PLUS: A record $1.1B seed round, a CEO's AI clone, and soaring AI compute costs
Happy reading
OpenAI and Microsoft have fundamentally restructured their landmark partnership, marking a major power shift in the AI industry. The AI leader is now free to offer its products on any cloud platform, breaking its previous exclusivity with Microsoft Azure.
This revised agreement significantly alters the competitive landscape by removing key exclusivity terms. What will this newfound independence mean for OpenAI's path to a potential IPO and its race against cloud giants like Google and Amazon?
In today’s Next in AI:
OpenAI breaks free from Microsoft exclusivity
A record $1.1B seed round for superintelligence
A bank CEO’s AI clone runs an earnings call
The soaring cost of AI compute
OpenAI Breaks Free from Microsoft's Grip
Next in AI: OpenAI and Microsoft have fundamentally restructured their landmark partnership, giving the AI leader new freedom to partner with other cloud providers. The revised AI deal overhauls key exclusivity and revenue-sharing terms, signaling a major power shift in the AI landscape.
Explained:
The biggest shift allows OpenAI to sell its products across any cloud provider, a major departure from its previous Azure-exclusive arrangement that opens the door to customers on AWS and Google Cloud.
Microsoft’s license to OpenAI's tech is now non-exclusive, and while OpenAI continues revenue-sharing payments until 2030, the total amount is now subject to a cap.
This new structure removes the controversial "AGI clause" and gives OpenAI more certainty as it reportedly explores an IPO, which could happen as soon as this year.
Why It Matters: This agreement marks OpenAI’s evolution from a heavily dependent partner to an independent power player charting its own course in the AI race. For the rest of the industry, this unleashes a more competitive landscape where OpenAI's leading models can now run on any major cloud platform.
The $1.1B Seed Round That Broke Records
Next in AI: Ineffable Intelligence, a new lab from a former Google DeepMind leader, just secured a record $1.1 billion seed round to pursue superintelligence. The funding, backed by giants like Sequoia, Nvidia, and Google, gives the months-old startup a staggering $5.1 billion valuation.
Explained:
Founder David Silver aims to build a "superlearner" AI that learns from its own experience through reinforcement learning, rather than relying on massive datasets of human-generated text and images.
The historic funding is co-led by Sequoia and Lightspeed, with major participation from Nvidia, Google, and the U.K.'s Sovereign AI Fund, signaling immense confidence in Silver's vision.
This move is part of a larger talent exodus from Big Tech, with top researchers from Meta, Google, and OpenAI leaving to launch their own heavily funded AI labs.
Why It Matters: This unprecedented seed investment shows investor appetite is shifting towards funding fundamental AI research at a massive scale, not just near-term applications. The migration of elite talent and capital from established giants to nimble startups is set to intensify the race for next-generation AI.
Bank CEO Lets AI Clone Run Earnings Call
Next in AI: In a bold move, Customers Bank CEO Sam Sidhu used an AI clone of his voice for an earnings call before announcing a multiyear deal that embeds OpenAI engineers to automate core banking operations.
Explained:
The bank is targeting significant efficiency gains, aiming to improve its efficiency ratio to the low 40s and slash loan timelines from weeks to just seven days.
Unlike a standard software license, this partnership involves embedding OpenAI engineers at the bank to co-create enterprise solutions that OpenAI could eventually sell to other financial institutions.
Customers Bank already uses AI to write half of its software code, saving an estimated 28,000 hours of work—the equivalent of about 15 full-time employees.
Why It Matters: This move signals a major acceleration of AI adoption in the highly regulated banking sector. It provides a blueprint for how smaller, agile banks can leverage AI to compete with larger institutions by building a digital workforce.
When AI Costs More Than Your Engineers
Next in AI: A new reality is setting in for companies deploying AI at scale: the cost of compute power is starting to outpace the cost of the engineers who build with it. This trend challenges the long-held assumption that AI is primarily a tool for reducing labor costs.
Explained:
High-profile examples are surfacing across the industry, with Uber's CTO having reportedly blew through his AI budget for 2026 and an Nvidia VP confirming his team's compute costs are far beyond their salaries.
This isn't just anecdotal—it reflects a massive surge in spending, with worldwide IT spending projected to reach $6.31 trillion in 2026, largely driven by AI infrastructure and services.
The intense costs are creating a new competitive front among AI labs, with providers like OpenAI and Anthropic now competing on token efficiency and adjusting pricing models to manage the spike in enterprise demand.
Why It Matters: The narrative around AI is shifting from a simple cost-cutter to a significant strategic expense that demands careful management. As these costs continue to rise, the pressure will mount for leaders to prove a clear and substantial return on their AI investments.
AI Pulse
DeepMind unveiled a new distributed architecture called Decoupled DiLoCo that trains LLMs across distant data centers with lower bandwidth and greater resilience to hardware failures.
GitHub announced its Copilot plans will transition to usage-based billing on June 1, moving from a system of premium requests to a token-based GitHub AI Credits model.
Qualcomm surged after a report from analyst Ming-Chi Kuo suggested the chipmaker is partnering with OpenAI and MediaTek to develop processors for an OpenAI smartphone expected in 2028.
Google is building a new Prompt API into Chrome, allowing web developers to directly access the on-device Gemini Nano model for generative AI tasks.