OpenAI COO Brad Lightcap Talks What's Next For OpenAI

Summary

Podcast Summary

Podcast: Bloomberg Talks
Episode: OpenAI COO Brad Lightcap Talks What's Next For OpenAI
Date: October 7, 2025
Host: Bloomberg (with Ed Ludlow)
Guest: Brad Lightcap, COO, OpenAI

---

Overview

In this episode, Ed Ludlow interviews Brad Lightcap, COO of OpenAI, live from OpenAI’s annual Developers Day—a pivotal day marked by major product announcements and a substantial infrastructure deal with AMD. The discussion explores OpenAI’s latest features, the integration of third-party apps into ChatGPT, the company’s strategic focus on both consumers and enterprises, infrastructure challenges, and how OpenAI measures its success.

---

Key Topics & Discussion Points

1. OpenAI’s Landmark Collaboration with AMD

• Context: OpenAI has signed a major deal with AMD to enhance its infrastructure and support massive demand for AI services.
  • “We are tremendously compute constrained...the AMD deal we’re excited about being, you know, directional a way for us to do that.” —Brad Lightcap [06:54–07:52]

---

2. Introducing Third-Party App Integration in ChatGPT

• New Capabilities: Users can now access third-party apps (e.g., Spotify, Figma) inside ChatGPT, unlocking deeper, contextual experiences and affecting markets.
  • "It's the contextual aspect of I'm doing X or I need Y...that allows you now to kind of use ChatGPT to solve that higher level task and then also integrate apps contextually to solve those specific problems.” —Brad Lightcap [01:38–02:27]
• Ecosystem Evolution: Builds on the original “Plugins” feature; now enhanced for richer app integration.
• Market Impact: Stocks of partner companies moved significantly based on announcements—demonstrates ChatGPT’s influence [00:29–01:06].

---

3. ChatGPT as an OS or ‘Super Assistant’

• Vision: Rather than being just a chatbot, OpenAI wants ChatGPT to become a personalized, productivity-enhancing super assistant.
  • “We never set out to build a chat bot. We always wanted to build something that was really true to you and what your preferences are, what your goals are that could actually help you achieve more.” —Brad Lightcap [02:46–03:16]
• Comparison to Mobile: Transition to new interfaces draws parallels to the shift from desktop to mobile apps.

---

4. Developer & Partner Dynamics

• Partner Buy-in: Companies are enthusiastic to build for new interfaces rather than worrying about losing direct traffic.
  • “Mostly people are really focused on building into new interfaces. Right? This is just like mobile in some sense...” —Brad Lightcap [03:25–03:52]
• Revenue Models: Economic models for app integration are still evolving.
  • “We're going to figure out the economics of this over time...we take this very experimental mindset…” —Brad Lightcap [03:58–04:21]

---

5. User Growth & Subscriptions

• Milestone: ChatGPT now has 800 million active weekly users [04:21–04:44].
• Subscription Model Success: Surpassed internal expectations for paying users, defying the belief that consumers rarely pay for software.
  • “You know, it’s surpassed where my expectations frankly were. People have this kind of conception that consumers tend to not pay for software...it's a healthy amount and more every day.” —Brad Lightcap [04:44–05:19]

---

6. Enterprise Strategy & Agent Technology

• Enterprise Focus: Tools for businesses to build “agents” (automation/custom AIs) visually and intuitively—2025 framed as the “year of agents.”
  • “Today's announcements actually I think target what are an important set of use cases for the enterprise...things like agent builder, allow enterprises to be able to build agentic experiences, powerful agent experiences on the go...” —Brad Lightcap [05:33–06:26]
• Speed of Development: Lead times for software have dramatically reduced, allowing for more rapid iteration [06:03–06:26].

---

7. Scale and the Compute Bottleneck

• API Load: OpenAI’s API now handles more than 6 billion tokens per minute, underscoring the need for robust infrastructure ([06:26–06:54]).
• Sora Launch: New app Sora’s rollout is limited by compute resources; scaling hinges on infrastructure expansion.

---

8. Measuring Success: Tokens Processed over Valuation

• Internal Metric: OpenAI focuses on tokens processed (usage/utility) as their main success metric, not just valuation figures.
  • “For me, it’s actually kind of a metric that we, we talked about is, is tokens...that is the purest for me, the kind of essence of utility is that consumption metric.” —Brad Lightcap [08:12–08:56]
• Growth Patterns: Tools like Codex have seen 10x usage growth, reflecting increasing utility across domains.

---

Notable Quotes & Memorable Moments

• On App Ecosystem:

  “This is just like mobile in some sense...there’s an opportunity for builders to create entirely new applications that are even native to chatbots.”
  — Brad Lightcap [03:25]

• On Product Vision:

  “We never set out to build a chat bot. We always wanted to build something that was really true to you and what your preferences are…”
  — Brad Lightcap [02:46]

• On Metrics That Matter:

  “If that number [tokens used] is going up, it means people are using us for more things. And that's the ultimate goal.”
  — Brad Lightcap [08:12–08:56]

• On Demand/Compute:

  “We see there are multiples of demand that are latent and untapped from what we have today...the AMD deal we're excited about being, you know, directional a way for us to do that.”
  — Brad Lightcap [06:54–07:52]

---

Timestamps for Important Segments

• [01:06] — ChatGPT integrates third-party apps and implications for users/markets
• [01:38–02:27] — Brad Lightcap explains the motivation behind app integrations
• [02:46–03:16] — Vision for ChatGPT as a ‘super assistant’
• [03:25–03:52] — Partners’ perspectives on integration and ecosystem analogy to mobile
• [03:58–04:21] — Evolving economics and revenue models for developers
• [04:21–04:44] — Active user milestone: 800 million weekly
• [05:33–06:26] — Enterprise strategy, agent technology, and software development acceleration
• [06:26–06:54] — API scaling, AMD deal, and compute limitations
• [08:12–08:56] — How OpenAI measures success: tokens processed over valuation

---

Tone & Language

Brad Lightcap’s language is optimistic, experimental, and candid—emphasizing rapid iteration (“experimental mindset”), a vision beyond mere chatbots, and a focus on “utility” and genuine user value. The conversation is direct, with transparent answers on strategic challenges (compute constraint, economics, growth).

---

Summary

This episode delivers a detailed look at OpenAI’s evolution from consumer chatbot to a powerful productivity platform with rich ecosystem integrations, its strategic enterprise bets, and how infrastructure partnerships (like with AMD) are crucial to future growth. Brad Lightcap underscores the experimental, user-driven philosophy of OpenAI, highlighting a preference for measuring success by genuine usage (tokens processed) rather than headline valuations. The discussion offers valuable insight for anyone following the rapidly changing landscape of AI products, business models, and the competitive infrastructure behind them.

Transcript

[Indiana University Narrator]

Indiana University is shaping the future of healthcare. Advancing discoveries that become treatments for Alzheimer's, obesity, cancer and rare diseases. And training the providers trusted to deliver them from the lab to the clinic. IU powers medical breakthroughs and the talent behind them. See how IU solves what's next IU Edu Impact.

[Podcast Announcer]

Bloomberg Audio Studios Podcasts Radio.

[Host/Interviewer]

News well, we started today with a big pop in shares of AMD. This after it inked a landmark deal with OpenAI to build out infrastructure given the chip maker a chance to challenge Nvidia in the computing industry. Now this comes on the same day as OpenAI's annual developers day, a Developers Day where we've seen a lot of stocks, including Figma, HubSpot, Cisco, Mattel, just to name a few. Also moving significantly on based on some of the comments coming out of that conference, joining us right now is Ed Ludlow. He is at that conference and he's joined right now by the CEO of OpenAI, Brad Lightcap. Ed.

[Ed Ludlow]

Yeah, and the headline is the ability to use third party apps within Chat GPT and the profound impact on the market was simply those name checked, saw their stocks move in a significant way. Delighted to speak once again with Brad Lightcap that it's as simple as that. Is that really, you know, the ability to be within GPT and access Spotify or FIG as two examples. When you were discussing the idea of that and API access with, with those partners, was it a tough negotiation to get them on board with the idea?

[Brad Lightcap]

No. I mean we've seen enthusiasm for this from the beginning. You remember way back when we lost something called plugins back in the early days of ChatGPT, this was one of our first attempts to start to build an ecosystem around ChatGPT so that ChatGPT can start to engage with and interact with the applications that are important to you in your personal life and at work, and now really have a much richer surface through MCP and other protocols to be able to bring applications into ChatGPT and for really to allow you to engage with ChatGPT in the kind of work around work. Right. It's, it's the contextual aspect of I'm doing X or I need Y, I'm on a road trip and I want to know what playlists would go well with this in the context of my broader trip planning that allows you now to kind of use ChatGPT to solve that higher level task and then also integrate apps contextually to solve those specific problems.

[Ed Ludlow]

We're in a place where ChatGPT has become more of an operating system where that was your ambition or not. Is that where you want to take it, to be an OS and a developer driven platform? It's almost like an app store, you know, on paper, based on what you announced this afternoon.

[Brad Lightcap]

Well, we've always thought of Chat CBT as like a super assistant. We never set out to build a chat bot. We always wanted to build something that was really true to you and what your preferences are, what your goals are that could actually help you achieve more. And so I think part of that is Chatbot having an appreciation and understanding of the applications in your life that are important. And I think enabling that kind of connectivity and interoperability makes Chat CBT richer. Also enables a lot of pass through for people to be able to engage with apps. They love it as well as new apps.

[Ed Ludlow]

The pass through bit is interesting. Were there concerns with some of those technology companies you're partnering with that it would take traffic away from their own domains?

[Brad Lightcap]

No, I think mostly people are really focused on building into new interfaces. Right. This is just like mobile in some sense where you have a new interface, you have a new form factor. People are going to want to use mobile form factors on the go. And apps like Spotify in some ways exist almost because they really now mobile. And so we think actually there's an opportunity for builders to create entirely new applications that are even native to chatbots. And of course for services you love to be able to benefit there too.

[Ed Ludlow]

Is there a revenue sharing agreement with those third parties whose apps are accessible through Chat GPT?

[Brad Lightcap]

So we're going to figure out the economics of this over time. This, you know, we're brand new here. Plugins was the first version of this and that was even an experiment. And so we're like everything at OpenAI we take this very experimental mindset to making sure we get it right. But the idea is we do, we do want to build something that's useful for developers. And of course there's going to be, you know, have to be some exchange in there of, of, of economics and value and we'll have to figure out how to get that right.

[Ed Ludlow]

You have hit 800 million active weekly users you announced on stage. Actually Greg Brockman told me that at 8 o' clock this morning and maybe we missed it, but it's a significant milestone all the time. I'm asked by all kinds of people, do we have any sense of within that eight, how many are base level free users and how many are premium level paid subscribers?

[Brad Lightcap]

Yeah, we have a very healthy Funnel of people that choose to pay for ChatGPT. You know, it's surpassed where my expectations frankly were. People have this kind of conception that consumers tend to not pay for software. And similar even to what I was saying before around how do you co develop the product alongside the business model? ChatGPT is a great example of that. Where the subscription model I think has been really a testament to how valuable it is for more users than I think we expected to be willing to pay for it. So we don't I think disclose the exact number, but it's a, it's a healthy amount and more every day.

[Ed Ludlow]

OpenAI in the beginning went after the consumer for users aggressively. You are now very focused on the enterprise business. What is the strategy for that and how do you prioritize your enterprise business?

[Brad Lightcap]

Yeah, glad you asked about that. So really today's announcements actually I think target what are an important set of use cases for the enterprise. Things we've been hearing at enterprises ask us about now for some time. So specifically one is we now have an ability for enterprises to build agents in a much more visual, much more intuitive way. You've heard us say 2025 has been the year of agents. We think that's true. Codex has been a great example for us of that. Our coding agent now available through, through an API.

[Ed Ludlow]

Also the lead time to make software is a lot shorter.

[Brad Lightcap]

It's gotten a lot shorter. I think you saw today we demoed, live demoed, I think three or four different things that we've built in real time. We expect that to continue to be to the things like agent builder, allow enterprises to be able to build agentic experiences, powerful agent experiences on the go, you know, iteratively and connected into the tools and, and sources of information that matter for the business.

[Ed Ludlow]

The data point that jumped out at me is your API is handling more than 6 billion tokens per minute. And that helps explain why the AMD deal, you know, which is focused on inference. You are involved in all of these domains of the company. I've already asked Greg, but I've got to ask you, how are you going to finance yet another infrastruct infrastructure project like is there going to be some debt here specifically for the AMD capacity and how do you move quickly to get it online?

[Brad Lightcap]

Yeah, well, the high level thing is we are tremendously compute constrained. It feels like we're in this kind of recurring theme of being compute constrained. And I think the reason for that is the answer to the question you ask, which is demand.

[Ed Ludlow]

Right.

[Brad Lightcap]

We see There are multiples of demand that are latent and untapped from what we, we have today. And even today, obviously by any standard, demand and revenue growth has been torrid in its, in its pa and so really we have to invest ahead of that. And I think that's going to be the rate limiter for us to be able to go capture demand, whether it's consumer or enterprise, and for us to be able to build new models, parallelize more experiences, more product experiences, and then enable users specifically to be able to use those products more actively in their daily life at work and at home. And so, you know, even things like Sora, the app we just launched, we wish we could invite more people onto it now, but we just need more compute. So the AMD deal we're excited about being, you know, directional a way for us to do that.

[Ed Ludlow]

I've got to ask about the report that OpenAI closed secondary or the ability for employees to sell shares at a $500 billion valuation. I already asked you this question, but what is the metric we're supposed to judge your Success by? The $500 billion valuation, the 6 billion tokens per minute to you, Brad, what is it?

[Brad Lightcap]

For me it's, it's actually kind of a metric that we, we talked about is, is tokens. It's, you mentioned 6 billion tokens minute on our API. That is the purest for me, the kind of essence of utility is that consumption metric. And, and so we actively track that metric to see how people's consumption of AI is growing over time. And you see this happen in amazing ways. So things like Codex, for example, we've seen grow 10x since August purely on consumption of tokens around coding. And you start to see that same pattern emerge across multiple lanes of use and across multiple areas of work. And that's the metric I look at because if that number is going up, it means people are us for more things. And that's the ultimate goal.

[Ed Ludlow]

Brad Lightcap is open AI CEO. There has been a fire hose of headlines and it has moved markets all day long.

[Podcast Announcer]

Hiscox Small Business Insurance knows there is no business like your business. Across America, over 600,000 small businesses from accountants and architects to photographers and yoga instructors, look to his Cox Insurance for protection. Find flexible coverage that adapts to the needs of your small business with a fast, easy online [e-mail] that's his C O x Com. There's no business like small business. Hiscox Small Business Insurance.