AI Recruiting at NVIDIA, Pika, and Eleven Labs: How to Build Exceptional Teams

The podcast episode opens by addressing the urgent talent challenges facing AI companies. Host Michael Mignano, a partner with Lightseed, introduces the topic of hiring trends in the generative AI era.

The segment highlights the scarcity and expense of AI talent, noting that "Bayon Company predicts that half of global AI jobs will go unfilled by 2027" and "44% of business executives say this talent shortage will make it harder for them to deploy AI."

The episode promises insights from leaders at Pika, 11 Labs, NVIDIA, and Every, plus input from Lightseed's VP of AI Talent Amy Anton, to address how to effectively staff an AI team in 2025.

Timestamp: [00:00-01:27]

The segment explores how successful AI companies are differentiating through their team composition, featuring insights from Demi Guo (co-founder/CEO of Pika) and Mati Staniszewski (co-founder of 11 Labs).

Demi Guo emphasizes the dual importance of technical and creative talent: >"For us, I think first of all like we tried to build a very um the best technical team both on research engineering sites... our founding research team members came from like the top AI lab like both from like the top AI industry labs such as Google Facebook AI and also the top um academic schools like Stanford MIT. At the same time i think we for us like art is also as important as science when it comes to building the team but also building the model so we have a lot of like team members are from creative background so who are filmmakers um artists, they provide a creative lens in the model development process."

Mati confirms this approach, adding the importance of focus: >"I 100% echo the team piece... I'm lucky to have a co-founder who has done a lot of the research before and has been able to assemble an incredible team of researchers so 100% that and in our case it's also being able to focus that team on a very specific set of goals."

The host observes that as software creation becomes easier, judgment and taste become increasingly valuable differentiators, suggesting more creative talent will be needed within AI companies.

Timestamp: [01:27-3:36]

The segment introduces Dan Shipper, founder of Every, "the first multimodal media company" that publishes various content types with AI assistance. Shipper has written about the transition from a knowledge economy to an "allocation economy."

In this new paradigm, the primary responsibility for managers will be allocating resources - choosing the optimal mix of AI tools and human labor to achieve better, faster, and cheaper results. This shift fundamentally changes what companies should look for in their hires.

"Given the many hyperintelligent AI tools at our disposal, the primary job for future managers will be allocating these resources, choosing the right mix of digital and human labor to get work done faster, better, and cheaper."

The allocation economy will require fewer subject matter experts and more multi-talented generalists who can work effectively across different domains and tools.

Timestamp: [3:36-4:35]

Dan Shipper describes how Every prioritizes people with diverse skill sets who can adapt to rapidly changing tools and workflows. He shares examples of team members who embody this multi-dimensional approach.

"A lot of the people that work for every have that sort of multi-dimensional skill set... we have this writer, Katie, when we have engineers write sometimes they're good writers sometimes they're not, and she does a lot of co-writing with them to help them get out their ideas. But she's also using Lovable all the time to build little tools for herself."

Shipper highlights another example: >"We have another writer, her name is Ria, who's doing the same thing, like building custom GPTs. And all these people are now curious and interested, and they sort of see how powerful it is. And once you get them going, they're like 'oh my god this is amazing.'"

This emphasis on multi-skilled talent is particularly valuable in the context of AI tools, which empower generalists to accomplish tasks that previously required specialists.

Timestamp: [4:35-5:19]

Dan Shipper explains Every's innovative approach to fostering creativity and experimentation through a quarterly practice called "Think Week."

"One of the things we do every quarter is we do this thing called think week where we don't publish anything new, we don't do any meetings. Every day there's a theme, but the idea of think week is to sort of recognize that most of the time in a startup you're spending time being sort of very reactionary, and you're just like constantly under fire trying to make sure things are not breaking."

He contrasts this reactive mode with the conditions needed for creative breakthroughs: >"The best creative work comes from a different sort of place where you're not reacting to your circumstances. You are proactively kind of playing around and following that thing that you just are psyched about. And think week is really about getting back in touch with that."

During Think Week, the team pauses regular operations to encourage exploration: >"One of the days this time we did like a day where it was like experiment with a new tool that you've been meaning to experiment with but you haven't. And that's how Katie started using lovable."

Shipper emphasizes this approach as critical for businesses navigating the AI landscape: >"That's actually so important for businesses right now and we do this with our consulting arm. This is like one of the big recommendations - find space to play."

Timestamp: [5:19-6:45]

Shipper articulates the fundamental tension that every worker faces in an AI-powered workplace - the choice between immediate productivity and experimental learning.

"It doesn't matter if you're employee, executive, whatever, founder - you have this decision you have to make every day, which is: do I do things the way that I know how and get them done, and then I have so much work to do that like if I worked the way I know how for 20 hours a day it still wouldn't be enough, but I can just get it done and I can go home... or do I spend like 2 hours playing around with this new tool that may not work and probably won't, and then I will have to go and do it the old way I knew how anyway?"

He acknowledges the natural tendency toward the familiar: >"Unless you're someone with a really curious early adopter mindset, everyone is going to do the first one and they're just going to get their work done."

Shipper argues companies need to deliberately create space for experimentation: >"You just need to be given the space... We do that as a concerted regular practice at Every and it also works for other types of companies too."

The host concludes that this approach leads to greater creativity and innovation: >"Writers who build tools, builders who edit podcasts, everyone experiments. I think we're going to see a lot more experimentation going on inside of startups as the models just give us more and more capabilities. They're bringing out more creativity in all of us, and I think that creativity is going to translate to newer, faster, and more differentiated products."

Timestamp: [6:45-8:04]

• AI talent is scarce and expensive, with predictions that half of global AI jobs will remain unfilled by 2027
• Successful AI companies like Pika and 11 Labs blend top technical talent with creative professionals who provide aesthetic judgment
• The economy is shifting from knowledge-based to allocation-based, where the primary skill is distributing work optimally between humans and AI
• Multi-dimensional generalists who can work across disciplines are becoming more valuable than narrow specialists
• Creating deliberate space for experimentation (like Every's "Think Week") is critical for innovation in AI companies
• Workers face a daily choice between immediate productivity and experimental learning with new tools
• AI tools are enhancing creativity across roles, leading to more innovative products and workflows
• The best teams include people who are willing to constantly experiment with new ways of working

Timestamp: [00:00-8:04]

Companies:

• Lightseed - The venture capital firm where host Michael Mignano is a partner
• Pika - AI idea-to-video platform co-founded by Demi Guo
• 11 Labs - Audio AI company co-founded by Mati Staniszewski
• NVIDIA - Major AI hardware company mentioned as featuring in the episode
• Every - "The first multimodal media company" founded by Dan Shipper that publishes various content types with AI assistance
• Bayon Company - Research firm cited for prediction about unfilled AI jobs

People:

• Michael Mignano - Host of "Generative Now" and partner at Lightseed
• Demi Guo - Co-founder and CEO of Pika (idea-to-video platform)
• Mati Staniszewski - Co-founder of 11 Labs (audio AI company)
• Dan Shipper - Founder of Every (multimodal media company)
• Amy Anton - VP of AI Talent at Lightseed
• Bill Dally - Chief Scientist of NVIDIA (mentioned as appearing later in the episode)

Concepts:

• Allocation Economy - Dan Shipper's term for the economic shift where value comes from allocating resources (human and AI) optimally
• Think Week - Every's quarterly practice of pausing regular work to encourage exploration and creativity
• Lovable - App-building tool mentioned as being used by writers at Every

Timestamp: [00:00-8:04]

The segment explores how to effectively scale AI teams, featuring insights from Bill Daly, who grew NVIDIA's research team from a small group to hundreds of PhD researchers.

Host Michael Mignano introduces NVIDIA as the provider of "the core GPU technology powering all of our favorite AI startups" that has "become one of the most important companies on the planet in part due to the incredible depth of scientific talent it has amassed."

Bill Daly shares his journey at NVIDIA: >"I came to NVIDIA in 2009 and inherited a team of you know I think it was like about 15 people most of whom were doing um ray tracing um you know computer graphics. From there, created groups doing architecture and circuits and doing AI."

Daly emphasizes the recruiting flywheel effect: >"When we first started in any given area it was very hard because no one wants to come to a place where they're the only one doing something, but by getting some really good people to anchor each place and then hiring really good people it then becomes easier to recruit talent because people like to join a team where there are other fun people to talk to and everybody is as smart as you are."

He stresses the critical importance of maintaining high standards: >"We found that we had to really set the bar high and hold it there. If we were to let that bar drop and start hiring mediocre people that would beget hiring more mediocre people."

Timestamp: [08:07-10:04]

Bill Daly explains how NVIDIA creates an environment that not only attracts top talent but also retains them, highlighting the importance of workplace culture in building AI teams.

"We try to create an environment where people like to be. We have very little turnover - people come and they stay because they get to do what they want to do. We have the resources to do fun experiments, they get to work with fun people, and they get to have an enormous amount of impact."

Daly highlights NVIDIA's unique value proposition for researchers: >"One great thing about Nvidia is because we supply the whole industry, if you develop whether it's a piece of new hardware for AI or a new type of model, a new training technique, it winds up benefiting everybody, benefiting the whole world. Whereas in some of the people we're competing with for talent, if they develop something then their company will use that but it won't be spread as widely as the things that we develop."

The host notes this approach is common among foundational technology companies, emphasizing that "A-level players attract other A-level players" and the importance of scaling headcount without scaling complexity.

He shares a critical insight about organizational growth: "There's a general rule of thumb among startup CEOs and founders that at each 50 person interval, the culture breaks down, the process breaks down, and you need to start over from scratch."

Timestamp: [10:04-10:54]

Amy Anton, Lightspeed's VP of AI talent, shares insights on what makes effective AI teams, emphasizing that the fundamentals of strong team building apply regardless of industry.

"What we look for when we're thinking about building strong AI teams is largely similar to what we think about for building strong teams period. I think especially in the beginning, there's a special uniqueness to people who not only have AI skills and experience but who also demonstrate high agency, who can be product thinkers, who may have displayed initiative to push inside the companies where they worked previously."

Anton highlights the importance of clear vision: >"We look for clarity of thought and vision. What is it that you want to do and why? Have you thought about the dynamics of the industry or the product or the vision and any kind of competing factors that may be at play in the market?"

She emphasizes the critical quality of charisma in early founders and leadership: >"People who are most successful are able to be talent magnets. They need to be able to recruit, they need to be able to retain top talent as their company grows, and they need to be able to raise money as well."

Timestamp: [10:54-13:25]

Amy Anton addresses the hype around AI and provides pragmatic advice about assessing actual talent needs versus chasing big names in the field.

"We're in a world right now where AI is such a sexy term and people feel like they need to weave AI into their company in some way for people to take them seriously or to attract talent or funding. What I am encouraging people to do is to be really crystal clear about what their needs are."

She challenges the assumption that companies need to hire the most renowned AI researchers: >"I frequently get asked the question 'how do we hire these people when they're so difficult?' And while I would say a lot of those folks aren't unreachable, the people who actually need a more junior person or somebody who isn't as kind of well-known or famous or who doesn't have the traditional credentials would be just as effective and potentially even more effective for the work that you're doing as a kind of earlier stage company."

Anton emphasizes tailoring hiring to specific company needs: >"I would definitely just encourage people to think about exactly what their needs are, have that conversation with whomever it is that they're working with to staff those teams. I think the answer could vary quite substantially depending on the specific needs of the company."

Timestamp: [13:25-13:46]

Anton tackles what she calls "the million-dollar question" of whether to staff quickly or hold out for the perfect candidate, offering a balanced approach to this common dilemma.

"Do you staff up quickly or do you hold out for the perfect unicorn? There is always tension in the system around these tradeoffs."

She recommends prioritizing based on role significance: >"My best advice to people is always kind of think about the criticality of the role that you're hiring. Of course a leadership position or if it's a role that's setting the direction in any way or is potentially a culture carrier, figure out what you need and what are the non-negotiables and don't negotiate with those."

Anton warns against unrealistic expectations: >"It's important also to recognize are you looking for something that doesn't exist. Are you looking for a purple unicorn? And you may be able to find one, but then can you hire them? Can you keep them? Can you compensate them?"

She encourages broadening search parameters: >"The bigger that you can increase your aperture of what the talent pool looks like, the better. And I think it's really important, depending on where you are in the journey of your company, to not let perfect be the enemy of the good."

Timestamp: [13:46-14:55]

Amy Anton emphasizes the critical importance of getting the leadership team right from the beginning, as they set the trajectory for the entire organization.

"It is really important that you get your leadership team in place right up front because everything sort of trickles from there. They're the culture carriers, they're the people who are setting the research direction, they're the people who are setting the product vision, they're the people putting in place all of the structures that the company will then kind of sit on top of."

She stresses the importance of thoughtful leadership selection: >"Being really thoughtful about your leadership team in the early stages I think will just pay dividends over time, really kind of empowering those people to lead and to hire and kind of relinquish control as a CEO or founder."

Anton highlights the motivation of early employees: >"A lot of people who join early stage companies are really excited to be a part of that growth and to be a really integral part of what happens with that company."

She advocates for distributed leadership: >"Really trying to empower people to own various parts - you have more brains thinking about more problems in different ways and you probably come up with more creative solutions with more brain power. Hire the right people in the beginning and then let them lead and get out of their way."

Timestamp: [14:55-16:03]

Anton discusses the importance of regularly reassessing whether your team members remain the right fit as your company grows and evolves.

"Hold yourself accountable for looking at the people who you have over time and making sure that they are still the right people as the company grows. I've certainly seen people who are hired for a company at 200 people and they may have wonderful experience to lead or to drive various parts of that company forward at 200 people. When that company gets to be 5,000 people, maybe it's expanded globally or across different regions, with the best of intentions, those people are not the right people to lead the company into the next phase."

She emphasizes the cascading impact of mismatched roles: >"Holding yourself accountable for really looking at who is still here and who should continue on in the company over time will help, because if you have someone who's in the wrong role or the wrong stage of the company, that really can trickle down and have some pretty negative impacts across the organization."

The host shares a common pitfall: >"One pitfall I've observed often is companies making up roles just to justify hiring somebody who's super impressive. What I've seen instead that works better is being really specific and intentional about the roles you need to hire for and then going and finding the exact person that matches that role, not the other way around."

Timestamp: [16:03-17:19]

The host summarizes the key characteristics of successful AI teams, emphasizing intentionality and adaptability.

"To wrap up, here's what the best AI teams have in common: they're not just technically great, they're intentionally built and they're constantly evolving. Hire for range, hold the bar, build your team like you build your product because in this space, your team is the product."

The episode concludes with thanks to the listeners and information about how to follow Lightspeed and rate the show, noting that "Generative Now is produced by Lightspeed in partnership with Pond People."

Timestamp: [17:19-18:01]

• Successful AI team scaling requires establishing a high bar for talent and never lowering it
• Creating an environment where talented people can do meaningful work is crucial for retention
• The culture and processes of a company tend to break down at every 50-person interval of growth
• Effective AI leadership requires both technical expertise and "high agency" product thinking
• Companies should clarify their actual AI talent needs rather than chasing famous researchers
• More junior or less renowned AI talent can often be as effective or more effective for early-stage companies
• For critical roles that set direction, know your non-negotiables and don't compromise on them
• Avoid looking for "purple unicorns" - candidates who have an unrealistic combination of skills
• Leadership team selection is foundational as they establish culture and direction
• Teams should evolve as companies scale - the right person at 200 employees may not be right at 5,000
• Don't create roles to fit impressive candidates; instead, define roles first, then find the right people
• The best AI teams aren't just technically excellent but are intentionally built and continuously evolving

Timestamp: [08:07-18:01]

Companies:

• NVIDIA - Major AI hardware company where Bill Daly grew the research team from 15 to 400 PhD researchers
• Lightspeed - Venture capital firm; Amy Anton serves as VP of AI talent
• Pond People - Production partner for the "Generative Now" podcast

People:

• Bill Daly - Joined NVIDIA in 2009, grew research team from 15 to approximately 400 researchers
• Amy Anton - VP of AI Talent at Lightspeed, provides insights on effective team building
• Michael Mignano - Host of "Generative Now" and partner at Lightspeed

Concepts:

• 50-person interval rule - The observation that company culture and processes tend to break down at every 50-person growth interval
• Purple unicorn - Term for an unrealistic candidate profile with an impossible combination of skills and attributes
• Talent magnet - Leaders who can effectively attract and retain top talent in competitive fields
• Culture carriers - Key team members who establish and maintain company values and practices

Timestamp: [08:07-18:01]