🪖 Google signs Pentagon AI deal despite internal revolt

💸 OpenAI missed revenue targets ahead of its IPO

🤖 AI still costs more than the humans it's supposed to replace

🇪🇺 EU orders Google to open Android to third-party AI services

📸 Apple to add AI-powered photo editing in iOS 27

+ 📊 Daily poll and results

+ 📈 Trending tools and resources

1,000+ ready-to-use prompts to solve problems in minutes instead of hours—tested & used by 1M+ professionals

Superhuman AI newsletter (3 min daily) so you keep learning new AI tools & tutorials to stay ahead in your career—the prompts are just the beginning

Google signed a classified AI contract giving the Pentagon access to its models for any lawful government purpose, the same week over 600 employees sent Pichai a letter demanding the company refuse military AI work.

The deal mirrors contracts OpenAI and xAI signed last month. Google quietly removed its no-weapons AI pledge in 2025, reversing a policy it adopted in 2018 after staff protests actually worked back then.

Anthropic is currently in court after getting blacklisted for keeping its safety guardrails in place, which means Google just watched that happen and signed anyway.

OpenAI fell short of its own revenue and user growth targets, spooking its CFO enough that she flagged concerns about funding future compute deals, according to a Wall Street Journal report.

The company is locked into massive infrastructure commitments, including a $300 billion compute deal with Oracle and an expanded $138 billion agreement with Amazon, all while racing toward a public offering.

Altman and CFO Sarah Friar issued a joint denial calling the report "ridiculous," which is either a sign the story is wrong or that two executives urgently needed to get on the same page.

Nvidia's VP of applied deep learning Bryan Catanzaro said compute costs for his team already exceed what he pays his employees, a claim backed by a 2024 MIT study finding AI automation is only cost-effective in 23% of vision-based roles.

Big Tech is spending anyway, with $740 billion in AI capital expenditures announced this year alone, a 69% jump from 2025, even as companies like Meta cut 8,000 jobs and Uber's CTO says AI tools have already blown past his budget.

Gartner predicts inference costs for large models will drop over 90% by 2030, meaning the math on AI vs. human labor could flip, but only if the technology also gets more reliable and less likely to destroy someone's database.

The European Commission concluded that Gemini gets unfair system-level advantages on Android, and wants Google to let rival AI tools respond to hotwords, view screen context, and access local data for suggestions.

The push comes from the Digital Markets Act, which designates Google as a gatekeeper subject to stricter competition rules. The EU has already forced Google to add search choice screens and allow alternative Play Store payments under the same law.

Google called the order an "unwarranted intervention" that would expose sensitive hardware permissions, but the EU has a consistent track record of making its DMA demands stick.

Apple plans to launch four AI photo tools in iOS 27: Clean Up (existing), plus new Extend, Enhance, and Reframe features that can expand backgrounds, improve lighting, and shift photo perspective using generative AI.

Bloomberg reports iOS 27 will also bring a Gemini-powered Siri overhaul and a standalone Siri app, making the photo tools just one piece of a broader Apple Intelligence upgrade coming at WWDC on June 8.

Google's Magic Editor launched in 2023 and still offers more radical image generation than anything Apple is planning, so Apple is closing the gap but not leading the race.

A) Yes, finding videos just got way easier - 32%

B) No, the search bar was never the problem - 68% 🏆