous. And so to your point, the fleets of agent, like that, that's why routines are so interesting. Cause almost like I used to be like, you know, doing a prompt and synchronously, and then I would like maybe like kick off different props asynchronously, but now I can actually have A routine that actually generates, you know, these, these prompts for me. So it's almost like the level of abstraction keeps pulling up a little bit. And so I would, yeah, I wish I have a crystal ball to see what this is going to look like. It'll be fun actually for you and me to revisit a year from now. Lenny Rachitsky No, I want to hear more about this. This is, so help us understand what is, what is routines? And then what is, when you say synchronous, you're writing a prompt and it just kind of goes off and is it writing it immediately? Talk about what this actually looks like. Fiona Fung Ah, yeah. So, uh, routines is like how you can, like, so if you remember, I was trying about my, what I used to do is always like, you know, wake up with my morning cup of coffee and Hey, look at through This Slack channel for me. But then now like I have a routine that I set to run every morning at a certain time to say, Hey, that's right. But then it's able to actually go and kick off agents on your behalf. And so that's because like, that's what like, and, you know, even, you know, before it would be, oh, you could like, yeah, do a cron job to automate like, but now it's, hey, look at these Feedback. And then if you're seeing some of these bugs that what are some polish fixes that you might be able to knock out and then it would like go kick off and then I wake up and I ended up having PRs That I could review versus before, which is still more of a, I can have different agents. And then it's still, I'm still thinking about, okay, now what do I do with this information? So it's like that higher level abstraction of, okay, now how can I actually write a routine that also basically does prompts for me for spawning different agents. ([Time 0:35:36](https://share.snipd.com/snip/dc4d9601-4633-43f4-8b8e-90989c1cd23e)) - **Measure Outcome Over Token Or Code Counts** - Measure outcomes not raw AI activity: Fiona warns against token-maxing or lines-of-code metrics and urges focusing on whether output delivers the desired product outcome. - She recommends listening tours across senior and junior engineers to capture qualitative value from AI tooling adoption. Transcript: Fiona Fung This end productivity is such a fascinating topic. Yeah, because I remember, you know, like even Boris mentioned, like, you know, we'll first start with like lines of code. And then that's throughput. And then I remembered once then there was a debate of, well, lines of code this engineer had this crazy lines of code but they just took some library and then just was porting it so they Checked it in and then I was like well maybe it's significant lines of code and then it's well but what if we're you know updating our frameworks and now we're generating less code but Like the output's still the same so now I was like okay maybe it's it's like time to land PR like but but very interesting if it's always been whichever metric, it's like if you're really Focused on the, like my advice here is, one is output, like is the output really going towards the outcome? Because yeah, like the token maxing, it's kind of like, it's almost like the line of code that we used to have, but I'm really much more about a, what is it that we're trying to do? Like all of this, you know, there was another saying, I really, like don't forsake motion for progress. Because if you're measuring like, you know, like tool user usage, then you're measuring the action, but is it really making whatever the end outcome of yours important? And so I really try to zoom out and focus on what is the problem we're trying to solve? What's a good way to measure that? And then that's what we focus mostly on versus kind of like the productivity measurement. I would say, though, like outside of metrics, especially if on teams, as you're having more people adopt AI tooling, I would say definitely go on a listening tour for the leaders listening To a podcast, especially, you know, I love to focus on the senior and juniors as well, like hear from them on what's working, what's not, how can we actually make it better? Because they will also help you multiply and scale that out across a full engineering team. And sometimes it's those conversations where you might get spark an idea that comes up and also really good shared learning versus, you know, like metric dashboards. It's very interesting. I've learned some good lessons for metrics. I'm all about like, it's really great when you have a metric that you can actually hill climb on, but always keep in mind of is, you know, kind of my whole growth mindset of, is it still serving You? Always keep in mind, like, is that metric really still serving the outcome that you were aiming for? ([Time 0:40:28](https://share.snipd.com/snip/26f266c6-5d9c-4eb0-91cc-3007fc89f844)) - **Classify Issues As Bad Or Sad To Prioritize Quality** - Define proactive quality thresholds and triage them: Fiona introduced 'bad versus sad' where bad is irrecoverable (crashes) and sad is recoverable pain points (flicker), letting teams prioritize fixes. - Each team declares its bad/sad definitions for their surface area (e.g., CLI crash rates = bad). Transcript: Fiona Fung I would say, and this is what honestly we want to keep doing more of and being better at it too, like the proactive quality. So especially for quality, making sure that what are the experiences that are key and making sure you actually, you know, actually speaking of metrics, those are really good things That you make sure you can like kind of like see trends over time. And so like on the quality front, we found like, and this is like the more proactive we can be of like making sure we can get an earlier detection into quality. And so like, that's been one thing that we've been paying a lot of attention to. Like, so like, you know, I started this, hey, let's have a concept of what's bad versus what's sad. And bad is like a very bad, irrecoverable error. And sad is something that's kind of like a pain point recoverable. But it's interesting when you stack up sads, it could, you know, generally go to bad. But even having a, like starting with a high level framework like that, and because if not, like I think sometimes with dashboards, you can have, you know, like time to load or all these Other, but when you're dealing with a lot of different product services, it's harder to go, wait, is that a good number or not a good number? And so one thing that's helped us is versus just raw, you know, like performance or, you know, like reliability numbers. ([Time 0:44:47](https://share.snipd.com/snip/266072b7-59dc-4426-8bba-919ec4de68e1)) - **Make Managers Ship Code Part Time** - Require managers to spend maker time as ICs: Fiona asks new managers to start as ICs or continue coding part-time so they stay connected to the codebase and product. - This practice builds rapport, helps leaders dogfood the product, and prevents losing touch with technical realities. Transcript: Lenny Rachitsky Talk about that, why that's so important in today's world. Fiona Fung I love it. It's funny when I first joined, because I have amazing recruiting partners, and I noticed we were, you know, this whole theme of kind of growth mindset, it's just because the landscape Is changing so fast. Like what worked well before, like may not make sense. And even what makes sense today may need to change tomorrow, right? Like that's what I have to keep reminding myself. So yeah, when I first joined Recruiters, like, okay, yeah, we have these couple like manager postings. And I'm like, you know, because actually how it came from like a listening tour I did with like all the members on the team. And I heard a lot of these, hey, I really appreciate the agency, but how can I make sure prioritization? And so I kind of through that, I, I, I realized, and then there was some really good feedback too, of making sure that it's not too many layers of reviews. Like there was good feedback. Some folks have might've joined from other companies or like, you know, and then I'm like, Hmm, when I actually think as a, as a leader, if you actually start as IC first without the kind Of like worry of supporting people, because that's a very heavy responsibility that I think matches. But before you have to take on that full responsibility, give yourself that maker time to actually tap deep into the code and learn the code base or the product, whatever it is. It doesn't have to, honestly, the PRs I do are like, but it's more about for me, it's important because it keeps me in the flow because we're making so many changes to co-work and code. So even me doing PRs, it's less about what it is I'm fixing. It's more about me using the, the priority date just to keep that touch. Cause as amazing as metrics and everything are, and I do look at those dashboards, you as a leader if if you're not you know like living and breathing your your product every day you sometimes Kind of like lose touch of the touch and feel of the product but anyways on the manager front I think giving time for managers that join to be able to go deep and to do that before supporting People and then they actually like end up building really great rapport with the team. Like, because if not, I think sometimes as managers, you know, you might come and join a new team and you instantly think, I have to manage, let me dig into my manager toolbox and do managery Things. But if you actually give yourself time to not have to worry about that first and actually learn what it's like to be an engineer and teammate on the team, that also goes really far to building Rapport. In terms of using the product, I think that's important. ([Time 0:49:57](https://share.snipd.com/snip/ba32b0a6-ad5a-4286-8123-7b6af65db3ec)) - **Combat Agent-Induced Isolation With Pair Programming** - Agents can make individual work lonely; deliberate social structures matter: Cloud Code started pairwise programming lunches and hackathons to recreate collaborative learning. - Watching peers use tools revealed diverse flows and transmitted tacit knowledge faster than docs. Transcript: Fiona Fung Interesting because we actually have this discussion on the team a bit. Like, are there things that we miss? Or like, actually, to your point, I'm always like thinking about like when someone's onboarding and ramping, like you have Boris who hand-rolled the code in the early days and of course Now does anymore. But he gained that knowledge from before because he was in the code days. So for especially all the engineers that join, I'm also like, make sure you're like taking the time too to all the work that you do, still get the understanding of the architecture or The change because it goes back to that trust, but verify. Maybe one day it won't matter anymore, but at the pace that we're going, I actually still think understand, like it's always double clicking on that layer that you depend on. Like maybe that to me is a matter, like it's always take that time to learn about your dependencies because this way, when your dependencies change, like you're kind of like more aware Or you might not be taking advantage of changes to, you know, dependencies. So I think it's always about kind of like doing that double click. But the other thing that we found interesting on the CloudCo team is after a while we felt it could start being a lonely experience because we all started just working with our agents So much. So recently we started maybe like a pairwise programming lunch because what we also learned was on Cloud Code, everybody uses a cloud code to work. Everybody uses a flow so differently. And so we found that, wow, when we do pairwise programming, we actually learn so much from each other. And then the other thing, too, is we make sure that we also have the maker time together, too. So, for example, hackathons is another thing we really like to do just to make sure we're kind of like interacting together as a team. ([Time 0:55:38](https://share.snipd.com/snip/8094628a-e2de-44ca-be90-799afbbda28f)) - **Automate Reviews Up To The Trust Boundary** - Build automated verification but keep experts where trust matters: push automated reviews where specs and framework checks suffice, and reserve human reviews for deep subject-matter areas. - Continuously ask if expert knowledge can be encoded and automated to scale confidence. Transcript: Fiona Fung Know, the iOS Android one, like, I think we're still, actually, it's funny because speaking of like the deep expertise, we definitely feel that it's really important. Like it's really important to still bring on folks with those expertise, but we probably don't need like as large because people are flexing. So it's, again, making sure we have, you know, like the Android and iOS experts, but then less of the larger kind of like a mobile or one per. And so, but again, that's still a balance that we're trying to figure out if do we have the right and enough expertise. The second one was, sorry, what was the second one? Lenny Rachitsky Oh yeah. It is, how far do you push fully automated review? Fiona Fung Yes. Actually, well, this is a fun one, like, you know, with the content design check, I think we're looking for across all of it, like, how do you actually get what good looks like? And so I have like, actually, the verification is still one that we, I think, and that's kind of like the second and the third, that I think we, there's a lot more opportunities there for Us. But the how far to push, I think, looking at where we think experts still matters, and then also, again, have to keep asking ourselves, okay, is there a way to leverage our expertise to Also automate? I think looking across the whole end-to experience, making sure we're not missing in. I think we come at it from an engineering standpoint, but making sure experiences, how we think those other areas. I think there's definitely still more that we can do there. Lenny Rachitsky Basically, how do you set up a verification that the experience is what you wanted it to be? ([Time 1:09:15](https://share.snipd.com/snip/83e46e5e-5e35-4535-b49d-656683717ab1)) - **Agent Fleets Create A New Context Switching Problem** - Context switching rises with many agents and async routines, creating a new flow problem: Fiona blocks focus time to catch up on agent work she spawned and notes this is unsolved. - She flags context-switch load as an open challenge as routines proliferate. Transcript: Fiona Fung Know, I think because with routines and everything being more async, I think there is starting to be a high load on our context switching. Because I even remember myself was like, oh, I kicked off. And so I think that's probably another thing we have to think about of how do we, you know, whether for team members or our users, like how can we actually make the experience better to Reduce that load? Because I do see the context switching a little bit increasing. Lenny Rachitsky Like if you have 20 agents running there's just endless checking in and reviewing and you have to remember what you're doing there it's like such an interesting world where like the Idea of flow we talked about before like engineers and most people like there's less of the just hours of flow but now the agents can kind of just remind you here's where you're at like The reset to switch almost is easier because you don't have to like relearn everything. You don't have to like re-understand the code base and the architecture. You can kind of just like, okay, but here's what we're trying to do. It's kind of like both. Got better, got worse. Fiona Fung Well, it's interesting because I used to do book out like focus time for like, you know, the focus time for coding because you want that dedicated and then the whole context switching. And then it's interesting now that because I can't context switch more with more ASIC agents, I'm noticing I do actually have to go back and block like a focus time for me to catch up on All the, you know, different ASIC work that I've kicked off. ([Time 1:11:11](https://share.snipd.com/snip/398748ff-f62a-48a2-900d-748ede8b61f6)) - **Rethink How We Train The Next Generation Of Engineers** - Education and onboarding must evolve: Fiona wonders whether apprenticeships or fellowship-style programs will replace traditional paths to teach the 'double click' systems knowledge future builders need. - She worries about how to compress deep systems understanding when many never read raw code anymore. Transcript: Fiona Fung One, speaking of like big open questions, I really do think how we grow the next generation, just because how you and me got to our engineering path is just so different. It's almost like, okay, now when you graduate from school, it's like, how do you kind of fast forward? But the important thing to me, but still understand kind of like, you know, that double click I talked about to the layer of a need. But that is a big question I have. And I wish I had the answers, but I wonder if it's for software engineering, it's almost like you go more towards a fellowship apprenticeship program. I know it's odd because technically we have like internships that would, but those were the kind of like three months and little projects. But I do wonder if, and I wish I have a crystal ball here, but it's, yeah, like how do you almost like cram in some of the, and maybe it won't matter, but you know, some of the life experiences That we all got, how do you actually enable us to kind of like teach that to, you know, the next generation of builders? Right. Lenny Rachitsky Like if you don't have to ever look at code, what's the incentive for a new software engineer to truly understand how infrastructure works and memory allocation, all these things that Are like kind of foundational. Fiona Fung And it's interesting. Maybe the malls will get good enough that it doesn't matter. You don't have to worry about it. You know, but I do think there's something about that double click because I think that's where there might be an opportunity to improve the product or the system. But figuring out how to learn that, not necessarily by, you know, like years of... Lenny Rachitsky Code. Yeah. Yeah. Like somebody's got to have to understand code at some level. It's like some COBOL engineer they have to like pull out of that retirement one day. Like, do you remember how to write Python? Fiona Fung You know, actually, what's really fun? One of my, you know, previous managers, I mean, he started software engineer when it was punch cards. And it's so fabulous. Like he's been messaging me everything he's been building with Cloud Code. And I'm like, wow, what a career that you go from. Like he's really like he, you know, he's really kind of like seeing this whole change. So maybe I think for us to think about is like, when I think about his career, how he got started with punch cards was also totally, totally different than now. And so it might be that I think it'll be interesting for us to see what remains important and then what changes in terms of importance. ([Time 1:13:24](https://share.snipd.com/snip/466e0e0e-8b7d-4784-bcf3-914b76ee33b0)) - **Treat Culture As A Living System To Guard** - Protect and cultivate culture intentionally as you scale: Fiona says culture is living fiber that shifts as teams grow and prioritizes diverse perspectives and open debates. - She pushes managers to surface issues early and avoid the 'everything's fine' trap so problems get solved collaboratively. Transcript: Fiona Fung You know, the thing that keeps me up at night is probably it's how we, so, you know, we talked about kind of like cloud code and co-work team culture. And the team culture is really important to me. Like it's the one team mentality. And I, you know, I share with folks. And by the way, culture is like a living, breathing thing. It's not just a poster you slap on a wall and it changes over time. And it's, it shows up in how we treat each other, how we, how we're there for each other. And so the culture of the team is important to me because we are, we are growing and, and, and since the culture shifts, like making sure that maintaining the things are important that We are, we still really, uh, like it's really important for me to have like diverse perspectives. So then we can have, you know, like good, healthy, open, honest debates out in the open. And we kind of like welcome those and kind of what I call that one team mentality that when you get close to the finish line, look behind you and see, is there someone from your team to help? Cause we kind of as a team. That's probably the thing that keeps me up at night. And it's, it's like, you know, there's so many other hard problems, right? But I think maybe a lot of the other ones are product or engineering challenges that yes, ([Time 1:18:06](https://share.snipd.com/snip/66cb1987-1c54-4333-a645-a71717dad7fc)) - **Remove Or Automate One Outdated Process** - Kill processes that no longer serve you: pick one manual, noisy, or expensive process and ask if it still has purpose, then stop or automate it. - Fiona changed six-month planning to lightweight monthly JIT planning and now considers automating the process to avoid tax-like upkeep. Transcript: Fiona Fung One thing is like, just a suggestion, because, you know, we talked about how can Cloud do, like, automate. Like, one other thing that's really big on Cloud Code and co-work team culture is explicit permission to kill processes that no longer serve us. And so maybe a suggestion is for, you know, any, anyone, you know, like on working on a team or leading teams, like pick your, like, what's one process that you either dread doing or is Really highly noisy or is like really expensive in terms of like just a lot of, like it might, or something that's just very manual, like pick one thing and first ask, is it still having Its purpose? Like, for example, when like even our planning, like actually that was my own big first learning when I first joined Clockhood. I'm like, hey, maybe we should, you know, do a six-month roadmap doc. But we're going to do it super lightweight because I don't want to waste a lot of time planning. And I felt we did a really lightweight process. That was such a good learning for me because the exercise was good to kickstart conversations and ensure aligned. But like three months into it, I'm like, wait, have we still referenced time? Because so much has changed. So that was also something I like. And so that was also something I changed that I myself brought in thinking, hey, maybe this will this will help. So always be open to learning and always ask yourself, whatever process you have, is it still serving its purpose just because the field is changing so fast? ([Time 1:23:03](https://share.snipd.com/snip/88342579-d772-4c76-8e72-8860dfe5a763)) - **Switch To Monthly Just In Time Planning** - Use Just-In-Time planning with monthly priorities: Fiona runs monthly lightweight planning sheets and weekly checks rather than six-month roadmaps to stay nimble. - She keeps themes for longer horizons but updates execution monthly because the landscape changes fast. Transcript: Fiona Fung Yeah, I call it JIT planning now, like just in time planning. So it is like around like, cause yeah, I think six months was too long. So now for sure, some projects will take more than a month, but we try to do like a month planning, like really lightweight. Actually, there's not even dots. It's really just us lining on a little spreadsheet of what we think is important. But even that one, I'm kind of thinking through, hey, like every week we should probably still keep like what we're trying is here are the month's priorities and we're going to try it Out. But I have a feeling like every week we'll probably want to do really quick like, hey, just to check. Yep, this is still this month's priorities. Good. But yeah, like but now we've shrunk it to monthly planning. Lenny Rachitsky So it's monthly meaning for the next month. Here's a little sheet slash Excel spreadsheet of what we're planning to do for the next month. And then every week, check in, is this still what we're planning to do for the next month? Yeah, it's like very, very weird. Fiona Fung But even that one, I'm also still feeling, how can we even automate this more? Because I never wanted to be feeling like a tax when someone has to update the spreadsheet. So this is actually, yeah, like just yesterday we're chatting, hey, how can we actually automate this? Speaking to that question, always ask yourselves, can we actually automate this better? ([Time 1:24:36](https://share.snipd.com/snip/4d757703-3173-49ff-89a4-1b30c70a3c6e)) - **Knitting Keeps Fiona Grounded And Creative** - Fiona learned knitting from her grandmother at eight and knits during calls; she likens knitting's stitch patterns to binary logic. - She dreams of opening a yarn store in her grandmother's name to recreate the community she loved. Transcript: Fiona Fung My gosh. Lenny Rachitsky Well i knitted this top uh recently you picked your own clothing unreal this is very meta for collad code building itself you know we're always actually actually this is whole fun thing Fiona Fung I i think between knitting and programming because it's kind of like two stitches knit in a pearl so it's zero and one and anyway so many concepts of stacks and queues you actually could Like i'm kind of like a compiler I'm kind of like you know generating an executable when I knit um yeah actually it was my my grandma that taught me to to knit when I was eight I kind of mentioned Her and I um going to that yarn shop and so every time I I knit I think of her um but I I always like to multitask as you could see like yeah kick off multiple agents and so anytime I'm sitting And because um I practice enough and got proficient I have to I don't have to look at what I'm doing so it's almost almost like how you know sometimes people have like fidget spinners and Such uh knitting is just so anytime I'm sitting still I'm like oh this is time to you know like generate more knitting executable because I got so much yarn that if I don't do this, uh, yeah, My, my yarn, I might have a slight yarn addiction problem. ([Time 1:35:18](https://share.snipd.com/snip/c596cf40-f372-459d-bc14-4bc512e19048))