Apple's Siri AI & OpenAI's IPO

Welcome back, Superhuman. Apple finally revealed its long-awaited Siri AI and Wall Street was wildly…unimpressed. Apple’s stock dropped 2% on the day. What do you think: Does Siri AI look underwhelming, or have expectations gotten impossibly high?

Today: Apple’s Siri AI, how to create stylized videos with Gemini Omni, and get the latest prompts and trending social posts.

1. Apple reveals Siri AI at Tim Cook's final WWDC: Siri AI will feature a dedicated chatbot-style app that delivers personal assistance by pulling context across your messages, emails, photos, and more — synced privately across devices. The keynote also debuted iOS 27, macOS Golden Gate, and a homeOS preview for Apple's upcoming smart home hub. See how Siri AI works or learn everything Apple announced during its keynote.

2. OpenAI files for IPO one week after rival Anthropic: OpenAI has submitted a confidential S-1, mirroring Anthropic’s move last week. The startup said it has not decided on timing yet and may wait because there are “things we want to do that are likely easier as a private company.” Separately, Sam Altman co-authored his plan for AI to benefit everyone.

3. WorkClaw launches AI coworkers that onboard like new hires: WorkClaw deploys agents that work inside Slack and Microsoft Teams to get tasks done. Each agent is compatible with 3,000+ apps and comes with a job description, manager, and its own isolated computer. It’s built for enterprise compliance by engineers from Meta, Google, and Microsoft. See how it works or try it here.

We’ve unlocked a WSJ article for you: Learn how long it will take for AI to reach its full potential.

Chatbots are just the beginning. Google for Startups’s Startup School: Agentic AI kicks off today, teaching you how to design and deploy autonomous systems that actually execute complex tasks. 

Master advanced tool calling, multi-agent orchestration, and complex reasoning chains directly from Google experts. Transform your AI features into scalable, independent workflows. 

Classes start today—save your spot now!

Entering the frontier. For years, Microsoft’s role in the AI race has been an investor and compute provider. That changed last week, when the tech leader rolled out its own general reasoning model — along with six others. The message is clear: Microsoft is stepping off the sidelines and into the race.

The breakup. Microsoft spent years as OpenAI's exclusive compute partner, biggest backer, and main distribution channel. The two companies jointly ended that agreement in April, leaving Microsoft free to launch its own models and compete head-on with the company it bankrolled.

Meet the models. Microsoft’s general MAI-Thinking 1 is still in private preview, so it hasn’t hit public rankings yet — where Anthropic, Google, Meta, and OpenAI reign supreme. MAI-Image-2.5 currently sits at #4 for image generation. The big question is: how does Microsoft plan to compete without the best model on the board?

Margins, data, & distribution. Microsoft’s models are trained on commercially licensed data, hosted on Azure, and can be fine-tuned using customers’ own information — a combination that makes them highly cost-efficient. For OpenAI, it’s an unideal twist: the company that once wrote the checks is now building its competition.

Papercraft Prompt: Create a 10-second one-continuous-shot video of a tiny paper fox walking through a layered papercraft forest at sunrise. Use visible cut edges, folded trees, handcrafted shadows, a slow camera push-in, soft golden lighting, and a whimsical mood.

Origami Prompt: Create a 10-second video of an origami crane unfolding on a wooden table and lifting into the air. Show crisp folds, elegant paper texture, gentle camera tilt, soft window light, and a serene poetic mood.

Claymation Prompt: Create a 10-second claymation scene of a small explorer crossing a rocky hill while clouds slowly move overhead. Use hand-modeled textures, slightly imperfect stop-motion timing, a locked-off camera, warm afternoon light, and a cozy adventurous mood.

Plush World Prompt: Create a 10-second video of a plush toy astronaut floating through a soft fabric moon base. Show visible stitching, fuzzy textures, slow drifting movement, a smooth camera glide, and a cozy magical atmosphere.

Prompt Source: Dheepanratnam

AI for the rest of us. No prompts to engineer, no setup. Viktor lives in Slack and Teams, so you just ask. "Pull last week's revenue." "Draft the launch email." "Review this PR." One teammate finance, marketing, and engineering already share.

Start at no cost, $100 in credits at viktor. com.

👀 A better Siri? One startup is building the ultimate Mac productivity tool that claims to “do everything Siri can’t do.” The teaser demo has gotten roughly 1M views.

🤖 300 Agents: Kimi Work is a new desktop agent that can launch up to 300 AI agents in parallel on your computer to get tasks done.

📊 New Feature: ChatGPT got a small but mighty update — the ability to generate graphics directly in-chat, making it easier than ever to learn interactively.

📝 Thought Provoking: This piece of advice from Anthropic’s co-cofounder is critical for staying sharp — and sane — in the age of AI.

🥭 MANGO: “FAANG” described the biggest tech companies of the 2010s. A new acronym just dropped for the AI era: MANGO.

^{* indicates a promoted tool, if any}

This detailed video tutorial breaks down how OpenAI co-founder Andrej Karpathy uses AI day to day, racking up 29K bookmarks and 1.5M views.

Grow customers & revenue: Join companies like Amazon, HubSpot, and Salesforce. Showcase your product to our 1M+ readers and 2M+ followers on socials. Get in touch.

Until next time — Zain, Theodore, & the Superhuman AI team.