- The Prohuman
- Posts
- Anthropic makes voice mode more useful
Anthropic makes voice mode more useful
Plus: NVIDIA bets $300 million on Korean AI
Rimsha Bhardwaj, Hasan Toor & Ihtesham Haider
July 24, 2026
In partnership with
Hello, Prohuman
Today, we will talk about these stories:
Claude turns spoken requests into actions
Korea’s AI strategy gets NVIDIA backing
Runway turns model overload into a product
200+ Claude Prompts Top Professionals Actually Use at Work
Claude can be your analyst, editor, and strategist.
But most professionals are using it to fix grammar.
These 200+ Claude prompts take it from grammar tool to your most powerful AI work assistant.
Sign up for Superhuman AI and get:
200+ ready-to-use Claude prompts to get real work done in minutes — researched, tested, and used by professionals at Google, Microsoft, and NASA
Superhuman AI newsletter (4 min daily) so you keep learning new AI tools and skills to stay ahead in your career — the prompts are just the beginning
Claude voice is becoming a work interface
Image Credits: Anthropic
Claude can now take a spoken request and act inside Gmail, Slack, Notion, Canva and Google Calendar.
Anthropic upgraded its voice beta so users can choose Opus, Sonnet or Haiku, while free accounts remain limited to Haiku and one connected app. Voice mode also carries over the last model used in text chat and defaults to its fastest version.
Voice is becoming useful.
My read is that Anthropic sees speech as a work interface, where task completion matters more than smoother conversation alone.
That gives Claude an advantage when rival voice assistants still leave users tapping through menus on a bright phone screen. It also raises the stakes for mistakes involving calendars, email and shared documents.
The gap is practical.
Will people trust a spoken command to change real work?
NVIDIA is funding Korea’s AI talent pipeline
Image Credits: NVIDIA
The money is the story.
NVIDIA and KAIST are launching a Seoul research lab through a planned $300 million collaboration, including $50 million in annual compute over five years.
At least 10 KAIST researchers each year will receive funding and NVIDIA internships, with the lab focused on Korean-language models and industry-specific AI agents.
My read is that NVIDIA wants a durable position inside Korea’s research system by funding the people, computing access and career paths around it. The practical test is whether Korean researchers gain meaningful control over the models and research priorities.
At the Seoul campus, that could eventually give local companies AI tools built around Korean language and operating needs.
Execution will decide the rest.
Runway is selling model selection itself
Image Credits: Runway
The best model keeps changing.
Runway’s new Media Router chooses among image, video, and audio models based on quality, speed, or cost. It runs through Runway Dev, whose customers include Adobe, Cloudflare, ElevenLabs, Expedia, Shutterstock, and Quora.
That is the real bet.
Runway’s own video models no longer lead major rankings, so the company is giving developers a reason to keep using its API anyway. I think this is a practical move because most teams do not have time to compare every model, price and output change themselves.
A bad lip sync on screen can expose the wrong choice immediately. Routing could become standard as token costs, provider rules and model quality keep shifting.
Can Runway remain trusted while sending customers to rival models?
Prohuman team
Covers emerging technology, AI models, and the people building the next layer of the internet. |  Founder |
Writes about how new interfaces, reasoning models, and automation are reshaping human work. |  Founder |
Free Guides
Explore our free guides and products to get into AI and master it.
All of them are free to access and would stay free for you.
Feeling generous?
You know someone who loves breakthroughs as much as you do.
Share The Prohuman it’s how smart people stay one update ahead.