Retter Reboot #19

The Retter Reboot : Your weekly dose of ai & tech news

1. Anthropic Lands SpaceX Compute Deal — 300MW for Claude

Anthropic signed an agreement with Elon Musk's SpaceX to tap the entire Colossus 1 data center in Memphis, gaining over 300 megawatts and access to more than 220,000 NVIDIA GPUs within the month. The deal directly addresses 80x growth in Q1 2026 that strained Claude Pro and Max reliability, and signals an unexpected detente between Musk and Anthropic amid his ongoing OpenAI lawsuit.

Source: Bloomberg

2. OpenAI Ships GPT-5.5 Instant as ChatGPT's New Default

OpenAI replaced GPT-5.3 Instant with GPT-5.5 Instant on May 5, reporting 52.5% fewer hallucinated claims on high-stakes prompts in medicine, law, and finance, plus tighter responses with 30% fewer words and fewer gratuitous emojis. Plus and Pro users on web get enhanced personalization pulling from past chats, files, and connected Gmail, with new Memory Sources controls letting users inspect and edit what context shaped each answer.

Source: OpenAI

3. OpenAI Launches Self-Serve Ads Manager — Targets $100B by 2030

OpenAI rolled out a beta self-serve Ads Manager at ads.openai.com on May 5, opening ChatGPT advertising to all US advertisers with CPC bidding, a Conversions API, and pixel-based measurement. The company is targeting $2.5B in ad revenue in 2026 and $100B by 2030, having already crossed $100M annualized — confirming that ads are now a core revenue line, not an experiment.

Source: Axios

4. Google Quietly Kills Project Mariner, Folds Tech Into Gemini Agent

Google shut down Project Mariner on May 4, ending its 17-month experiment with a screenshot-based browser agent and absorbing the technology into Gemini Agent and Chrome's Auto Browse feature. The move reflects an industry-wide pivot away from visual browser agents toward API-centric and command-line agents like OpenClaw, which are cheaper, faster, and more reliable than continuous screen processing.

Source: The Verge / Wired

5. Anthropic Goes Hard at Wall Street with Claude Opus 4.7

At an invite-only briefing in New York featuring JPMorgan CEO Jamie Dimon, Anthropic launched Claude Opus 4.7 alongside ten pre-built AI agents for banks, asset managers, and insurers. The package includes full Microsoft 365 integration — Claude operating as a single agent across Excel, PowerPoint, Word, and Outlook with shared context — Anthropic's clearest move yet to become the operating layer for financial services.

Source: Fortune

6. Microsoft Pledges $10B to Japan for Sovereign AI

Microsoft announced its largest-ever Japan commitment — $10 billion across 2026–2029 — covering AI data center expansion with SoftBank and Sakura Internet, deep cybersecurity cooperation with the government, and training over one million engineers by 2030. The investment maps directly to PM Takaichi's Sovereign AI strategy, keeping sensitive data inside Japan's borders while delivering the full Azure AI stack.

Source: Crescendo AI

7. EU Council and Parliament Agree to Streamline the AI Act

On May 7, the EU Council and Parliament reached a position to simplify and streamline AI Act rules, aiming to cut compliance friction without rolling back the framework's core obligations. The move comes as European enterprises and AI developers have pushed back hard on overlapping reporting requirements and tight implementation timelines.

Source: Consilium (EU)