🎧 Listen Ads-Free: Tired of interruptions? Subscribe to AI Unraveled directly on Apple Podcasts at
Summary: The first Monday of Q2 2026 reveals massive fractures in the capital structure of the AI industry. We perform a forensic analysis of the internal conflict at OpenAI, where CFO Sarah Friar has been sidelined after projecting a catastrophic $200 billion cash burn, directly challenging CEO Sam Altman’s aggressive IPO timeline. We also deconstruct Altman’s controversial new policy blueprint calling for a tax on automated corporate labor. At the infrastructure layer, we analyze Nvidia’s brilliant strategic hedge: releasing frontier-class open models (Nemotron 3) to prevent OpenAI and Anthropic from monopolizing the software ecosystem and dictating hardware prices. Finally, we look at Anthropic’s margin squeeze, forcing third-party agents off flat-rate billing, and the dangerous rise of “Cognitive Surrender” in enterprise decision-making.
Important Topics Covered:
The OpenAI IPO Civil War
The “Automated Labor Tax”
Nvidia’s Strategic Hedge
The Death of Flat-Rate Compute
Cognitive Surrender:
LinkedIn’s “BrowserGate”
Netflix VOID
Keywords: OpenAI IPO readiness, Sarah Friar, Sam Altman AI Tax, OpenAI $200B cash burn, Nvidia Nemotron 3 open source, AI hardware monopoly, Anthropic Claude OpenClaw ban, Penn study cognitive surrender, LinkedIn BrowserGate, Netflix VOID physics-aware AI, DjamgaMind, AIRIA.
This episode is made possible by our sponsors:
🛑 AIRIA: Secure your AI workforce. AIRIA unifies orchestration, security, and governance into a single command center, using micro-VM sandboxing to protect sensitive data from agentic goal-hijacking. 👉 Govern your agents: [LINK]
🎙 DjamgaMind: High-Fidelity Intelligence for the C-Suite. If you are a modern decision-maker, DjamgaMind delivers strategic audio forensics in Healthcare, Energy, and Finance. Stop reading headlines and start understanding the systemic impact with our human-verified, technical-grade analysis. 👉 Explore the Forensics: https://DjamgaMind.com/regulations
🛠️ The AI Executive Toolkit: Stop scrolling through generic lists. Get the hand-picked, forensic-vetted implementation stack to bridge the gap between raw innovation and professional-grade governance. Exclusive listener perks on tools like:
Chatbase: Build a secure, custom AI chatbot trained exclusively on your firm’s internal data and regulatory documents. (https://link.chatbase.co/etienne-noumen)
ElevenLabs: Transform lengthy compliance bulletins into high-fidelity “Audio Intelligence” for your team to consume on the go. (https://try.elevenlabs.io/4z7r3skyymar)
Fireflies.ai: Automate your meeting forensics by transcribing and extracting action items from your clinical consultations and board meetings. (https://fireflies.ai/?fpr=etienne62)
Google Workspace: Professionalize your firm’s infrastructure with secure, cloud-based collaboration and branded communication. (https://referworkspace.app.goo.gl/Q371)
Full toolkit at: https://djamgamind.com/toolkit
⚗️ PRODUCTION NOTE: We Practice What We Preach.
AI Unraveled is produced using a hybrid “Human-in-the-Loop” workflow.
Sam Altman proposes AI tax and regulation blueprint
OpenAI CEO Sam Altman released a 13-page policy blueprint on Monday that proposes new taxes, a public wealth fund, and regulation to prepare for AI’s expected impact on jobs and the economy.
The document calls for taxes “related to automated labor” to protect funding for programs like Social Security and SNAP, and recommends giving every citizen a stake in AI-driven economic growth.
Altman also suggested employers and unions push for four-day workweeks with no pay cuts, expanded training for human-centered jobs, and guardrails on how the government can deploy AI systems.
Why Nvidia chose open models to reshape AI
If you’re wondering why AI chips leader Nvidia is now building open models that compete with the Chinese open-source champs, and even proprietary models from OpenAI and Anthropic, then you’re not alone.
Last month, Nvidia launched Nemotron 3 Super, a 120-billion-parameter reasoning model that outperformed expectations in benchmarks. This is a mixture-of-experts model with a 1-million-token context window. In other words, it’s a serious model made to compete with the frontier labs. Meanwhile, the company promised that a model 4x its size, to be called Nemotron 3 Ultra, is coming soon.
And because Nvidia opens the weights, datasets, and training recipes, it’s among the most open models in the world, especially for a model of this capability. Some of the only models that could claim to be more open would be the ones from MBZUAI, which The Deep View covered in depth in January. But Nvidia’s open models are far closer to full-stack openness than most of the open-source models, which only offer open-weight releases.
So why would the leading hardware company of the AI era make software that competes with its leading customers?
“We’re not trying to control AI. We’re trying to grow it,” Bryan Catanzaro, VP of applied deep learning research at Nvidia, told The Deep View. “And so our incentives as a company, our business is aligned with open models and with supporting the ecosystem in a very direct way.”
Kari Briski, VP of generative AI software at Nvidia, told The Deep View another perspective: “The model is the byproduct. It is not core to our business, which allows us to just open up the data, open up the recipes, open up everything.”
If we break it down, there are three benefits Nvidia gets from making its own models:
Extreme hardware co-design: Making their own models allows Nvidia to optimize the heck out of their GPUs, CPUs and other hardware to run AI. They don’t have to wait to get the latest models from the frontier labs to plan the next stage of optimizations.
Hedging against proprietary monopolies: If the frontier labs that need the latest and greatest hardware dwindle down to only a handful of players, then Nvidia could end up at their mercy. When you have a smaller number of customers you rely on for huge numbers of orders, then those customers gain more and more control over your prices. They can demand lower prices because they know so much of your business depends on them.
Letting a thousands flowers (a.k.a. customers) bloom: By releasing open models that other hardware and software makers can use as a rapid on-ramp to build their own AI products and serve the various niches in the industry, Nvidia is powering up the ecosystem, helping companies with limited resources have models they can use to compete and potentially creating a lot more future customers when those companies succeed and grow.
“You don’t want one person winning [because] then they decide all the rules. You need a big open ecosystem for everybody to come along,” said Briski.
Anthropic boots third-party agents from Claude plans
Anthropic just blocked agent platforms like OpenClaw from running on Claude plans, requiring users to pay separately via usage add-ons or API keys, as the company confronts agent-driven demand its flat-rate pricing was never built to absorb.
The details:
Agent tools hit Claude with nonstop requests that exceed what its normal plans typically cover, despite Anthropic models being the leading driver for the tech.
Anthropic’s Boris Cherny announced the change, saying it is a step towards “managing growth to continue to serve our customers sustainably long-term”.
Anthropic is handing out credits worth a month’s subscription, discounting add-ons up to 30%, and offering refunds amid cancellation requests.
OpenClaw creator Peter Steinberger criticized the step, saying, “First they copy popular features into their closed harness, then they lock out open source.”
Why it matters: Anthropic was already catching heat over tighter rate limits, and walling off its agentic power-user community won’t help the goodwill problem. It’s a tough situation with Anthropic’s agent usage likely playing a role in degrading normal user experience, but OAI is now there as the alternative at a crucial time in the rivalry.
When AI thinks, humans stop questioning
AI might be causing us to forget how to think for ourselves.
Recent research from the University of Pennsylvania found that AI users were often willing to accept flawed AI reasoning, readily incorporating it into their decision-making with “minimal friction or skepticism.”
The research documents the rise of “cognitive surrender,” a phenomenon in which users adopt AI outputs while “overriding intuition… and deliberation.”
In a study of nearly 1,400 participants across 9,500 trials, researchers found that subjects accepted unsound AI reasoning more than 73% of the time and only overruled models’ decisions about 20% of the time.
Additionally, participants with higher trust in AI and “lower need for cognition and fluid intelligence” tended to fall victim to this more often.
“Across domains, AI tools are not merely assisting decision-making; they are becoming decision-makers,” The research reads. “This shift opens new theoretical ground: How should we understand human cognition and decision-making in an age when we outsource thinking to artificial processes?”
The study adds to a growing body of research on how AI may be impacting the way that we think. One of the most commonly cited studies comes from the MIT Media Lab, in which a group of test subjects was asked to write SAT questions with three different tools: one with OpenAI’s ChatGPT, one with Google search, and one with no help at all. Consistently, the ChatGPT users “underperformed at neural, linguistic, and behavioral levels.”
Even some of AI’s biggest names are questioning its effects on our brains. Anthropic CEO Dario Amodei said in a March interview with podcaster Nikhil Kamath that deploying AI in the wrong ways could easily make people “become stupider,” but only if they choose to forgo learning entirely. “Even if an AI is always going to be better than you at something, you can still learn that thing. You can still enrich yourself intellectually,” Amodei told Kamath.
The researchers, however, posit that cognitive surrender may not inherently be a bad thing. If an AI model is generally better at reasoning and decision-making than the person using it, with fewer mistakes, “deferring to a statistically superior system may be adaptive or even optimal.”
The bigger issue, however, comes down to agency. The researchers noted that this trend could mark a profound shift in cognition itself, “one in which users may not know when or why they have deferred, and where the line between human and machine agency becomes blurred
Netflix opens physics-aware AI for video editing
Image source: Netflix Research
Netflix just released VOID, an open-source framework built to erase video objects while rewriting the physics associated with them, instead of typical erasing and inpainting tools.
The details:
Existing removal tools just paint over backgrounds, without actually reasoning about the cause-and-effect those edits introduce across the broader scene.
VOID uses a mask that maps what to erase, what’s physically affected, and what to keep, with a judge model then charting the consequences.
VOID can handles physics it never trained on, with demos like a balloon floating when a holder is removed or blocks not falling when one in the chain is erased.
25 evaluators compared VOID against six baseline models including Runway, preferring Netflix’s results nearly 2/3 of the time.
Why it matters: This is Netflix Research’s first public AI release, and its a sign of where the video space is heading — intuitive systems that don’t just erase objects in footage like an image editor, but can actually simulate and alter the physics of the scene based on the changes for more controllability and real production use.
LinkedIn secretly scans over 6000 browser extensions
A report called BrowserGate accuses LinkedIn of running hidden code that scans for over 6,000 browser extensions on users’ computers, linking specific software choices back to real people and their employers.
The investigation claims LinkedIn can infer personal details like religious beliefs, political views, or job-seeking activity, and also scans for over 200 competing products like Lusha, Apollo, and ZoomInfo.
LinkedIn denies the accusations, saying it checks for extensions only to stop scammers and scraping, while the report’s author is a developer whose account was restricted for breaking platform rules.
China forces Apple to remove Jack Dorsey’s Bitchat
Apple pulled Jack Dorsey’s decentralized messaging app Bitchat from its China App Store after Beijing’s internet regulator, the Cyberspace Administration of China, said it violated rules on services capable of social mobilization.
Bitchat works entirely over Bluetooth and mesh networks without internet connectivity, letting messages hop between devices — a design that has made it popular during government-imposed connectivity blackouts in multiple countries.
The app has passed three million total downloads across platforms, and this is the second time China has targeted a Dorsey-backed decentralized app, after banning the Nostr-based Damus in 2023.
OpenAI CFO questions readiness for 2026 IPO
OpenAI’s CFO Sarah Friar has told colleagues the company is not ready for an initial public offering by late 2026, putting her in direct conflict with CEO Sam Altman’s goal to list by Q4.
Internal projections show OpenAI burning through more than $200 billion before reaching positive cash flow, with losses for 2026 alone projected at roughly $14 billion against $2 billion in monthly revenue.
Friar no longer reports to Altman directly, has been excluded from key financial meetings, and the company has quietly retained Goldman Sachs and Morgan Stanley to manage a possible offering.
Vibe coding boosted App Store submissions in 2025
App Store submissions surged 84 percent year-over-year in Q1 2026, and the growth of vibe coding tools like Claude Code and ChatGPT Codex is believed to be driving the increase.
For the full year of 2025, submissions grew 30 percent versus 2024, nearly hitting 600,000 total, with momentum building each quarter and accelerating sharply into early 2026.
Apple says its review team processes 90 percent of submissions within 48 hours, but developers and consumers have complained about lower-quality apps flooding the App Store as a result.
What Else happened in AI on April 06th 2026?
OpenAI is navigating a leadership change, with Fidji Simo on medical leave, COO Brad Lightcap on special projects, and CMO Kate Rouch stepping down for cancer recovery.
Anthropic acquired startup Coefficient Bio for roughly $400M, folding the team into its healthcare and life sciences group focused on drug discovery.
Mercor confirmed a data breach tied to an attack on open-source library LiteLLM, with hackers claiming access to up to 4 TB of data from the $10B AI training startup.
Pika Labs released PikaStream 1.0 in beta, a real-time model that lets AI agents join Google Meet calls as video avatars with voice cloning and live conversation.
OpenAI rolled out ChatGPT in CarPlay, allowing users to access Voice Mode in their supported vehicle for hands-free use.
MIT study models AI ‘sycophancy’, warns of ‘delusional spiraling’ in chatbot interactions [Link]
18-month New Yorker investigation finds OpenAI’s Sam Altman lobbied against the same AI regulations he publicly advocated for, pursued billions from Gulf autocracies, and how he tried to hide a post-firing investigation that produced no written report [Link]
Japan Wants to Build a Solar Ring Around the Moon That Will Provide Endless Clean Energy to Earth [Link]
UK confirms drone-killing DragonFire laser weapon for Royal Navy destroyers by 2027 —laser downs 400mph high‑speed drones, costs $13 per shot [Link]
Meta salary data reveals a VP of AI can make $650,000 in base salary [Link]
Microsoft says Copilot is for entertainment purposes only, not serious use — firm pushing AI hard to consumers and businesses tells users not to rely on it for important advice [Link]
OpenAI proposes superintelligence governance plan - taxing automation, establishing AI wealth funds, 4-day work weeks [Link]
