Event Replay: Careers at The Frontier: Hiring the Future of OpenAI Part 2

I'm going to give folks just a few moments to filter in, but I already say Professor Mahaffey from University of West Florida, welcome, so good to see you. Thank you for making it late on a Thursday evening for you on the East Coast.

Melanie McDonoghue, see you again, Chief Innovation AI Officer for the City of Lebanon, New Hampshire, so good to see you.

Awesome, let's get this party started.

Welcome everyone, I'm Natalie Cone, Head of the Open AI Forum, the expert community hosting our conversation this evening.

In the forum we spotlight discussions that reveal how AI is helping people tackle hard problems, and we share cutting-edge research that deepens our understanding of this unprecedented technology's societal impact. AI is an innovation on the scale of electricity. It's transforming how we live, work, and connect.

OpenAI's mission is to ensure that as AI advances, it benefits everyone. We build AI to help people solve the toughest challenges, because solving hard problems creates the greatest benefits, driving scientific discoveries, improving healthcare and education, and boosting productivity across the world.

We launched a new series in partnership with our incredible people team and recruiters to demystify what we look for in future OpenAI teammates. Our goal is to set every prospective candidate up for success, helping you walk into interviews confident that you understand our values and comfortable because you already feel like you know us.

Welcome to Careers at the Frontier Part Two, Inside OpenAI's Recruiting Process.

Today we'll be joined by Sarah Warkov, a Sourcer at OpenAI, where she spent the past two years building teams across Applied AI and Research. She's helped recruit engineers behind ChatGPT, Sora, and Operator. Often starting from the very first hire. And now Sarah supports hiring for OpenAI's robotics organization. Before OpenAI, she recruited technical talent at Google, Meta, Amazon, and in startups. Outside of work, she enjoys family time, long walks with her 90-pound and binge-watching 90-pound dog.

Oh, Sarah, when you come on, you'll have to tell us what kind of dog you have. And binge-watching reality TV.

Red Avila is a recruiter at OpenAI where he spent the past five years building teams across people, legal, research, engineering, policy, and more. Five years is a very long time in a fast-paced organization like OpenAI. He's recruited talent behind our core research and even participated in Red teaming models like GPT-4. He enjoys being a utility talent partner, working across different domains, and keeping his skills fresh. Outside of work, Red spends most of his free time with his longtime partner learning about wine, foods around the world, and using ChatGPT to help him stay fit and active. Please help me welcome Sarah and Red to the stage.

Hey, guys.

Hello.

Hello, hello.

I just love collaborating with the People team, and I'm so honored to be welcoming you to the forum tonight.

Thank you.

Happy to be here.

Honored to be here.

And Red, what is that shirt that you're wearing? I know and I love it, but I want you to share it with our audience.

Yes, so this is my favorite piece of OpenAI clothing and it's our Sora shirt. So it's a fun little art here and it's a long sleeve, which I love. Yeah, this is me. It is super, super fun.

Guys, I actually wanna start with a question that when we began planning this talk, it was my question because I am not a recruiter and I just really wasn't incredibly familiar with the process.

So Sarah, will you tell us what the difference is between a sorcerer and a recruiter?

So, you know, you're, Sarah's a.

Sourcer, everyone in the audience, Red is a recruiter, and maybe it would help if we just started by discussing like what the difference is.

Yes, I'd love to. So Sourcers and Recruiters play different roles, but the end goal is the same. We want to hire exceptional people to align with our mission. So Sourcers are talent researchers. We research to understand what's happening in the market. So who's joining where, who's working on cool ideas, or who raised a round of funding, and we help get candidates interested in interviewing, and then we'll guide them through early interviews.

So typically that's a call with myself or whoever your Sourcer is, and a technical screen, and then we advise how

hiring teams on how to proceed based on their skills and experience. From there, recruiters focus on later stages, so final interviews, the offer process, and beyond. We actually work really closely together to help make sure that the candidate experience is as seamless as possible. And actually, there's a small team typically that supports the interview process end to end.

So there's coordinators, hiring managers, interviewers. I do support robotics. So the recruiter is Vicky. I'm the sourcer. The coordinator is Jared. And we're all working together to help candidates get through the process.

Oh my gosh, that sounds like such a fun job, Sarah. And in terms.

of your research, not the funding rounds. We know how you can find that, but do you also get to spend a lot of time researching research, like learning about the bleeding edge research and who's contributing to it?

Totally, yeah. We have pretty unique, I would say, research capabilities that we do.

One of my favorite sourcing strategies actually to find candidates is people who have won Rubik's Cube competitions in the past. I'll go to the World Cube Organization website and a lot of smart people are winning Rubik's Cube competitions. That's a super fun fact.

Thank you so much, Sarah. Again, welcome to the OpenAI Forum.

And I'm seeing now that we should be collaborating more because you can tell me what the bleeding edge is and what kind of conversations we should be hosting here. So onto you, Red, you're a recruiter. You've been here for five years. What does it mean to be a recruiter? What kind of work do you do at OpenAI? And how can candidates expect to engage with people in your role at OpenAI?

Yeah, definitely. I think recruiters play two interesting roles here. One, you're sort of the orchestrator for the hiring process and experience. So you think about what is the plan? What timeline are we on? Who's involved in this process and why? But I think more important to that, you also are one of the entry points for candidates to get to know.

open AI. We try to remind ourselves regularly that we're being interviewed and being considered just like candidates, and so it's up to us to give everyone a great experience. And so someone like Sarah is someone who's going to support the process and help us find great talent, and it's my job to make sure that process supports that. And so normally I'm also talking to candidates every day, especially as their interviews go into later stages. I'm usually their point of contact, so I'm a call, text, email away from them, and hopefully eventually being the one to actually deliver their job offer to them and help them through that process. So that's a key point for me.

That must feel really good to be the person that gets to deliver the good news.

Yes, it is.

And so it's an interesting point to be in. Definitely. But I can understand how you're really good at that role. It's been so fun collaborating with you since the very beginning.

Just so the audience knows, when we hosted Joaquin, he just took over the role of head of the people team. We were hosting him in the forum. And we originally intended on hosting Joaquin for a totally different subject, but things changed at a point.

And I really fast. And so, you know, we decided to pivot, make this about recruiting and people at Open AI, which was super successful. So here we are for part two. And read, you just jumped in and you had so much institution.

knowledge and you really helped us craft that talk. So I can see how you're so good at helping people feel comfortable and navigating the hiring process, super rad. Awesome.

Okay, well, let's move on to some of the discussion points that we decided we wanted to share with our audience. So Sarah, we already talked about recruiters and sourcers, but how about when you're looking at resumes, what kind of signals stand out to you?

Also, how many resumes do you review on average? Yeah, I'll answer that second question first. I mentioned I support robotics. It's a really big industry, and I think it's a really important one for us. So I think it's really important that we have a lot of people that support us and we have a lot of people that support us and we have a lot of people that support us.

really exciting area. So I'm reviewing over 1,000 applicants per month. It's really exciting. And so typically, when I'm looking at a profile or a resume, it's rarely about a single credential. What stands out is ownership. So when we see candidates who have built something end to end or take an initiative in ambiguous situations, that's a much stronger signal than just listing every technology that you've ever used.

We're looking for curiosity, adaptability, and clear evidence of impact. I love that. And we kind of touched about that in part one.

But tell me in explicit terms, if you can just recall it.

Like, what does evidence of those qualities look like in a concrete way?

Yeah, I would say it can come out in a few different ways. Actually, there are some patterns that we have seen in some hires that we've made. So one of them is technical breadth. Many people work across multiple systems and combine research and engineering, whether it's publishing papers, contributing to open source, or building production systems.

Another is end-to-end ownership. So they've repeatedly taken projects 0 to 1, whether they're building early prototypes, developing infrastructure layers, or scaling.

products that reached millions of users or supported critical internal workflows. We also see visible impact. So this is kind of where I was getting to with the resumes. They want to emphasize outcomes, not just responsibilities. Metrics and results like improving onboarding conversion rates or saving millions in storage costs or unlocking multimillion dollar deals. Those really stand out as I'm reviewing them.

Two more kind of patterns that we've seen across our hires. One is adaptability. So switching domains quickly. One candidate that I can think of recently taught themselves Android and Google Glass SDK in just a few months.

and shift a working prototype, finally scrappiness, so wearing multiple hats, building outside of work through side projects, hackathons, or open source contributions with real traction.

Yeah, thank you, Sarah. That is super helpful and definitely goes a bit deeper than we did in our first talk.

This one is for you, Red. So you have been here a long time, and for some people in the world, five years might not sound like a long time, but OpenAI is basically moving at the speed of light.

So tell us how you usually describe OpenAI's hiring philosophy, because of all people, you would be able to do this best.

Yeah, thank you.

I think it's actually pretty simple, frankly, and I think the simplicity comes from the fact that we want to find the world's best talent across many disciplines who want to work together to build the most transformative technologies we'll probably ever see. And then we want to make those technologies accessible to the world and then in turn ensure the world benefits from them. And then we need to create an environment where that can happen. So it's actually like very easy.

But it's actually not easy. No surprise. There's a lot more to it. But the details do matter. And that actually feeds into the longer, more serious answer, which is that we have found that consistently our

happiest and most impactful employees and team members are really the ones who come in with this sort of builder, owner, partner mindset. And if I kind of break it down simply, we think of builders as people who roll up their sleeves and aren't afraid to like create something, including zero to one.

Owners are people who see things all the way through from ideation to outcomes and take ownership over those results. And partners are people who collaborate and share the successes with others.

And so, it's not about being a perfect employee or having a perfect reputation or work. It's really about being scrappy, adaptable, and I think ultimately motivated by our mission. It matters deeply to the day-to-day.

And speaking of, if you start working at OpenAI, you will definitely start working on day one. We have an onboarding process run by an onboarding PM, Janelle, but yes, we'll see your impact very quickly. Definitely, you might get started before you get started. You potentially, yeah. You gotta at least learn a few things, sure, yes.

And Sarah, I know that your role is a bit different, but I would actually just love to hear, in your own words, if you could describe our hiring philosophy. Yeah, I mean, it's hard not to repeat exactly what Red, but he's been here a long time, and I think Red has really helped shape our hiring philosophy.

philosophy. We are looking for people that are really curious about what we're doing. One thing I like too is that we want people to be concerned about the safety of what we're building. So safety is a huge thing that we look for when we're speaking with people.

And then also just humility. It's well known here that we all have the same title. It's a member of technical staff, member of people's staff. And so I think just kind of taking away like the titles part of it, we really want to make sure that no matter what level or title that you are, you're contributing and having the impact that you want to bring to OpenAI.

I love that. Thank you so much, Sarah.

Red, let's turn back to you. You have infinite wisdom at OpenAI. You've helped hire some of the most influential researchers in the world, like Lukasz Kaiser.

For those of us who are not technical in the audience, he basically is one of the contemporary founders of AI because he co-authored the Transformers Attention is All You Need paper. And by the way, he will be presenting in the forum very soon in October. So please make it back if you are interested in meeting him.

But Red, what traits are we seeing again in successful hires and any kind of behind the scenes?

you might give us at how we ended up landing some of the most amazing thinkers, what they have in common, you know, just whatever you just, you feel like sharing with the audience today, I'd love to hear from you.

Yeah, yeah, definitely. I mean, Lukasz is a great example. I mean, you know, you look at his background and it speaks for itself, right? To your point, credited on some of the most influential research projects and papers and technologies in recent times. But there are some things that make Lukasz kind of unique and special, I would say. And he may not say this about himself, but I'll say it. I think it is partially his openness to collaboration and truly being willing to lean in and understand.

understand other people's ideas and perspectives, work with them and adopt them. And that is also across different disciplines that curiosity is important. And you put that into someone who is quite humble, I would say. Humility is a big part of being successful at OpenAI because we're all doing things every day that might be a little uncomfortable or new. And those are things that I think are traits that matter and end up in successful hires over time.

So we care a lot less about like your title or how senior you are necessarily, you know, your age index, and a lot more about the impact you've had.

Very much what Sarah was talking about, like what have your ideas actually done?

contributions look like. And I think you take that and you go into an environment like OpenAI, where you actually have the ability to scale those things as well. A lot of times people ask about, you know, is there a secret to OpenAI's work or success? And I think, honestly, not necessarily. The secret sauce is really blending all of that talent together, and allowing them to work on a shared goal. And frankly, it's as simple as, is everyone on board with the mission? Do you understand it? Do you have a connection to it? And do you believe in it? That by itself, like that will push people much farther than giving them a big title or all the resources in the world. And I think OpenAI has an environment where it's kind of an interesting blend of like academia.

a private industry. Um, and it's something that we hear from new hires regularly. They talk about that unique blend being the reason they chose to come here. Um, you know, I truly believe you cannot fake people's genuine passions and what they can do collectively with their ambitions. I think that is that is the secret sauce, frankly.

Yeah, that all really resonates red and and maybe I can just contribute a little to that story in that.

Well, one, I do also love collaborating across disciplines with researchers, with the people team like you guys with our go to market team. Every single event that we host in the forum is a cross functional collaboration.

And something that we've talked about before in the forum, and I'm going to bring back up now, which I think can be added, is that we are the kind of people that even though we're afraid, we jump into things head first. And actually, my partner says that sometimes I jump in face first. Because the reason the fear, it's unknown. It's unknown. We're experimenting, and we're trying things, and we're moving much quicker.

My background before I came here was in the arts nonprofit sector. And I don't know if I would ever survive in an environment like that. I love it. It's so fun and intellectually rigorous, but way slower. Here we move so fast.

And if no one has ever done what you're doing before, there's no roadmap, there are no instructions. And so it's totally possible that you're gonna fail and you're gonna fail in front of a lot of people.

And I actually wanna give credit to my colleague, Sarah here today, because when I reached out to her about a month ago and asked if she wanted to do this, initially Sarah said, I would really love to, but I struggle with stage fright.

And then we started jamming on that. And I told her that when I started doing this work four years ago, I also had incredible stage fright. And by the end of our chat, Sarah said, okay, I'll do it. And here she is today.

today, and there couldn't be a better person here communicating these ideas to us. So Sarah, thank you so much for being here, even though you were nervous and you jumped in.

I was so nervous. Thank you. I was so nervous. I still am. I mean, this is like exposure therapy to the max. I feel like I'm still scared right now.

But one thing that also worried me was during the last event that we did, Joaquin took out his guitar and started singing. So I was like, I don't know if I'm the right one, because he really put on a show. And I was worried to follow that up.

No, you've been a perfect collaborator. You've helped both Reds.

and I move this idea, you know, into actual execution. So thank you so much.

And Red, similarly, when I was trying to pull together that talk with Joaquin, you jumped in and you created and organized like a five page document.

You got all of the people team to collaborate. And really that is why that event was so beautiful.

If it's because there were like 20 people working on it but you organized all of it. And I didn't even ask you to, you just jumped in.

So that is OpenAI for those of us looking. And there's so much humility. And it is all about just lifting up that particular initiative that rolls up to the mission.

And we all enjoy working with each other. So I feel very lucky that you guys are here today. We get to share this story. Thank you.

Okay, moving on past all the love, love, love, love. We like each other.

Okay, Red, even our managers and leaders are still hands-on. Oh boy, is that true? We've got so many stories that can give that more color. They're coding, prototyping, mentoring. So it's not about moving away from the work. It's about staying close to it while scaling yourself and your team. Can you tell us more about this?

Yeah, I think, you know, the responsibility that leaders have at OpenAI.

is quite high like any successful organization, but I will say one difference I hear from leaders time and time again is they're excited but also surprised at how much hands-on work they're both able to do and kind of required as well. And there's no escaping it. You can hide in a corner for only so long.

So our managers and leaders are hands-on. They have to be. This is an incredibly fast-moving organization that is incredibly ambitious and it is the next big company of the future in tech. And I think that requires them to be coding, to be prototyping, mentoring, digging into details.

So it's not really about moving up and away from the work. It's more about

staying close to it and actually helping scale it out, including your team. And so I think that combination of leadership and hands-on execution is sometimes rare, frankly, but it's a part of what we look for at those levels and we need that. The way I think about it is we don't actually have a lot of room for kings, queens, royalty, or armchair philosophers.

Every day that we show up here, our efforts do count and matter, and that includes all levels of the organization. So I think it's maybe unique to us, but it is what makes us successful and we want more of it. Definitely.

Sarah, do you-

to add anything to that, no pressure, but if you had any thoughts.

No, I completely agree. It's amazing actually that we are able to find people that are still hands-on but also have that people leadership. I think that, and another thing I would add is just people that are like building outside of work that we really like side projects. We really like, I don't know, it's like the Rubik's Cube thing I brought up.

I really like when a candidate like shows his camera over and he's like building something from scratch or I had one candidate that's like, oh look at this story that I wrote and it's just cool to just see people tinkering and you know.

know, they, a lot of times they have those interests outside of work that we just, we love to see it. That makes a lot of sense, because it's about intellectual curiosity, and it's about building. So that really makes a lot of sense. And we promised our members and the guests here today that we would give concrete examples. So I'm going to go ahead and jump in, because I have two that are front of mind to just just really give this expression some more concrete depth.

So, Joaquin, for instance, the leader of the people team, he, before before that, he was the head of the preparedness team. And you know, he was helping evaluate our models for catastrophic potential. And

And he had gotten to a point where he felt like he had been a manager for such a long time that he hadn't had his hands in the weeds of the technical pieces for a while, and knew that he needed to stay close to that at OpenAI. So he stepped back from that role, a leadership role, to become an intern and start working with code again, before he jumped back into a leadership role. Because he wanted to know, how can you lead a team if your hands aren't in the weeds with the things that they're doing?

Similarly, actually my manager just today, I was telling you guys this as we were preparing, we're looking over the calendar, we're trying to scale, OpenAI is scaling all over the world

to the forum is scaling all over the world and launching global chapters. And we cover lots of different domains and disciplines. And I'm looking at the calendar and I tell my manager, I honestly don't know how we're gonna do that. Our team of two is gonna do that.

And our manager has only been the head of our team for a couple of months. And she said, which events do you want me to lead? I'll do it, teach me how to do it. And she jumped in and we distributed the labor and she's learning it for the very first time, taking on some pretty high profile talks.

And that's exposure and it's scary, but it is absolutely.

a testament to those characteristics that you guys are saying that we'll look for. So, yeah, definitely a special place.

So are you guys okay if we move on now to advice for candidates, and then we'll do some Q&A? Okay. Totally.

Sarah, what advice would you give to someone who's preparing to apply or interview at OpenAI?

This is actually my favorite thing to do. If you've hopped on a call with me in the initial stages, I spend half the time preparing you for the interviews, because my biggest piece of advice is preparation.

So if you're preparing for coding interviews, brush up on data structures and algorithms, get comfortable with the code.

CoderPad platform, it's not like coding in a full IDE where you have autocomplete in a debugger. So you will be expected to explain your thought process out loud, and it really helps to practice this ahead of time. So also come prepared with stories of your impact.

Joaquin mentioned this last time, but really the simple framework of star, situation, task, action, result. Write your examples down, revise them, and practice until you can tell them clearly in two to three minutes.

And the last thing I'll say is use data whenever you can. For example, I scaled a system to handle 10x more users, or I improved uptime.

from 95% to 99.9% or I unlocked $20 million in annual savings by optimizing infrastructure. Those are the kind of details that make your story really memorable. Thank you so much, Sarah.

I wish you would have been supporting me along the interview process. It sounds like you're a really lovely sorcerer to have on your side.

Okay, Red, from your perspective then, we wanna walk away from an interview understanding what you could bring...

Oh, I'm sorry. What, from your perspective, do we want to walk away with?

No, you're right. I mean, it's what you were mentioning. Like, you know, anytime you walk into an interview, it's nerve-wracking, of course.

And honestly, on both sides, because you just, you don't know each other yet, right? You're going to get to know each other in this space.

But I think as a candidate, your goal really should be to try and impart your experience, your knowledge, your skills to the interviewer. And that also means that it's not really about putting on a show necessarily, but more about relaying your experience accurately and well.

And so if I ask you like, how have you dealt with a sudden change in a plan or strategy in your experience, in your role? Your brain should jump to an example. You should be able to walk me through that. Almost like I'm watching a movie. I can sort of visualize everything happening in your role in that.

And that's much better and much more effective.

effective than generalities. So maybe to Sarah's point, actually, you can think of the STAR method or format as a way to structure your responses. It's simple, but it's really helpful, frankly.

Yeah, thank you so much, Red. That makes a lot of sense. And Sarah, when it comes to questions that recruiters might ask, maybe we can actually just touch on a few concrete ones and help the audience prepare in a real way. Would you like to share some common questions that they might get asked?

Yeah, totally. I would say common ones, like one that I usually ask all the time is, what interested you in open AI? That's kind of easy. I'm just really curious what got you to talk.

to me in the first place. And then how does AI impact your life today? Tell me about a time that you had to learn something completely new quickly.

And often I'll ask candidates, tell me about the biggest technical challenge that you've faced and how you walked through it. I'm not really looking for the technical challenge itself, but I want to know like how you approached it, the obstacles along the way and what you did to overcome them. Cause we want to understand like not just your skills, but how you think and adapt and really connect to our mission.

Thank you, Sarah. I love that. And one thing that we've also learned from economists in the forum and from our chief economist, Ronnie.

is that social skills, along with AI literacy, are becoming very, very important. And part of these adaptability and the ability to move something from ideation into a product really quickly is the ability to collaborate and communicate well. So that really makes a lot of sense to me of why you would need those skills here at OpenAI.

And Red, do you want to just elaborate a bit on those questions? Because I know you come at it from a little bit different of a perspective in your role.

Yeah, I think sometimes I'm meeting a candidate for the first time after they've had initial interviews. So they've gone through some process already. And I think some of the interesting questions to ask further along are things like, if you were to join.

team what would success feel like to you, right? What are those signals that you'd be capturing? And honestly the opposite can also be really helpful for the candidate and for us.

If we fast forward six months and you're working here and maybe you're feeling like this role isn't working out for you, what do you think the reason would be? Like that self-reflection and ability to talk through those things is actually really helpful and can really solidify a match or tell us that maybe it's not the right one.

And after all these years, Red, what does a successful career at OpenAI feel like to you? Like you don't know what happened for the last five years.

No, I mean, I think you will feel a reward in knowing that the technologies are advancing and that people are actually using them, that they bring something helpful, some value to them.

I mean, every day you go and you search on social media or the news about stories of people's lives being positively impacted by AI.

And I think you can kind of draw the line back to what your work is. You don't have to be Lukasz to have an impact. And I think that's a very important theme here at OpenAI, is everyone's role really does matter.

And so, yeah, I think that's what it feels like. Definitely.

And I actually, we just presented two

nights ago in the forum how people are actually using ChachiPT, a collaboration with Harvard and our economists, and more than 10 million people are using our technology across the globe.

So, Red, you've actually seen and felt the way that adoption has increased. You felt it from the inside. So, like, oh, my gosh, you actually talk about seeing people adopt the tools that we built.

And, Sarah, what does success in your role at OpenAI feel like after two years? I would say, well, I guess I'll talk about my role, but just kind of broadly, I think

trying other teams at OpenAI is really cool. We make internal mobility pretty seamless. Hiring managers love internal transfers into their teams. So I think getting curious enough, and you should contribute to multiple teams.

And oftentimes, if a team is going through a launch, other teams will shift over and help them through it so you can see what it would be like to contribute to that product.

So yeah, I think success is being able to say that you have shifted around other areas. I supported applied AI before this, which is our product and infrastructure teams. And now I'm on the research side. And so it's very cool to be able to just try.

other things out that you're curious about. I totally agree. So I don't know if you guys know this, but I started my time at OpenAI building the forum in the research org two and a half years ago. And I've been in global affairs for one year now and both were very exciting in their own ways. But I love the work that I'm doing now too. And we still, the community, many people that are here today watching us, they still plug into the research org. They've evaluated our models in the post-training process. So we still do all of that while now we're also more involved in storytelling and connecting with people across the world instead of keeping like the community.

The community now has, I think, 57,000 people in it. We started with 200 in the research org. So I agree, it's really rad to be, I'm so grateful for the time that I had in the research org because I also understand what we do so much better because of that. And it makes, it informs this role so much better.

So agree, Sarah, it's really cool to be able to move around and I think you can do so after six successful months or is it a year in your current role? Or maybe it's not a hard and fast rule.

Okay, one more question for you guys and then we're gonna turn to the audience because they have lots of questions for us.

So before we wrap up, Red, I'm just wondering, sorry to keep leaning on all your history, but five years is a really long time. It's pre-Chat GPT. So can you share a story that captures the human side of OpenAI?

Ooh, yeah. Well, first of all, I'll mention when I joined OpenAI, we had this goal that we were gonna be a company of 300 people. And I just mentioned that because I thought, okay, cool, what is that gonna look like? Is that possible? I was excited for it, but I didn't know what it would actually look like. Now we're past 3,000 globally. And so the culture of OpenAI has always been one of like,

creativity, a little weirdness, rigor, grit, and that has been consistent. I actually remember one of the early things that got me interested is I visited the office and there was a piano, and I thought, oh, interesting, and there was sheet music, there were books, there was a guitar in the corner, and I was like, who's doing, like, who's playing music here? What does that have to do with this?

And, of course, like, realizing, oh, people have a lot of different passions in life, including researchers, and there was this environment where people, like, were kind of encouraged to actually bring that because it all kind of interconnects, and so we've had pianos in the office for years since then, and actually some of our research projects and even breakthroughs were inspired by musical exploration.

generation, throwback to projects like Jukebox and Musenet, they kind of grew out of that spirit of wanting to apply language models to different domains. And I think it is actually a very real, tangible example of people bringing in their interests to develop AI and to also connect it to the rest of the world. So that's one that kind of always sticks in the back of my mind is whenever I see a piano, I think of that. Yes.

And I think you might be talking about Christine McLeavy. She's one of the people that definitely... For our audience... Well, you can see her playing the piano in Music as Math in the forum. But Christine came to us through the residency program, which is a program...

program where you don't necessarily have to have an AI researcher background. You obviously have to have some technical abilities, but we welcome a breadth of backgrounds. And Christine was a classical, a professional classical pianist. And that is what gave us the idea for the events. Music is math.

And then Jess Chang, who's on our security team, she came and she played the viola on the stage, you know, so the, and to this day, that was probably just one of the most beautiful, memorable experiences ever.

These people have, you know, they moonlight with these second careers and their creative passion.

So thank you for reminding me that.

Okay, team, we have now 20 minutes for audience Q&A, and there's so much. So let's dive in.

So the first one from Lorenzo Colombiana, board member at Wharton AI Studio asks, what advice would you give to a jack of all trades? As far as applying, I mean, I don't know, Red, I'll take this first, it's just like applying and then maybe if you want to go into kind of maybe interviewing, but if I'm looking at a resume with a with a jack of all trades, I do still want to see similarities and overlap with the job description.

description. So it's awesome that, you know, you're able to do all kinds of things, you have all kinds of skills. But I also want to make sure that, you know, it's it's the right fit for the role that that I'm hiring for. So being able to still match all all those skill sets that you have to some of the things that we're listing in the job description, I think is really helpful.

And also emphasizing, like, I want to know all those trades. And I want to know, like, metrics from those trades, I want to see what that is, because, like, that's awesome. And I think that those types of people do really well here. So that's kind of my advice as far as like, in the beginning of the process.

Thank you, Sarah.

Yeah. Red, do you have anything to add to that?

Yeah, I actually think it can be really tough to be someone who's got a lot of different skills across a wide range of areas, but you're trying to narrow that down to a particular job. To Sarah's point, there's a job description, right?

And so, you know, maybe practically speaking, you want to think about how do I market myself to that job? One easy way is, you know, think about how your resume presents and what information connects to that job. If it's unclear, if you can't draw lines that directly connect them, maybe you should think about rephrasing or emphasizing things that are a direct connection.

One other thing I'll say as well is that sometimes there are jobs that are very senior and very deep in a particular expertise, and if you're someone

has a lot of breadth, it can be hard to match that. So do not rule out earlier stage career opportunities in order to make the transition to something specific. It can actually be your key to getting in and getting ahead quickly.

Totally. And maybe I can add a little bit to that because I have a serious breadth of experience and I think it really took me a while to figure out, what do I really want to focus on, become an expert in, but yet all these other experiences inform it and imbue me with expertise and power in this specific place.

So one thing that I want to share is you do eventually

if you're coming specifically from an academic background, like a lot of people in the forum are, you might actually need to retool just a bit with some hard skill sets.

And then to Sarah's point, you have to be able to show that you've built something. And so the way I did that, because I spent 10 years in restaurants, managing very, very high volume restaurants and then events director of Michelin star and James Beard restaurants.

And then I got a degree in art history and then I ran arts organizations like Yerba Buena Center for the Arts and Counterpulse and did fundraising and programming curatorial.

So it's like, can you imagine in the beginning trying to apply to jobs?

I'm like, look, I've done all this stuff. I'm a powerhouse, but it was all over the place. But then I retooled a bit. I got data analytics and visualization, nanodegree, got a PMP to learn how to formally project manage. And then I offered my skills to the nonprofits I'd worked for pro bono to build a portfolio of data analysis and visualization and all sorts of other things that were really tailored to the types of jobs I was trying to get into.

And at that time, I knew I wanted to get into community building in AI. Sounds crazy, but I saw that AI was going to touch every single discipline. I knew community had to do with making people

feel like they had a place at the table. I knew how to do that, knew how to host events. And so all of that to say, you know, I had all those skills, but if I would have listed them all out and hadn't had the product to show or the hard skills to execute in a technical environment, I wouldn't have made the cut.

And to your point, Red, I started out at the very bottom. My first role in tech was at Scale AI and that was only four years ago, but then I got the opportunity once I built the community and built it well to come to OpenAI and build it.

So in the beginning, that might not feel good if you've been a director, but I can attest.

it's totally worthwhile. And in this environment, when you deliver, and you produce, and you collaborate well, there's no way you're going to stay in one place. The sky's the limit, right?

I really like that question. I think that's going to be helpful for a lot of people, for the Jack of All Trades profile.

And I don't know if you guys have heard this story, but one of our research residents back in time was a professional poker player. So it's kind of like the Rubik's Cube competition. And something that I've heard many times from technical folks in the forum.

you know, we somehow will be talking about some sort of leading edge research and it always kind of goes into how do you get a job? How do you prepare for a job at OpenAI? And a couple of things that they've said, especially for young people coming like right out of their undergrad or even graduate degrees, like how do you get experience? And like competitions, you know, math competitions, coding competitions, capture the flag competitions, and collaborative projects in code, in like public code bases. So that like, you just got to be scrappy and build things, even if no one's telling you to or paying you to in the beginning.

Totally. Yeah.

Okay. This one, friends, is from

Andrew Holtz, AI tech lead at EPAM Systems. They say, I'm curious, what's the most memorable thing you've seen in a candidate you've hired? So maybe Sarah, we can start with you and then we'll shift to Rhett.

Oh man, that's- Memorable. Yeah. I mean, I did already use the example of the candidate that taught themselves Google Glass SDK and Android in months like that. That was like, wow. So like anytime candidates can tell me like, oh yeah, I contributed to this open source project and I look at it and it's like thousands of stars. And so I would say, gosh, I think of like a specific example

But that makes sense. So they're like, oh, this is just my side project, my hobby. And then you look at it and you're like, oh my God, what is this person?

I think Sam recently talked about one of our researchers that we hired way before my time. Red, you might have been here. But he said that we found him on GitHub maybe, and he was not a classically trained AI engineer. So his language was different and the vantage point from which he approached problems was different. And that he taught us a whole new language, but he ended up being a superstar.

Do you know who Sam was talking about there? There are a few people, honestly.

Yeah, it's like that doesn't narrow it down much. I think, you know, frankly, like all of the people we hire have some unique quality or aspect to them. I will say I didn't hire this person, but something that I think was a fun example, we have someone here who built an incredible, basically art exhibit that combined AI and mechanical technical work and physical work for the annual Burning Man event.

Oh my gosh, who? Ben Bartlett. Oh my gosh, I didn't know that. Yes, and I actually randomly saw it on social media and I saw.

I saw the exhibit and I was like, wow, that's really cool. And someone posted it in our internal community. And I was like, oh, someone opened an app for that. Oh my gosh. I mean, kudos to them. And so we're doing like an internal talk about it now. But like that to me expresses like the type of creativity people bring that you want to see and support.

Oh my gosh, exactly. That kind of gives me goosebumps. And also how do people have time? What? That's what I, yeah, because I support the robotics group. So a lot of candidates I talk to will show me like robots that they've built. And they're like, oh yeah, but that's on the side.

Okay, so we don't sleep. Yeah, they're just tinkering with thing they like they'll break things on on purpose, just to rebuild so yeah I mentioned tinkers in the beginning but that's just like a quality that we really like to see. Oh, I love that. Thank you, Sarah.

Okay, back to the community grace Kim director, director of accessible for all says, though, for those of us making a career pivot and coming from outside the tech sector, would you recommend using a functional resume to demonstrate key qualifications for a role.

I'm not sure what a functional resume is you guys are going to have to prove. Yeah.

say, because I guess the assumption is because you're outside of tech, so it's like maybe not a ton of translation into this, but I think you would be surprised at actually the amount of overlap that you would see at a tech company. There's plenty of people here that are that didn't come from the tech industry.

So I would say definitely back to what Red and I mentioned earlier about kind of matching up to the job description and just listing similar topics that you see in the job description.

And I also wouldn't shy away, like I definitely encourage that transition.

Please don't feel like you can't, like, come to a tech company because you're, you know, are outside the tech industry. It's definitely an exciting change, and it happens a lot, and just matching up qualifications to the job description. It happens a lot more often than I'm certain people realize, and when I worked in arts orgs in San Francisco, it really, it did take a while to break into tech, because it really was a matter of being able to communicate how the skills map over, and then the other thing that we've talked about, Sarah and Red, you know, building a product or a team.

some sort of output, a concrete output that you could would have to show for yourself. So what one thing that I share, so I have a lot of friends that are teachers, a lot of friends that are artists and work in the arts or so I am sorry to be taking over all these ideas for how to get into tech if you don't work in tech.

One thing I just want to communicate is that I used to also not understand who was working in tech companies. I really thought it was an absolute bubble and like your parents must have worked in tech companies, you must be an engineer and I humbly admit how absolutely wrong I was.

As Sarah has mentioned, there are so many people working at OpenAI that are first-generation college graduates, that are first-generation immigrants, that, you know, I mean, have transitioned from liberal arts and social sciences.

And we also, by the way, have a policy team, and there are just so many different types of people at OpenAI, and they are absolutely lovely.

So first thing I just want to communicate about that is, if you don't feel yet like you see yourself reflected in us, you just haven't had enough touch points, because you're going to find people that are your people here, and don't let that be the reason that you don't apply.

Yeah. And then, um, another thing I just want to add that I tell my friends that try and transition are to take your resume and not just your resume, but like write out all the things that you've done, all the projects you've contributed to all of your experiences and use chat, GPT, take the job description and your resume and your cover letter and all the documents that describe who you are and the work you've done and ask chat, GPT, how it maps to that job description. So just like a little tip and I've seen it, I've seen it work.

Yeah, that's actually great advice. I say that too. I love that.

Red, do you have anything else to add there?

Um, I think, yeah, I agree with you that it can tech can definitely feel like you're trying.

break into something right like there's no door you just have to crack and I totally appreciate that like it is it is a growing industry and it's a successful one so it naturally attracts people who are excited about joining that.

I will say to add to your points community goes a long way and it's not about messaging every recruiter on LinkedIn and it's not about you know just throwing your resume wherever it'll go but you can be intentional about making connections going to events participating in things like this even to learn more to make those connections and I truly believe you know destiny has a way of working itself and so if it is going to work it will.

And those are some of the efforts I think you can put into it. There will be an opportunity for you, I promise.

Yes, Red. Yes, if you don't give up. Yeah. If you don't give up. You can't, you have to be OK with rejection. And that's just all it is.

If you've applied once and got rejected and then you said this is not for you, then this is not for you. But I totally agree with that, Red.

And social capital, it just takes that one person that's willing to say, OK, I'll give you a chance.

And for me, it was the VP of Marketing at Scale AI. And he had a different background. He believed in scrappy people. He believed in generalists. And he took a chance.

And that, you know, and it took, it took about a year, honestly, and probably like 120 rejections. So you just got to stick with it.

I so agree with that, Red, that like, there's, destiny has a way. And Mary Oliver said that she truly believes what is hers will find her. And that's so true.

Okay, one last question, and then we will let you guys get back to your lives, but I like this one. So what makes OpenAI's research environment distinct from academia or big tech in a way that retains these world-class researchers?

Red, do you want to go first?

Yeah. I mean, I think it goes back to what we said earlier, which is that there's a genuine ambition and collective interest in seeing AI technologies get ahead.

I actually read this recently from a researcher here who said, I think part of the journey at OpenAI in research is feeling like the world is missing something and we want to help bring it, that we can help advance humanity. It's not that we solve the world's problems, that's not the position we're in or trying to be in, but we think we can help other people do that in small and big parts.

And frankly, I think that mission is very clear to us and it has been.

very consistently since I've been here. The format, the products, the growth, all of that will change over time. That mission has been consistent. And I think, frankly, it helps us stand out from other companies out there who have successful research labs, but maybe have stumbled upon aligning themselves on who they want to be and who they are. It's really hard to fake that. And if you do, you can't do it for too long. And I think we have been very genuine about that over time. And the people who join us that are happy here, they feel that. They understand that. I can go into details, but I mean, frankly, I think that is the overarching theme. It's also why people come back to OpenAI. And so I think it's worth noting.

Thanks, Brad.

Yeah, I was gonna say, I was gonna say mission alignment like just people feeling really empowered by our mission excited about the mission and really wanting to contribute to it.

The other thing I think the research culture does pretty well is, is, like, we really like people that are working on hard projects that are not worried about failing. We've actually promoted people that failed on on a project because the work was so hard. So I think it's like it's this culture of like, please do the work it's it's really hard but if you're interested in it if you're excited about it if you're trying to solve it then yeah we want you to be putting

your effort towards it. So I think it's like just enabling people to work on something that they're curious about, something that excites them. I think our research org does a good job of that. In my orientation, Sam said that he wants us to do the best work of our lives here, and that really motivated me. And I always think, you know, in building community, what motivates people? And you can, I think, in some ways, simplify it into three different categories.

Achievement. So that's definitely something that motivates the people that work at OpenAI. They want to do something huge, which is create a tool that really helps solve some of humanity's

hardest problems. Another category is affiliation. And I am not a technical contributor. So I'll just speak on behalf of, I'm in the global affairs team and policy, but I feel so motivated and reinvigorated on a weekly and monthly basis.

Not to, I mean, things do get, they're hard here. Like Sarah, we fail and it's hard. It's hard on our egos and it's hard to spend so much time doing something that might not have worked out, but I'm so motivated by affiliation, being a part of a team, like rooting for the people that you're working with.

I was really excited to host you guys today.

and share your beautiful work. So I think that that's another key. And I don't know if that actually motivates technical contributors as much as non-technical contributors. Like, wouldn't make that assumption. But what do you guys feel about that piece? Like the affiliation and collaboration part. Does that, is that part of what keeps you around?

Totally. I feel like everyone's very smart here. I feel lucky to be working with everyone. And yeah, that's a big part of it.

Yeah.

Well guys, we did it. We're at time. I just want to say it was really sweet and fun.

Sarah's like, I don't think.

think that we have enough to talk about. I was like, oh, we definitely, definitely do. We definitely do. I hope this is just the beginning for all of our collaborations, Sarah and Red, and your members in the community.

So if you're a member in the forum, not a guest, but at the member level, that you can actually DM any other members in the community. So if there are questions that weren't answered, reach out to Red and Sarah. Please don't hold them accountable to any sort of 24-hour deliverable timeline to respond. But I'm sure they would love to hear from you.

And Sarah, thank you so much for helping us bring this to fruition. You really pulled us across the finish line here.

And Red, thank you for all of your heart and all of your soul and the depth of wisdom that you brought to both of our careers at the Frontier event. And I really hope to host you guys again soon. Thank you so much. This is fun. Yay. Good night, guys. Bye.

Oh, my gosh. I just really love the people team here. And I also really hope that if you're in the audience and you're interested or if you were thinking about working at OpenAI, we just inspired you to go ahead and take the leap. And I hope you feel more comfortable now and confident crafting your resume. At the wake of the last event, we created a one pager that was just a broad overview of.

tips from recruiters, so we're going to drop that in the chat now. You can look at that.

You can also see our last event, Careers at the Frontier, the first one with the head of the recruiting team, Joaquin, and there are some really, really awesome tips there too.

And then we'll continue working on this series together. In the meantime, we have other cool events on the horizon.

One is with the California Teacher of the Year, Casey Cooney, who uses AI to make good teaching even better.

So if you are a K through 12 teacher, all of your students are using ChatGPT, and you want to come up with some innovative, creative ways to also leverage ChatGPT to make your teaching even more impactful than it was.

to begin with and build your capacity, maybe give you a few more hours in the week to do the work that you love to do and let AI do the work you don't wanna do, please come to that event. It's September 25th, next week, 5 p.m. Pacific time. Casey is an awesome speaker and I can't wait to host him in the forum.

And then in October, Lukasz Kaiser, who we discussed earlier, a researcher at OpenAI, has been around for a while. He's written many papers that are foundational to contemporary AI models, such as our reasoning models at OpenAI. We're hosting him October 8th at 12 p.m. It's a lunchtime talk, so you can drop in while you're having your snack.

and it's virtual, it's open to the public, everyone is invited. I really hope to see you all.

Thank you everybody for being here. Happy Thursday, have a beautiful night and I will see you next week.