Podcast Summary: TBPN – Satya Nadella LIVE | GitHub Universe Special

Date: October 28, 2025
Hosts: John Coogan & Jordi Hays
Guests: Satya Nadella (Microsoft CEO), Alexander Embiricos (Codex Team), Kyle Daigle (GitHub COO), Jay Parikh (EVP Core AI, Microsoft), Jared Palmer (SVP GitHub, VP Core AI), Michael Grinich (CEO, WorkOS)

Episode Overview

In this live show from GitHub Universe in San Francisco, TBPN hosts John Coogan and Jordi Hays dive deep into the recently announced next phase of the Microsoft and OpenAI partnership. Satya Nadella, CEO of Microsoft, headlines the episode, discussing the origin, strategy, and future of Microsoft's AI bets. The conversation explores AGI, platform strategy, agentic software, and enterprise tech trends, with insights from Microsoft and GitHub leadership and leading figures in the AI developer ecosystem.

Microsoft & OpenAI: The “Greatest Deal of All Time”?

Background and Context

• Microsoft’s partnership with OpenAI began in 2019, with a $1B investment and Azure as the exclusive cloud provider.
• Today, Microsoft owns 27% of OpenAI, with an even deeper product, IP, and financial relationship.
• The structure of the deals, including revenue sharing and IP rights, is cited as almost unprecedented among Big Tech partnerships.

Notable Quote:
"At least today Microsoft makes like a billion dollars in revenue every single day... so do you treat that deal like it’s just another day at Microsoft or something that there’s weeks of negotiation?" – John Coogan (05:00)

Key Points:

• The deal evolved from a traditional minority investment to a highly strategic alignment, granting Microsoft commercialization rights, exclusive licenses, and vast Azure commitments.
• Unlike past tech investments, Microsoft’s override of legal and structural obstacles (nonprofit setup, revenue sharing, rights management) was what made the investment so unique, not the magnitude of the check.

Venture & Industry Perspective:
“There’s a little bit of sour grapes from the venture capital community. The amount of capital that the seed investors deployed... certainly they made a great return on paper, but did they actually make a great return relative to the risk?” – Jordi Hays (04:09)

Segment: Key Microsoft–OpenAI Deal Updates (10:45–14:44)

• AGI (Artificial General Intelligence) Declaration: Now to be validated by an independent panel of experts.
• IP Rights Extensions: Extended through 2032, including post-AGI models with safety guardrails.
• Research IP: Microsoft retains rights to confidential research methods until AGI or 2030.
• Consumer Hardware: Now explicitly carved out from Microsoft’s rights.
• Third-party Products: OpenAI can now co-develop with third parties; some API products exclusive to Azure.
• Azure Commitment: OpenAI has contracted to purchase $250B more in Azure services.
• Open Models & Government Access: OpenAI can offer API to US national security customers on any cloud; permitted to release open-weight models under certain criteria.

Notable Moment:
“My immediate thought is: how many of these things are going to be critical to iron out before the IPO?” – Jordi Hays (14:04)

Satya Nadella Interview (18:18–44:01)

The Genesis of the Partnership

• 2016: Early connection with Sam Altman (“since his first company, pre-YC days” – 18:35).
• OpenAI started on Azure; Elon Musk directly reached out for Azure credits.
• Microsoft’s obsession with natural language AI and Office positioned them to recognize OpenAI’s relevance as scaling laws and LLMs became apparent.

Nadella on the Billion-Dollar Investment:
"It was not that hard to convince anyone this is an important area, but... Bill [Gates] even said, 'Yeah, you’ll burn this billion dollars, right.' And we kind of had a little bit of high restraint, but we said, let’s give it a shot." – Satya Nadella (21:12)

Milestones

• 2021: “I first saw GitHub Copilot, said, 'man, this is worth it.'” – (22:12)
• “Blip” of OpenAI’s November 2023 board drama: “It’s a blip.” (22:27)

Microsoft’s Platform Strategy

• "Microsoft’s a platform and partner company... the value capture by the platform must be higher than by the platform itself." – Satya (23:01)
• Vision for AI in developer tools: Autonomous agents (Agent HQ/Mission Control), integration of all models, facilitating developer workflows.

On the (Elusive) Meaning of AGI

• “It’s become a bit of a nonsensical word ... we now know what the issue is. We know everybody describes even the intelligence we have, which has been exceptional, as jagged.” – Satya (25:00)
• Advocates for “broad intelligence” first—models need to eliminate “jaggedness” and build robustness in domains like code and knowledge work.

Microsoft’s Internal vs. External AI Efforts

• Platform over fiefdom: “My mindset is all platform, man. On Azure, do you run Windows? Linux? SQL Server? Postgres? ... I’d love to have Anthropic, Mistral, Google’s Gemini on Azure.” (29:55)
• Culturally: Fosters an internal competitive ecosystem, with teams vying to innovate rather than favoring internal silos.

The Importance of “Deals Guys” in Tech’s New Era

• "You have this great investment and it has great return and no carry; all the value goes to shareholders. That's awesome." (32:21)
• Recognizes that tech/resource deals (energy, chips, infra) are as critical as product innovation.

Investing Against the Future

• “Long before it was conventional wisdom... we put $10 billion, $13.5 billion, before it became a thing.” (35:07)
• “Tech is 4-5% of GDP. In 5–10 years, will that be higher or lower? It’s going to be higher... the only rate-limiter is economic growth.” (35:50)

Adapting to Rapid Knowledge Work Shifts

• On managing headcount amid AI disruption: “It’s learning the new production function... you’re rewiring yourself. Unlearning is the hardest part. Learning is easy; unlearning and learning together is much harder.” (39:17)
• Microsoft has transitioned through several platform and business model shifts; the AI era is unique for combining technical, business, and production/distribution paradigm changes.

Product & Developer Platform Insights

GitHub Copilot and Codex at the Center

• Copilot now “criminally underhyped” due to not requiring VC rounds—already massively adopted (80% of new users, 36M new users this year).
• The platform’s real value is “unifying the ecosystem,” letting developers choose their tools but collaborate and build atop GitHub-enabled infrastructure. (Kyle Daigle, 74:34)

Agentic Development and Multi-Model Future

• “This feels like 2010, before every app had an API. Now we’re in a kind of walled garden moment, and it’ll swing back to platform openness.” – Kyle Daigle (75:00)
• Move towards local and distributed inference for agents (“We’ll have more and more inference happen locally... but you need to trigger it earlier, on true intent, not just a 'now it's time to work' invocation.” – 78:42)
• Benchmarks are losing relevance for enterprise adoption—what matters is developer preference and integration ("Do you like it? Do developers love using Codex in their daily flow?" – Alex, 67:39).

Open Source, Enterprise, and AI-Native SaaS

On Open Source Business Models:
“Don’t just assume that your free open source users will directly translate into paying customers... set expectations early.” – Jared Palmer (113:32)

• Barbell approach: Go broad/free for indie devs, focus on extracting enterprise value from big users, but beware early enterprise deals derailing focus.
• For Cobalt or legacy languages, specific fine-tuning offers more value; most enterprises don't gain much from custom model training unless they have truly unique codebases.

Enterprise Readiness in the AI SaaS Era:

• AI startups today are “pulled upmarket far faster” than classical SaaS. Security, compliance, and sensitive data access requirements are front and center.
• WorkOS powers enterprise features (auth, provisioning, logging) for new AI SaaS companies "almost from day one."
• “In AI, a product is only valuable if you get access to all your stuff. It becomes a huge security concern. You need to scrutinize it more and move faster.” – Michael Grinich (124:29, 126:14)

Culture, Talent, and the Meaning of “Developer”

Should You Learn to Code or Do Deals?

• "I think yes, you should learn everything you can about these systems... Ultimately, it's less about knowing the code and more about systems thinking." – Jay Parikh (95:55)
• On deal-making: “I think that’s a life skill—knowing how to collaborate, negotiate, compromise, and see if there’s a greater good.” (97:10)

Developer Identity and Next-Gen Builders:

• “The developer label might melt away in 5–10 years. You used to need to be a photographer to take photos. Now everyone is. AI will democratize development the same way.” – John Coogan (108:12)
• Still, ambition, product focus, and high-agency “builders” will always be necessary to make the leap from idea to impactful product.

Notable Memorable Quotes & Moments

• “The biggest gaming business is the Windows business... [But] after Activision, we want our games everywhere: consoles, PC, mobile, TV, cloud.” – Satya Nadella (40:29)
• On AGI: “It’s become a nonsensical word... Everyone defines it differently... We have to get rid of these jagged problems before true general intelligence.” – Satya Nadella (25:00)
• “More demos, less memos.” – Jay Parikh (99:21)—on culture of rapid iteration in the AI age.

Timestamps for Key Segments

• [00:00–14:44] Microsoft–OpenAI Deal History and Update
• [18:18–44:01] Satya Nadella Interview: Vision, Bet, and Culture
• [73:54–78:03] GitHub Copilot, AI Developer Adoption (Kyle Daigle)
• [81:12–84:15] Enterprise & AI Data: Custom Models, Pretraining
• [87:21–99:21] Jay Parikh: Unlocking Creativity, Developer Culture, Corporate AI Strategy
• [106:12–117:54] Jared Palmer: Open Source Business, Developer Experience, Vendor Lock-In
• [120:00–133:50] Michael Grinich: WorkOS, AI SaaS Moves Enterprise-First, Security Risks

Final Observations

• The episode paints a picture of AI in 2025 as less about “the big bang of AGI” and more about messy, practical, entrepreneurial opportunity—navigating business model shifts, platform lock-ins, and cultural change in technical organizations.
• Microsoft’s early, risk-taking strategic investment is hailed as visionary, with Satya Nadella emphasizing platform openness, partnership, and readiness for next innovation cycles.
• Across GitHub, OpenAI, and the startup ecosystem, the new organizing logic of “developer as builder, agents as teammates, and rapid enterprise adoption” dominates—a stark shift from the consumer, small-team ethos of the 2010s.

For more deep dives and interviews as they happen, follow TBPN across all platforms and subscribe for text summaries and more.