Welcome back, Superhuman. Many people didn’t take the idea seriously at first. Launching data centers into orbit sounds like the plot of a sci-fi movie. Not a corporate strategy. But after yesterday’s partnership, this once-fringe plan looks set to become reality.
Today: OpenAI calls out Apple, how to automate any task with OpenAI Codex, and get the latest prompts and trending social posts.
TODAY IN AI
Learn how Nvidia and SpaceX plan to send data centers into orbit. Source: SpaceXAI.
1. SpaceX and Nvidia team up to send data centers into space: The two tech leaders unveiled plans to co-design Starmind AI1, a solar-powered satellite that will beam AI compute to Earth, running on Nvidia’s chips. The approach is unproven and may never be commercially viable, but, if successful, could address three challenges with Earth-based data centers: emissions, cooling, and power. SpaceX also posted its first earnings since June's IPO, reporting $7.8B in Q2 revenue, up 92% annually, on a $541M net loss.
2. Bland releases benchmark-topping new voice model: Bland Speech v3 claimed the top spot on Design Arena’s Audio Realism Bench, beating leading models from Microsoft, SpaceXAI, and OpenAI. The model generates realistic human-sounding voices with natural stumbles and pauses from either written text or at least 10 seconds of existing audio. It’s built for customer support calls. Listen to it here.
3. OpenAI calls Apple’s trade-secret lawsuit “careless, aggressive, and oddly personal”: Apple sued OpenAI last month, accusing the lab of stealing corporate secrets. OpenAI just published a response, saying it doesn’t “have, nor want” Apple's trade secrets, along with internal messages it says contradict Apple’s story. The ongoing legal battle could represent a major obstacle for OpenAI’s hardware ambitions. Read the full blog here.
Yesterday’s most-clicked story: Alibaba launched Qwen3.8-Max, a new open-weight model rivaling the performance of US frontier models that it calls an “always on workmate.”
SPONSORED BY YOU. COM
Teams often pick an API by checking a benchmark table and calling it done—a shortcut that could miss what really matters in production. This guide from You. com explains why raw latency is a misleading signal and what to measure instead.
Stop optimizing for the wrong number. Download the guide.
FROM THE FRONTIER
AI isn’t just changing how we get work done — but also who does it
Made with Midjourney
Traditional roles are starting to blend
Have you used AI to tackle a project outside of your normal domain? If so, you’re part of a young but growing trend. OpenAI found that 43.5% of occupation-specific ChatGPT messages involved tasks normally associated with another role, like a salesperson spinning up a custom website, or an engineer running financial projections. The AI lab calls it task crossover.
Don’t be surprised if “task crossover” comes to your company
Using AI for tasks outside of your job description is still the exception, not the rule. Out of the 800,000 messages OpenAI studied, only about 17% crossed into another occupation's territory. But expect that percentage to continue climbing as companies grow more open to using AI, employees get savvier with the tools, and the tools themselves get more capable.
AI is also enabling professionals to venture out on their own
Entrepreneurship is also growing in popularity. Census data shows that monthly professional services business applications have climbed from 55k to over 80k since 2025. Both reports point to the same trend: AI is slowly untethering employees from traditional job titles.
IN THE KNOW
What’s trending on socials & headlines today
Meme of the day
📈 The Claudefolio: Last March, Anthropic gave Claude $50k with a goal — invest the money and beat the S&P 500’s return. The results so far? Claude is up 19%, compared to the index’s 12%. See which stocks Claude is invested in here.
😮💨 Job Outlook: A software engineer argued that tech jobs can’t survive much longer, given how advanced AI is getting. The comments pushed back, listing four reasons why that outlook is backwards.
🎤 Claude’s Performance: Several weeks ago, a spoof of God’s Plan by Drake went viral, called Claude’s Plan. The viral song just resurfaced, this time featuring an entire live performance (3.5M views).
😆 ChatGPT Voice: Screenshots from four-year-old Reddit posts have been going viral because they all sound exactly like ChatGPT. It’s a subtle reminder of just how much ChatGPT leans on the social media site for information (see where else it pulls from).
SPONSORED BY IBM
CrushBank used IBM Bob, an AI development partner, to help customers move from disconnected applications and databases to governed data that powers search, assistants, knowledge retrieval and analytics. With IBM Bob, the company was even able to expand an IT support business into a data and AI solutions provider that spans their entire enterprise.
PRODUCTIVITY
👩🏫 Ankon: Turn any idea into a whiteboard video with AI.
🤝 Koji: Know what your customers want and why using AI powered insights.
✍ VideoTranscript: Turns videos and audio into searchable and reusable transcripts.
TUTORIAL
How to automate any task with OpenAI Codex
Open OpenAI Codex > Plugins > enable Computer Use > start a new task
Switch permissions from Default to Full access, approve the required access prompts
Then give Codex a real task to complete
Sample Prompt: “Open [app/website/file], complete [task], then summarize the results and ask before making any major changes.”Codex can also handle repetitive desktop workflows across your apps, from organizing files and updating spreadsheets to testing websites and more
Note: AI tools can make mistakes, please be careful with sensitive data and tasks.
PROMPT
Turn Rough Role Notes into a Job Description
ChatGPT Prompt: You are a recruiter turning hiring-manager notes into a job description. Write the job description using the context below. Context:
Role notes from hiring manager: [notes]
Team and reporting line: [team]
Must-have vs nice-to-have skills: [skills]
Compensation range if shareable: [comp]
Company context: [company]
Instructions:
1. Separate true must-haves from nice-to-haves explicitly — inflated requirement lists ("10+ years for a role that needs 3") are a leading cause of strong candidates self-selecting out, especially from underrepresented groups.
2. Write the responsibilities section around what the person will actually do in the first 6 months, not a generic list of duties copied from a similar posting.
3. Avoid gendered or exclusionary language patterns (aggressive competitive language, unnecessary jargon) that research shows suppresses application rates from some candidate groups.
4. Include compensation range if shareable — postings without a range increasingly get filtered out or distrusted by candidates.
5. End with a clear, low-friction application step.EXTRAS
Want more?
Check out our Top 125 AI Tools
Get better at prompting with our Top 1,000 Prompts
Learn how to use AI at work with our 50 Tutorials
Grow customers & revenue: Join companies like Amazon, HubSpot, and Salesforce. Showcase your product to our 4 million+ readers and followers on socials. Get in touch.
What did you think of today's email?
Until next time — Zain, Theodore, & the Superhuman AI team.
