Inside Video Game Giant Electronic Arts’ Multi-Year Deal With Stability AI: ‘One of My Biggest Worries Is That We Become Observers’
Following Electronic Arts‘ big go-private deal announcement last fall, the video game publisher made another splashy move: revealing a multi-year partnership with the James Cameron-backed Stability AI.
A few months into the teamup, which also included “The Sims,” “Battlefield 6” and “Madden NFL” maker EA taking a strategic investment in Stability AI during its latest round of funding in October, and the two companies still have quite a ways to go before gamers see the results of the pact.
It’s all still in the very early stages, though EA chief strategy officer Mihir Vaidya told Variety in a Jan. 16 interview that conversations between EA and Stability AI about a partnership started well before EA was being courted by Saudi Arabia’s Public Investment Fund (PIF), Silver Lake and Jared Kushner’s Affinity Partners for a $55 billion go-private move.
“One thing we’re super proud of at EA is that, over the course of our history, I think we’ve taken great pride in our heritage as electronic artists,” Vaidya said. “And to be a great electronic artist, you need to have the very best technology and tools available to you. That’s been a core principle of how we’ve kind of thought about technology, is we always are looking out there for the very best technology and solutions for our creatives. Over the course of the last few years, naturally, we’ve been exploring like, what are the possibilities, as it relates to AI, from a tools and technology standpoint, to make available to our people?”
Vaidya said multiple AI tech companies were considered and EA remains in ongoing discussions for more opportunities, but the Stability AI deal was closed because of what it could offer that other companies couldn’t.
“As we went along the journey, what really stood out to us and why we ultimately decided to partner with Stability was one of the things that we talked to all of our developers was this notion that we can’t control the tools that are available,” Vaidya said. “Certainly there’s gaps in capabilities, but even when the tools were capable, they’re not controllable. That’s incredibly frustrating for anyone in order to see their utility in the context of their work. And when we unpack, well, why are they not controllable? Turns out a lot of foundation models, generic, general foundation models, they’re black box. They’re incredibly capable across a range of tasks, but not very good at very specific tasks. So how do you make them more capable of specific tasks? Turns out you need deep customization. Customization means you have to be able to unpack the model, understand where representations of different concepts and capabilities lie, and then be able to preserve those representations but adapt them at the same time to your specific needs. It’s not a trivial task. It’s very, very difficult. And you need a partner that is able to do that. And Stability was not only able to do that, it was willing to do that.”
Regardless of how generative AI can help a developer, it’s a scary addition to the industry, as many developers believe their jobs will become easier, but ultimately obsolete, with the progression of how AI is used.
Stability AI CEO Prem Akkaraju refutes this point by invoking the economic lump of labor fallacy.
“If you add or introduce technology into making something more efficient what ends up happening is the workforce actually expands,” Akkaraju said. “Labor is not fixed, and so if it was fixed, you’re absolutely right, that would truly eliminate the jobs.
Akkaraju offered the example of the invention of the ATM machine: “When those were rolled out in America in the 80s, every bank teller in America was up in arms, and they said, ‘This is what we do. We give withdrawals, we take deposits, and you’re taking our jobs away with this machine.’ And there’s more bank teller jobs today than in the ’80s, if you go point to point. And why is that? Because the banks did an analysis, and they went from 30 people that they needed for each location down to 13, and they did what came naturally, and they expanded. Now, there’s a bank around every single corner. And the average salary for bank tellers went up because their skills went up.”
Vaidya says he “absolutely” does not want to “trivialize” the “fears and the conversation around this notion of replacement,” because “there are real concerns.” However, he believes EA’s plan for the use of generative AI will lead to more of a change in what the available jobs look like than a great reduction in workforce.
“AI can play a role in augmenting a particular task, what’s naturally going to happen is there’s going to have to be a recomposition of certain tasks,” Vaidya said. “I think we have to get to that level in order to truly appreciate the impact on the workforce. Over the course of time, there’s naturally new job families that also emerge. And especially in the creative industries, more creatives that are actually part of the industry because the reality is that entertainment is certainly not a zero-sum game. It’s a pie that expands over time. There’s almost an insatiable demand for entertainment since the beginning of time. And so I think if these tools and technologies enable us to create, to increase velocity, to increase quality, to increase volume, to increase scope — demand increases. And as demand increases, it supports a bigger creative population. But the shape of that population might be different, and not necessarily in the wholesale case, but in select cases, because of the fact that AI affects things at the task level, which results in a little bit of a recomposition of what are the collection of tasks in any job.
Vaidya added: “I think the most important thing to do, if this is the case, is to be on the front foot and lead the conversation, because otherwise what happens is this notion of disruption kicks in because you’re not ready for that change. And as a result, this is especially the case if that disruption comes from folks that are not insiders in the industry, but outside of it. And that’s one of my biggest worries is that we become observers, as opposed to participants and leaders in this conversation, because market forces will lead to AI being deployed at scale. And there are many players outside of the creative industry today that are trying to leverage AI in the space of sport and entertainment.”
A few months into the teamup, which also included “The Sims,” “Battlefield 6” and “Madden NFL” maker EA taking a strategic investment in Stability AI during its latest round of funding in October, and the two companies still have quite a ways to go before gamers see the results of the pact.
It’s all still in the very early stages, though EA chief strategy officer Mihir Vaidya told Variety in a Jan. 16 interview that conversations between EA and Stability AI about a partnership started well before EA was being courted by Saudi Arabia’s Public Investment Fund (PIF), Silver Lake and Jared Kushner’s Affinity Partners for a $55 billion go-private move.
“One thing we’re super proud of at EA is that, over the course of our history, I think we’ve taken great pride in our heritage as electronic artists,” Vaidya said. “And to be a great electronic artist, you need to have the very best technology and tools available to you. That’s been a core principle of how we’ve kind of thought about technology, is we always are looking out there for the very best technology and solutions for our creatives. Over the course of the last few years, naturally, we’ve been exploring like, what are the possibilities, as it relates to AI, from a tools and technology standpoint, to make available to our people?”
Vaidya said multiple AI tech companies were considered and EA remains in ongoing discussions for more opportunities, but the Stability AI deal was closed because of what it could offer that other companies couldn’t.
“As we went along the journey, what really stood out to us and why we ultimately decided to partner with Stability was one of the things that we talked to all of our developers was this notion that we can’t control the tools that are available,” Vaidya said. “Certainly there’s gaps in capabilities, but even when the tools were capable, they’re not controllable. That’s incredibly frustrating for anyone in order to see their utility in the context of their work. And when we unpack, well, why are they not controllable? Turns out a lot of foundation models, generic, general foundation models, they’re black box. They’re incredibly capable across a range of tasks, but not very good at very specific tasks. So how do you make them more capable of specific tasks? Turns out you need deep customization. Customization means you have to be able to unpack the model, understand where representations of different concepts and capabilities lie, and then be able to preserve those representations but adapt them at the same time to your specific needs. It’s not a trivial task. It’s very, very difficult. And you need a partner that is able to do that. And Stability was not only able to do that, it was willing to do that.”
Regardless of how generative AI can help a developer, it’s a scary addition to the industry, as many developers believe their jobs will become easier, but ultimately obsolete, with the progression of how AI is used.
Stability AI CEO Prem Akkaraju refutes this point by invoking the economic lump of labor fallacy.
“If you add or introduce technology into making something more efficient what ends up happening is the workforce actually expands,” Akkaraju said. “Labor is not fixed, and so if it was fixed, you’re absolutely right, that would truly eliminate the jobs.
Akkaraju offered the example of the invention of the ATM machine: “When those were rolled out in America in the 80s, every bank teller in America was up in arms, and they said, ‘This is what we do. We give withdrawals, we take deposits, and you’re taking our jobs away with this machine.’ And there’s more bank teller jobs today than in the ’80s, if you go point to point. And why is that? Because the banks did an analysis, and they went from 30 people that they needed for each location down to 13, and they did what came naturally, and they expanded. Now, there’s a bank around every single corner. And the average salary for bank tellers went up because their skills went up.”
Vaidya says he “absolutely” does not want to “trivialize” the “fears and the conversation around this notion of replacement,” because “there are real concerns.” However, he believes EA’s plan for the use of generative AI will lead to more of a change in what the available jobs look like than a great reduction in workforce.
“AI can play a role in augmenting a particular task, what’s naturally going to happen is there’s going to have to be a recomposition of certain tasks,” Vaidya said. “I think we have to get to that level in order to truly appreciate the impact on the workforce. Over the course of time, there’s naturally new job families that also emerge. And especially in the creative industries, more creatives that are actually part of the industry because the reality is that entertainment is certainly not a zero-sum game. It’s a pie that expands over time. There’s almost an insatiable demand for entertainment since the beginning of time. And so I think if these tools and technologies enable us to create, to increase velocity, to increase quality, to increase volume, to increase scope — demand increases. And as demand increases, it supports a bigger creative population. But the shape of that population might be different, and not necessarily in the wholesale case, but in select cases, because of the fact that AI affects things at the task level, which results in a little bit of a recomposition of what are the collection of tasks in any job.
Vaidya added: “I think the most important thing to do, if this is the case, is to be on the front foot and lead the conversation, because otherwise what happens is this notion of disruption kicks in because you’re not ready for that change. And as a result, this is especially the case if that disruption comes from folks that are not insiders in the industry, but outside of it. And that’s one of my biggest worries is that we become observers, as opposed to participants and leaders in this conversation, because market forces will lead to AI being deployed at scale. And there are many players outside of the creative industry today that are trying to leverage AI in the space of sport and entertainment.”