In partnership with
In Today’s Issue:
💼 Elon Musk takes the stand against Sam Altman in a $134 billion trial
☁️ OpenAI expands enterprise reach by embedding GPT-5.5 and Codex directly into AWS
🧠 NVIDIA drops Nemotron 3 Nano Omni
💰 An NVIDIA executive admits compute costs now exceed human labor costs
🏆 GPT-5.5 Pro claims the top spot on the AI leaderboard
✨ And more AI goodness…
Dear Readers,
Elon Musk just took the stand against Sam Altman in what might be the most consequential AI lawsuit ever filed, and the $134 billion question at the heart of it could reshape how we think about nonprofit tech organizations forever.
But while the courtroom drama unfolds in Oakland, the industry itself is not waiting around: OpenAI is embedding GPT-5.5 and Codex directly into AWS, NVIDIA just dropped an open omni-modal reasoning model that processes text, images, audio and video through a single system, and an NVIDIA executive openly admitted that compute now costs his team more than people do, puncturing the myth that AI is already a cheap labor replacement. Meanwhile, GPT-5.5 Pro has broken the three-way tie atop the AI leaderboard, and an open-weight model from Xiaomi is giving Claude Opus 4.5 a run for its money in coding.
Today's issue is packed with shifts that matter, from billion-dollar legal battles to benchmark wars to the uncomfortable economics of actually running AI at scale, so grab your coffee and scroll on.
All the best,
Kim Isenberg
🚀 OpenAI Expands Deeply Into AWS
OpenAI and AWS are broadening their partnership by bringing GPT-5.5, Codex, and Bedrock Managed Agents into AWS environments, aiming to make enterprise AI deployment feel less like a platform jump and more like a native workflow. The move is especially notable for companies already committed to AWS, since it folds OpenAI’s models and agentic tools into familiar security, governance, billing, and compliance systems.
The announcement also proves a sharper enterprise push: Codex on Bedrock targets software and knowledge-work automation, while Managed Agents powered by OpenAI offer a path from prototypes to production-ready workflows. Beneath the product news is a strategic message: OpenAI wants its frontier capabilities to live wherever large companies already build and operate.
🚀 NVIDIA Unveils Multimodal Agent Model
NVIDIA has launched Nemotron 3 Nano Omni, an open omni-modal reasoning model designed to let AI agents process text, images, audio, video, documents, charts and interfaces through a single system rather than stitching together separate models. The company frames it as a major efficiency step for agentic workflows, claiming up to 9x higher throughput and strong benchmark performance in areas like document intelligence, computer use and audio-video reasoning.
What stands out is NVIDIA’s emphasis on deployment control: open weights, datasets and training techniques, plus availability across Hugging Face, OpenRouter, build.nvidia.com and partner platforms. For enterprises, the pitch is not just faster agents, but agents that can perceive complex digital environments more coherently while fitting into regulatory or localized infrastructure needs.
💸 AI Costs Outrun Human Labor
NVIDIA’s Bryan Catanzaro says the cost of compute now exceeds employee costs for his team, challenging the popular narrative that AI is already a cheaper substitute for human workers. Fortune’s article draws a striking contrast between massive AI investment and ongoing tech layoffs, noting that companies are spending hundreds of billions on infrastructure while evidence of broad productivity gains or job displacement remains limited. Experts frame this as a short-term economic mismatch: AI may eventually become cheaper and more predictable, but today its hardware, energy, software, and oversight costs make it more of a costly complement than a clean labor replacement.
AI’s rapid progress in math suggests models are moving from helpful problem-solvers toward research collaborators that can search deep literature, reason over longer timelines, and potentially contribute to original mathematical discovery. OpenAI’s Podcast Ep. 17 digs into this topic.
Musk vs. OpenAI Goes to Trial
The Takeaway
👉 Elon Musk is suing OpenAI, Sam Altman, and Greg Brockman for breach of charitable trust and unjust enrichment, claiming they betrayed OpenAI's founding nonprofit mission by converting it into a for-profit entity.
👉 Musk seeks up to $134 billion in damages (directed to OpenAI's charitable arm), the removal of Altman and Brockman, and a reversal of the October 2025 corporate restructuring.
👉 OpenAI argues Musk was part of early discussions about going for-profit and only sued after losing a power struggle and launching his own competing AI company, xAI.
👉 The trial, expected to last two to three weeks, could complicate OpenAI's planned IPO, strain its Microsoft partnership, and set a legal precedent for nonprofit-to-profit conversions across the tech industry.
The biggest courtroom drama in AI history just kicked off. Elon Musk and Sam Altman, once allies who co-founded OpenAI together in 2015, are now facing each other in a federal court in Oakland, California. Musk took the stand on April 28 as the first witness, accusing Altman and OpenAI president Greg Brockman of turning a charity into a profit machine.
The core allegation: Musk claims Altman and Brockman deceived him into bankrolling OpenAI by promising to maintain it as a nonprofit dedicated to developing AI that benefits humanity, only to later restructure into a for-profit entity. He is seeking up to $134 billion in damages, the removal of Altman and Brockman from their roles, and a full reversal of the corporate restructuring.
OpenAI's lawyer fired back, telling jurors that Musk sued because he lost a power struggle and his rivals succeeded without him. With OpenAI valued at over $850 billion and eyeing an IPO, while Musk's own xAI is expected to go public as part of SpaceX, the stakes could not be higher. Expected witnesses include Altman, Brockman, and Microsoft CEO Satya Nadella. The judge expects to issue her ruling by mid-May 2026.
Why it matters: This trial could fundamentally alter the trajectory of one of the most powerful AI companies on the planet, potentially delaying OpenAI's IPO and disrupting its partnership with Microsoft. The outcome will set a precedent for how nonprofit tech organizations can restructure into for-profit entities, with implications far beyond AI.
Sources:
🔗 https://openai.com/de-DE/elon-musk/
Data-driven global scaling for 2026
Stop basing talent decisions on outdated figures. Deel’s 2026 Global Hiring Report provides salary benchmarks and growth trends from 150+ countries. Learn about the 283% rise in AI roles and how the talent landscape is shifting. Use these insights to optimize your spend and scale your team with total compliance.
Xiami mimo-v2.5 pro MIT license surpasses Opus 4.5 on arena.
We now have an open weight model that outperforms the recently SOTA coding model.
GPT-5.5 Pro Tops AI Rankings
The AI leaderboard just got a shakeup. OpenAI's freshly launched GPT-5.5 Pro has claimed the top spot on the Artificial Analysis Intelligence Index, breaking a three-way tie between OpenAI, Anthropic, and Google that had persisted for weeks.
But what makes this especially interesting is a tool called the Epoch Capabilities Index (ECI), built by the research org Epoch AI. Think of it as an IQ test for AI models: it stitches together 42 different benchmarks into a single score, so you can compare models even when individual benchmarks get too easy too fast. On that scale, Claude 3.5 Sonnet sits at 130 and GPT-5 at 150, giving you a rough sense of the range.
Before GPT-5.5, the top ECI spots belonged to GPT-5.4 Pro and Gemini 3.1 Pro. Now, GPT-5.5 Pro dominates across key benchmarks like FrontierMath Tier 4 (39.6%), ARC-AGI-2 (85.0%), and Terminal-Bench 2.0 (82.7%), all significant leaps over its predecessor. Epoch AI's own research shows the rate of AI capability progress has nearly doubled since April 2024, jumping from about 8 ECI points per year to 15. That acceleration is not slowing down.
Gladly Connect Live '26. May 4–6 in Atlanta.
The room you want to be in. This is where CX leaders are tackling the hard AI questions and sharing what's actually working. For CX and ecommerce leaders. Atlanta, May 4–6. Space is limited — secure your spot now.
