Unpacking Our Conversation With a Former DOGE Staffer

Zoë Schiffer is joined by WIRED’s Vittoria Elliott to discuss her conversation with Sahil Lavingia. Lavingia worked at Elon Musk’s so-called Department of Government Efficiency, within the Department of Veteran Affairs, until he was fired for speaking out about his experience. Lavingia said his time at the VA was marked by a lack of transparency from DOGE leadership, and chaotic day-to-day operations—the ramifications of which are still being felt today. Just this week, senators called for a federal investigation into the Trump administration’s killing of hundreds of contracts for the Department of Veterans Affairs.

You can follow Zoë Schiffer on Bluesky at @zoeschiffer and Vittoria Elliott on Bluesky at @telliotter. Write to us at [email protected].

You can always listen to this week's podcast through the audio player on this page, but if you want to subscribe for free to get every episode, here's how:

If you're on an iPhone or iPad, open the app called Podcasts, or just tap this link. You can also download an app like Overcast or Pocket Casts and search for “uncanny valley.” We’re on Spotify too.

Note: This is an automated transcript, which may contain errors.

Zoë Schiffer: Hi, this is Zoë. Before we start, I want to take the chance to remind you that we really want to hear from you. If you have a tech-related question that's been on your mind or a topic that you wish we'd covered on a recent show, you can write to us at [email protected]. And if you listen to and enjoy our episodes, please please rate the podcast on your podcast app of choice. It really helps other people find us. Welcome to WIRED's Uncanny Valley. I'm WIRED's director of Business and Industry, Zoë Schiffer. Today, on the show, an in-depth interview with a former DOGE staffer. Our colleague, Vittoria Elliott, sat down with Sahil Lavingia, who used to work as part of Elon Musk's so-called Department of Government Efficiency before he was fired for speaking out about his experience. Lavingia shared how his tenure at the Department of Veterans Affairs was marked by a lack of transparency from DOGE leadership and chaotic day-to-day operations. His statements come at a time when DOGE is in a moment of transition, but the DOGE ideology and DOGE projects have become fused with the larger Trump administration. Vittoria Elliott, senior writer at WIRED, is here to tell us more. Welcome to the show, Vittoria.

Vittoria Elliott: Thanks so much for having me, Zoë.

Zoë Schiffer: Vittoria, let's start by telling listeners who Sahil is. We know he's a former DOGE staffer, but who is he outside of the organization and how exactly did he end up joining the DOGE ranks at the VA?

Vittoria Elliott: Yeah, so Sahil is the founder of a startup called Gumroad, and you can sort of think of it like a sort of fusion of Etsy and Patreon. It's a place for digital creatives to sell their work and he has been working on that for the better part of a decade. He's a pretty experienced startup founder, and Gumroad sort of has this loyal usership but has not achieved the scale of something like Instagram. But he has always had a sort of interest in politics and in using technology in government. He is what I would sort of describe as a true believer in the stated mission of DOGE, which is to make government more efficient and more transparent and work better for people, the sort of work better for the users, which are the American citizens. And to that end, he actually applied for the US digital service years before the Trump administration, back when it was sort of what it was under the Obama administration, which was this small internal tech consulting team that could be deployed at different government agencies to make certain services better. Peak example, the deployment of healthcare.gov. When that failed, USDS sort of became this way for the government to roll out better digital services, and Sahil was really inspired by that. So Steve Davis, Elon Musk right-hand man, the president of the Boring Company, and someone who's been part of a bunch of Musk's projects, was the person who interviewed him and told him that the vision for DOGE was to in-source technology expertise into the governments. Instead of outsourcing to a bunch of technology contractors, that the goal of DOGE was to in-source technological expertise into the government and deploy software that would make government services easier and better for people. That's what Sahil says he thought the vision of DOGE was, and he was really bought into that.

Zoë Schiffer: OK. Well, that makes a lot of sense on a few levels. I mean, this is how Elon Musk likes to run his companies, as you and I both know, with vertical integration. It's part of what has made Tesla and SpaceX, in particular, really successful is he owns a large portion of the supply chain. He's not relying on other companies to build kind of core component parts like other manufacturers often are. And so, like you're saying, Sahil comes in as almost an idealist, and the mandate of DOGE is to root out fraud, waste, and inefficiency. But one thing that really struck me was that he didn't actually find a ton of inefficiency at the Department of Veterans Affairs where he was stationed and that some of the measures that DOGE was implementing, not just at the VA but at a bunch of agencies, like asking employees to send an email listing five things that they'd accomplished in the week, were actually creating a fair amount of inefficiency.

Vittoria Elliott: Yeah. And I think the VA is also so particular, right, because the VA is one of the biggest government agencies, and it is known amongst the federal government to be one of the more technologically forward ones because if you think about it, veterans come back from wars with a lot of disabilities. So there's a lot of contractors and in-house technology experts at the VA that spend a lot of time thinking about how to make services the most efficient and accessible they can possibly be for people with a varying range of access and abilities. And so I think, very much in particular, the VA is a very tech-forward organization with a lot of really smart people working on stuff. And I think Sahil really saw that. And he said one of the first things that he was working on when he came into the VA was not necessarily deploying software but looking at contracts to cut. And he told me that if he had known that sort of his primary focus was not going to be building tools and shipping code but just kind of figuring out who to fire and what to cut that he might not have taken the job.

Zoë Schiffer: Well, I don't want to give away too much of your really, really amazing interview with Sahil. So I guess my final question is just, what is the thing that stood out to you most from this conversation, from his experience at DOGE and what DOGE was like, how it runs?

Vittoria Elliott: I think what really struck me for an organization really obsessed with efficiency and transparency, that there seemed to be very little of that internally. He really described an organizational structure where he was not collaborating with other DOGE teams, sharing learnings, even really knowing what other people were working on. There was very little clarity around who to go to with questions or feedback. And so, even internally, there was sort of this lack of transparency. Externally as well. I think one thing I definitely noticed about Sahil is that he is someone who has really strong principles about certain things, and transparency is one of them. He really believes that if you open source your code or you ask for feedback honestly, that that will make you better. And so he didn't see a problem talking to me or talking to other press because he sees that as fundamental to things improving, and he didn't really understand why DOGE wasn't interested in that in the same way that he was. So I think that's one thing. And then the sort of efficiency that DOGE has really externally said the government lacks, it seems like there's actually quite a lot of inefficiency in how stuff was managed. The fact that a lot of people there, because they didn't have government experience, were kind of remaking the wheel in a lot of ways. There wasn't a lot of, as he described them, easy wins because people were actually doing their job within the limits of the rules in which they have to do them for the federal government.

Zoë Schiffer: I think that's a good place to stop. When we're back, we'll hear Vittoria's sit-down interview with former DOGE staffer, Sahil Lavingia. Stay with us.

Vittoria Elliott: I guess let's start from the very beginning. What drew you to DOGE, and were you previously interested in government work before DOGE?

Sahil Lavingia: I think what interested me to DOGE was this idea that software engineers would have a prioritized place in the federal government. And I think the idea that, as someone who writes code, you could actually ship code for the federal government. This thing that felt like this black box that you didn't really have access to just felt really cool. I think many people would want to do it for the impact, but it just feels like this untouchable thing. And I actually applied to USDS, like the original DOGE, in 2016 I think, 2016 or '17, never heard back. Part of the government black box. And then, eventually, I got... someone had mentioned that they had met someone socially in New York that was working at DOGE. This was late December time period I think, and I was like, "Oh, could you intro me? I would love to talk to them." So I've always been interested in just the ability for software to make the world better, and a huge chunk of the world is the government, like public services, things like that.

Vittoria Elliott: So you said it was kind of opaque about how to get in, but you did talk to people. You had sort of like prescreening interviews, but when we initially talked to you, you mentioned it was kind of disorganized and unclear what was going on. So can you tell me a little bit about what that was, like and what the steps were in that process and the kind of information you were given?

Sahil Lavingia: Yeah, it was pretty vague. I mean, they really... everything felt like it was on this need-to-know basis like you were joining the military and there was a lot of sort of enemy actors out there that you didn't want to give them any information. I got connected to that one person. All communications went through Signal. We were intro'd in sort of a three-person group chat. That was kind of like the default way DOGE kind of operated was these three-people group chats because you'd have the person intro'ing the two people who wanted to talk, and that was almost the entirety of DOGE was this network of three-person group chats. The more formal conversation with this guy, Baris Akis, who's... I guess, again, no one has titles officially, but in my mind, he was kind of a recruiter. And so, he called me, and we talked, and his questions mostly were around the operational, the logistical. Like, "Can you actually move to DC? Are you serious? How soon can you move here?" Well, I didn't know my title, the role, what the work was going to look like, my salary, if I had one. I assumed it was going to be a volunteer gig, basically. I wasn't in it for the comp, you know, the equity. And then he connected me to this guy Steven Davis, who was kind of the acting sort of head of DOGE. I asked him, "Hey, what is DOGE? I can sort of guess at the gestalt of DOGE at this point, but what... could you give me any sort of hints?" And he said, "Do you know any winning technology companies that outsource their software?" The answer is no. "Well, we want to insource the federal government software." I don't think honestly, at that point, the concept of reducing the debt by a trillion dollars, at least to me had come up. It was about insourcing the federal government software. If you look at how much money the federal government does spend on third-party contractors who deliver IT is a significant amount. But that's what really I was like, "Oh, this is actually what I want. I wanted to ship software for the federal government, and this seems very aligned with DOGE and the pre-inauguration days. A lot of people have asked me since, "Did you know it was going to be weird or how political or ideological or..." A friend of mine was like cruel. He used the word cruel, and I was like, "I don't really want to join a cruel organization. I don't think anyone does, but I just felt like if they're really about shipping software and the USDS original mission with Healthcare.gov and all that, I'm totally in. I think that would be awesome."

Vittoria Elliott: Can you tell me about your interactions with the other DOGE people? Did you make friends? Did you make colleagues?

Sahil Lavingia: Yeah.

Vittoria Elliott: ... at least? Do you know... What was your relationship with the other DOGE people, and what were your impressions of them?

Sahil Lavingia: It was pretty friendly. I mean, it was kind of joining a little troop. We were working together 12 hours a day, and we laugh and joke and you had to because a lot of the work itself was pretty boring, to be honest. Reviewing contracts and sitting in on government meetings can be quite boring and frustrating sometimes. Not writing code, for example, can be annoying if you're a software engineer. So yeah, I feel like we were pretty friendly. Unfortunately, when I left, basically all communications ceased. I tried to talk to them. They're just like... I don't know if they were told or if they're just the default implicit promise of if you're working at DOGE, you don't talk to anybody who, once they've left, they've kind of left the church. I was sort of excommunicated like, "Don't talk to this person." Surprisingly little, honestly, communication with the other non-VA DOGE people because they are so worried about maybe information leakage or something like that. They kind of don't... You think you're joining the group, but you're really joining a subgroup of a subgroup of the group. And then, every once in a while, we'd have what Steve called an E-meeting, quote, unquote, which was kind of like a meeting with Elon, and we'd all kind of meet, and it would be this sort of candid Q&A. That was not very goal-oriented. It wasn't like, "OK, let's work." It was kind of like a nonwork meeting.

Vittoria Elliott: Can you talk a little more about who was there and what your impressions of them were? It sounds... I remember, when we talked, you really expressed being sort of a true believer in the sort of mission of USDS/the new mission of DOGE, which it was or the mission of DOGE as transmuted through USDS, which is making things more efficient, using technology for these reasons, but ultimately kind of a believer in government in a way.

Sahil Lavingia: Yeah, totally. I think the meaning overall was more quiet than I was expecting. I was seeking feedback from other people, and I guess the non-feedback I got was it just showed that people joined DOGE for a bunch of different reasons, and I think my reasons for joining DOGE were not aligned as much as I was expecting with the general thrust of the organization and of the other folks who were there. Some people were just really pro-Trump, just really wanted to help Trump do what he wanted to do, whatever that meant. Some people were really fiscally conservative, like libertarian types where they really felt like they wanted to just... any opportunity to cut the size of the federal government didn't really matter exactly what it was a good thing. And I think that's generally sort of Elon's probably thrust. Elon is not as much as I was expecting a software sort of activist. So I still don't really know because I think if Steve had told me in that call, "Our goal is to use software to cut this out to the federal government as much as possible" then I wouldn't have joined. I joined because it was like, "We're going to be the software engineering firm for the US government." So maybe there was just a loss in translation thing, or maybe they needed to say what they needed to say so that they could hire people like me. I don't know. But yeah, I think what I got out of that meeting, most people were nontechnical there. Most people were not software engineers. Most people there were there, and if you're not, I assume you're not that interested in using software to make the government more efficient necessarily. So yeah, I think I was a little bit naively optimistic about how aligned DOGE would be with my stance and with the core sort of original USDS. It wasn't necessarily that people were unopposed, but they were like, "This is not the highest priority thing." I messaged someone at GSA who's sort of in charge of that, and I said, "Hey, I would love to work on this project. I could migrate this whole thing to tailwind, and it would be so much easier for people to work on." And it was just like, how does this correlate to... who's going to do the work? Basically, it was always the first question. I would always say, "Well, I will do it. I'm happy to do the work." But it wasn't clear if you wanted to suggest anything to Steven or Elon how you were to do that. I made some suggestions in that meeting, even though the suggestions seemed to be well received, they were not really implemented. One was implemented, but most were not, and I got negative feedback from someone saying, "Hey, you shouldn't use that meeting for that purpose. That's not what this meeting is for. You kind of hijacked the meeting to talk about your own personal agenda." But really I thought my personal agenda was to make DOGE more successful. So I didn't think it was misaligned there in that way. But yeah, I think I learned over time that basically you're hired to do a certain thing and not to question whether you should be doing that thing or not.

Vittoria Elliott: It seems like, for you, a fundamental value is really around transparency.

Sahil Lavingia: Yeah.

Vittoria Elliott: Is that why you've been talking to me and everyone else? And did you feel misaligned with DOGE, given that transparency is so important to you?

Sahil Lavingia: Yeah, I mean, one of the... in that all hands, I said Elon asked for some feedback. It was just right after they had done the Fox... the first Fox interview. It was the first public thing that DOGE had done, and he's like, "Hey, we should... what do you guys think? Should we do more of this? It feels like people liked it." And I was like... I suggested that we should livestream that meeting. I just said, "Why don't we livestream these meetings? I think that would be cool, and a lot of people are curious about what's happening, and it would be useful to the world and good for us," and etc., etc. And he was like, "Sounds great. We'll do it next week." And he computed it. He was like, "This is how we'll do it. We'll do it here. There's this risk." He really sort of spent some time thinking about how to actually make it happen, and then it just never happened, and I don't know why it didn't happen. But just, generally, I find that, yeah, I think I was aligned with the marketing of it, which is like, we're going to be the most transparent organization in government, maximally transparent, etc., etc., and it just never manifested. It just never was a priority. I think, for me, transparency is a real priority of actually the way you make things more efficient is by being transparent because then you can see, and other people can see and suggest how to do a better job, right. In hindsight, I was pretty misaligned with DOGE in that way where, if you're running a maximally transparent organization, you should allow your employees to talk to the press. You should allow anyone to talk about anything, the good, the bad, what they struggled with, what they learned, and also allow for internal discord, right. Actually, allow for people to say, "Hey, this thing that DOGE is doing I think maybe we should not do."

Vittoria Elliott: Can you tell me who you got the impression was sort of in charge of DOGE or the sort of nerve center? Who were the handful of people that you felt like knew what was going on across these different teams?

Sahil Lavingia: It felt like basically just Steven Davis. It really felt like Elon was at the head of the table. Next to him was Steven Davis. The way I think about it is, Elon is kind of like the chairman of the board. Steven Davis is the CEO. Baris Akis was this sort of COO chief of staff, and everyone else was kind of like a direct report of Steven Davis.

Vittoria Elliott: And you mentioned that everything's over Signal. Obviously, there's rules around maintaining records around government work if you're a government worker. Did anyone ever talk to you about, "Hey, you have to archive these messages. Hey, you have to save these emails. Hey, you can't delete that." Did anyone discuss your... any legal obligations around that with you for your work at DOGE?

Sahil Lavingia: No. No. There was never any DOGE or onboarding, right. So, no communication around talking to press, not talking to press, taking photos, not taking photos, disclosing anything you've done with anybody. And yeah, nothing about archiving, maintaining records. It was just done on Signal, and it was never communicated like, "This is why we use Signal, for example." It was just like, "Steven is going to call you on Signal," and that's about it.

Vittoria Elliott: And a lot of the DOGE engineers have been assigned to multiple agencies. Were you ever assigned to another agency outside of VA, and did you have a sense of why there were so many DOGE engineers across multiple agencies simultaneously?

Sahil Lavingia: Yeah, I was not. I was assigned to VA, and I never got detailed, I guess, is the term detailed to any other agencies. My feeling is that there were just way fewer engineers at DOGE than people would think. There were 10 or less. And so I think they were just insanely limited in how many people they could actually... If they had 500 engineers, they wouldn't have to do that, but I think they just had so few, and specifically so few that they trusted, right. I think there probably was this maybe schism and sort of the sense that if you joined post-inauguration, you were seen as more of an external contributor volunteer. And if you joined pre-inauguration, you were taken a bit more seriously, is my guess, and just had more time to build trust with Steve and Elon, etc. So I think they just only trusted a very small contingent of people, so they just needed the same group of people to just run around a lot and go from agency to agency.

Vittoria Elliott: Got it. Given that you were touching possibly really sensitive data, was there any controls that were put into place? Did anyone sit you down and say, "You're going to touch a bunch of sensitive stuff. You are or not allowed to do this." Did you get a sense that there was any sort of, I guess, conversation at DOGE about how to handle this data, how to make sure it's protected safely, how ensuring that people might go back to the private sector, that it's not used for their own benefit?

Sahil Lavingia: Yeah. There was nothing DOGE-specific around that, but there were. I think it depends on the agency, and the agencies all have their own protocols that they follow. And in our case at VA, there was some data that they said, "If you want access to this, you have to go through this two-hour sort of course online where you go through a bunch of tests and have to say that you're able to... you understand the risks." But in terms of the HR data, I was able to get access to all the HR data, all the contracts with no... I didn't have to sign anything or say anything that I agreed to not share it with the public or do anything like that. It is kind of implied. I was surprised, though. I mean, I was... I asked them one time. I was like, "I like how many people have access to the HR data?" And they're like, "Basically everyone in HR." I'm like, "You know this data has everyone's social security numbers in it. There's like 473,000 social security numbers." And they're like, "Yeah, thousands of people have access to that." I was like, "That's kind of weird." And I do think the government could improve on these controls and making them better. There was one time where I got access to this thing called CDW, where... which is I think stands for the Corporate Data Warehouse, which effectively all the data in VA was in CDW at one point and they revoked my access a few hours later when they realized that someone had given me access to this and they were basically like, "Dude, why did you give him access to this?" And they were like, "Well, he's senior advisor to the chief of staff. That's... And our protocol is to give you access to all the data when you asked for it." I think there was this feeling in the beginning that DOGE had a lot more power than people thought, and so everyone was just kind of doing what we asked them to do and unnecessarily that I'd be like, "No. By the way, follow your own process. I'm not asking you to do anything different."

Vittoria Elliott: There have been members of Congress and some reporting that there have been possible instances of data exfiltration with DOGE, and I'm wondering if that's anything that you were ever aware of or if there was any concern expressed about that?

Sahil Lavingia: Not in my experience. Yeah, I mean, I've read some articles about it. Someone maybe doing... interagency sending from one... data from one government laptop to another government laptop. I haven't heard of any data leaking or going to the public or state actors getting the data. And in my case, yeah, I don't think any of that happened. I was pretty good about... I would actually leave my VA laptop within the VA headquarters. I wouldn't even take it back to the hotel because not much you could do anyways. And yeah, so it wasn't a huge issue. I think the core issue I'd be worried about, it's just employees who are just not careful with... Just the average employee gets a lot more data than I was expecting pretty easily.

Vittoria Elliott: Was there anything about DOGE that made you uncomfortable? Either something you were asked to be doing or anything you saw happening? Because again, you sort of spoke about this sense of maybe misalignment, so I'm curious if there was ever points where you felt uncomfortable with what was happening.

Sahil Lavingia: I'd say the only thing was they'd say, "Hey, we're trying to conduct this riff." And I was like, "What do you mean?" And they're like, "Oh, just put together, here's a Google Sheet, a sheet, an Excel file of all the subdepartments and how many people work in each one and the job codes. And let's help the VA actually do it, not just give them the org charts, but actually give them the number of people to riff at every layer." And I basically was just like, "I don't know how useful that is, right." How useful is it to be like, "Hey, you have 4,000 software engineers or IT specialists as they're called, and you should have 2000?" I can do that. I guess I can change some numbers around in Excel, but who's... what is the utility of this product? Who cares? Or when Trump signed some EOs that said All the memos, you have to review and get rid of anything antisemitic or things like this. A bunch of lawyers came to me because they heard I was there, and were like, "Hey, can you help us?" So I wrote a little script I published on GitHub so people could see, but it was basically just loops through all the memos similar to the contracts and just checks if they're non-compliant with Trump's EO. And I asked them like, "What if I just said, no, I just don't want to help you." And they're like, "Well, you don't have to help us. We'll just do it manually." And I'm like, "You know there's like 500,000 memos though. How would you do that manually?" And they'd say, "We'd sit in a room for weeks and just review every single memo." And I felt uncomfortable broadly with that idea that we're just creating work for no reason. It's just a lot of these things have zero value, knowing that basically nobody was reading the five bullet points that people were sending every week.

Vittoria Elliott: How do you know no one was reading them?

Sahil Lavingia: Well, because apparently the DOGE people were supposed to be paying attention to that kind of stuff, and we weren't. They were like, "Are you reading them?" I'm like, "No." And they're like, "OK." I was like, "Well, were we supposed to be reading them? We don't even have them. How would I read them?"

Vittoria Elliott: I remember in our first conversation you talked about trying to cancel contracts, and you had mentioned that there was a Palantir contract that you had wanted to cancel at the VA. Did you ever manage to get that through?

Sahil Lavingia: No. No, I don't think so. Yeah, I felt like, ironically... Now I realize that this idea of insourcing the government software and becoming the government software engineering firm, it seems more likely that Palantir is going to become the government's software engineering firm. That effectively the way things are going, if they continue to go in the direction that they're going in, that actually all these contracts basically would just go to Palantir over time, and they'll have Palantir do it because the government's not willing to hire software engineers. And Palantir is... honestly, their software is pretty good, relatively good. We're using at VA. It's just really expensive. We're paying tens of millions of dollars a year for software that I could build in weeks by myself. It just a lot of margin in. I mean, you could look at Palantir's margin is probably pretty good, is my guess.

Vittoria Elliott: Can you talk to me a little more about what you saw at the VA, and did you feel like it aligned with Musk's and DOGE's assertion that the federal government is inefficient?

Sahil Lavingia: The VA specifically is not that inefficient. Just numerically quantifiably, the VA is giving out hundreds of billions of dollars in disability payments and healthcare and is staffed. It's about $50 billion a year in wages, so it's about one-seventh of the money goes to the employees. And that includes, by the way, every nurse, too, right. You can't really provide healthcare without nurses, right. So, at the end of the day, VA specifically has a big budget because it provides a lot of services to Americans. And inefficiency, if you define it as the percentage of discretionary spend versus sort of the entitlement spend, is just relatively small. And the only way you can really reduce the size of VA is by cutting entitlement by actually just reducing benefits. And that was never discussed. It was never discussed. At any level, I never heard a conversation around, "We need to reduce entitlements. That's the only way we're going to get to $1 trillion." It was like, "No, we're going to do this by purely using software and cutting contracts," which just felt like not really exactly true or possible in any real way.

Vittoria Elliott: Do you think that many of the problems that DOGE's leadership was purportedly out to solve was the result of not really understanding how government works?

Sahil Lavingia: Yeah, I think that's a fair point. I mean, I think that at the end of the day, there's a level of rigor. No one really said, "What if we're wrong?" It's like we have this unique opportunity to participate in this technological pseudo-coup of the federal government, and this is such a unique time and place, and Elon's here, and this is crazy. No one wants to do the research and learn like, "Oh, by the way, it's not possible." And I think there wasn't really a ton of interest in understanding the federal government, the agencies because I think that that would just kind of make things less fun. And I think a lot of it was, as Elon says, the most entertaining outcome is the most likely. I don't think he was necessarily optimizing for the most efficient outcome as much as he was optimizing for the most entertaining outcome.

Vittoria Elliott: I wanted to talk a little bit about the end of your time at DOGE. You obviously have told me and said in your blog that you were fired.

Sahil Lavingia: Mm-hmm.

Vittoria Elliott: Can you tell me what happened?

Sahil Lavingia: Yeah, so this writer wrote an article about Gumroad, and he had mentioned... he had sort of had this conspiracy that I had open-sourced Gumroad such that it would make it easier for me to work at DOGE and work for the government because then I could still work on Gumroad stuff without having worked there. Something like that, I think, was his posture. And even though I didn't agree with him, similar to my experience with WIRED, I appreciate his work and his attempt. And so I talked to him about Gumroad and about DOGE, and he wrote an article. He ended up publishing it in Fast Company, and Baris actually texted me about it, and he said, "Hey, did you run this by anyone at DOGE?" And I said, "Nope, sorry. By the way, I didn't... it wasn't meant to be in Fast Company. It's not like I was going out and talking to the press. It was actually his personal blog, and he had a good deal. He sold it, and he made some money. It is what it is." And I never actually heard back. Actually, that was the last message. It's a day later my [inaudible] access got revoked, and I just stopped getting any VA information, DOGE information, everything just kind of feel like... Basically, I just got ghosted. Basically is what happened was I said the wrong thing, and I just got ghosted for life. And since then, no one has... no one from DOGE has contacted me. The only people who've talked to me about DOGE are people who have left DOGE at this point or plan to join and are curious about my experience.

Vittoria Elliott: And do you know a lot of people who have left DOGE?

Sahil Lavingia: Yeah, quite a few now. Almost all my friends, all my social friends who've joined DOGE, have now left. The only people there are kind of the people I didn't know before I joined all the kind of autistic software engineer-type people.

Vittoria Elliott: Why do you call them autistic software engineer-type people?

Sahil Lavingia: I mean, I don't know if they're autistic or not. It kind of just was a meme, I guess, in the early days of DOGE where someone broke the story of these four or five.

Vittoria Elliott: It was me.

Sahil Lavingia: Was it you? Really? Wow, cool. Well, it was just this meme on Twitter that sort of the face of DOGE became these sort of very young, what looked like this sort of archetypal Silicon Valley software engineers who just worked all the time, slept where they worked, this kind of the very startupy vibe, wore hoodies, that sort of thing, call it dropouts, focused on the work, not very engaging on a personal level, was hard to converse with this kind of 19-year-old Silicon Valley engineer that is more focused on shipping code than basically anything else.

Vittoria Elliott: Were most of the engineers young?

Sahil Lavingia: Yes. There were only... There's only one other engineer I met who was over the age of 30, a woman actually. So it was me, her. We were kind of like the boomers or something. And then it seemed like every other engineer was really young, between the ages of 19 and 25 sort of, which I think makes sense. At the end of the day, you're moving to DC, you're saying no to getting paid lots of money. And so it ends up sort of filtering out, unfortunately, filtering out a lot of really great software engineers. The best software engineers I've ever met, never... they all considered working for DOGE. I've talked to some of the best software engineers literally in the world, and they're all, "I considered it, but I couldn't move to DC or I couldn't work full time."

Vittoria Elliott: I mean, sort of in that vein, it sounds like that's a mistake you felt DOGE made in terms of eliminating maybe really great talent. Were there other... Are there any other mistakes that you feel like DOGE made?

Sahil Lavingia: I think the biggest mistake that DOGE made was just not admitting when it made mistakes. And it felt like DOGE had sort of picked a side and was not willing to sort of give anything to the other side if they pointed out a mistake. And I think I wish DOGE did a better job of informing the team and the public, "What have we tried, what worked, what didn't work." I think, generally, I think speaking, it lacked a feedback loop. It sort of said, "We're going to work on these projects no matter what. We're going to save a trillion dollars by cutting contracts and reducing the force no matter what." And I think that's not how you succeed. So yeah, I think lots of things DOGE could have done better.

Vittoria Elliott: How would you rate DOGE's success in the government so far?

Sahil Lavingia: I mean, relative to what Elon, I think, I don't know if he promised it, but this idea that we could save trillions of dollars just on a numerical basis. I think DOGE will save best-case low hundreds of billions of dollars. So even best case, it's sort of like 10 or 20 percent, which is like an F grade in high school. But I think in terms of if you measure it from the average impact a single person had, it can be quite positive. Like I was able to save the VA, let's say, $10 million a year. That's pretty good for 55 days, like $10 million a year in savings. Hopefully, more than that over time as my code continues to do its thing. So I think, as an individual, I was able to be very impactful. DOGE as an entity, I think, was not as successful as they wanted to be, and it was successful in the way that Trump wanted it to be successful, which is I think it was able to absorb a lot of the heat. And Trump is amazing at being the current thing and taking up all the headlines. And I think DOGE was very successful in helping him stay of mind and let his team do whatever they wanted to do, like cut USAID, etc., without them taking the heat for it and letting DOGE and Elon take the heat.

Vittoria Elliott: Well, thank you so much for your time.

Sahil Lavingia: You're welcome.

Zoë Schiffer: Thank you so much for that conversation, Tori. Just one last question before we go. Where do you think DOGE goes from here?

Vittoria Elliott: I think what we're really seeing in this moment of real tumult and transition is this. One of the things that Sahil says is that Trump really benefited from the visibility of DOGE and Musk and the fact that it kind of seemed to run in a parallel track from the administration, right. And so that meant that the unpopular things that DOGE was doing kind of were getting pinned on Musk and not the administration as a whole. But the reality is that cutting down the number of federal workers, cutting down the number of federal contracts, combining data sets to enforce the immigration agenda, that's all stuff that the Trump administration kind of wants to be doing with or without Musk. And I think the future of DOGE is that those projects, which really line up a lot with some of the stuff in Project 2025, are going to continue and are going to be supported by the administration. It's just that without someone like Musk really being very public about it, without the shock and awe of these young guys being brought in to execute on these ideas that it likely will continue, but it's just kind of going to be the government now.

Zoë Schiffer: That's our show for today. We'll link to all the stories we spoke about in the show notes. Make sure to check out Thursday's episode of Uncanny Valley, which is about why Disney and Universal recently joining the AI copyright battle matters. Adriana Tapia produced this episode. Amar Lal at Macrosound mixed this episode. Jordan Bell is our executive producer. Conde Nast's head of global audio is Chris Bannon, and Katie Drummond is WIRED's global editorial director.