- The Prohuman
- Posts
- Deep Research rewires the note taking loop
Deep Research rewires the note taking loop
Plus: The real cost behind OpenAI’s surge
Hello, Prohuman
Today, we will talk about these stories:
Google turns NotebookLM into a real researcher
Microsoft’s cut tells a bigger story
A new course meets tech-shy travelers
Voice AI Goes Mainstream in 2025
Human-like voice agents are moving from pilot to production. In Deepgram’s 2025 State of Voice AI Report, created with Opus Research, we surveyed 400 senior leaders across North America - many from $100M+ enterprises - to map what’s real and what’s next.
The data is clear:
97% already use voice technology; 84% plan to increase budgets this year.
80% still rely on traditional voice agents.
Only 21% are very satisfied.
Customer service tops the list of near-term wins, from task automation to order taking.
See where you stand against your peers, learn what separates leaders from laggards, and get practical guidance for deploying human-like agents in 2025.
Google gives its notebook a working brain

Image Credits: Google
The update lands with the soft click of a tab opening.
Google is adding a “Deep Research” tool to NotebookLM, letting the assistant build multi step research plans, browse the web, and return a sourced report you can drop straight into your notes. It also adds support for Sheets, Drive URLs, PDFs, and Word files, which means the notebook finally accepts the messy formats people actually keep on their desktops.
This move shows Google wants NotebookLM to be more than a smarter scratch pad, because a tool that can fetch sources, structure a plan, and synthesize a briefing starts to edge into territory once reserved for junior analysts. It also reflects how users are working across split screens and mixed file types, especially since the mobile apps landed earlier this year.
Expect a shift in study habits once people realize they can run a deep query in the background while pulling numbers from a spreadsheet without breaking focus. The workflow gets tighter and the ceiling gets higher.
Someone will try this on a late night deadline and feel the room change a bit.
OpenAI’s bill arrives in plain numbers

The leak hit like a file dropped on a metal desk.
New documents show Microsoft collected $493.8 million from OpenAI in 2024 and $865.8 million in the first three quarters of 2025, a window into how fast the revenue machine is running and how heavy the compute burn has become. The payments stem from a 20 percent revenue share, paired with Microsoft’s own kickbacks from Bing and Azure OpenAI, which complicate the math inside both companies.
One thing stands out, OpenAI’s inference bill may be outpacing its income, with estimates putting 2025 inference spend at $8.65 billion by September. That figure lands differently when you remember most training compute comes from credits, while inference hits the checking account.
If the market leader is straining under cash costs, smaller players will feel that pressure in sharper ways, especially as cloud providers keep raising prices for peak demand. Investors will do their own math once they see how much depends on subsidized compute.
Somewhere in Redmond, someone is staring at a quarterly sheet and wondering how long this equilibrium holds.
AI tools meet the practical traveler

Image Credits: Boise State University
The classroom smell of old carpet meets a lesson on the future.
Boise State’s BroncoLearn just launched “AI Travel Tools for Adults and Seniors,” a course built by lecturer Margaret Sass that aims to make tech feel less like a locked door and more like a working map. It comes with a free Pressbook packed with examples and simple walk-throughs, including how to use translation apps at a bus stop or pull safety alerts before boarding a train.
Seniors rarely get courses built around their real travel friction, and this one shows how a little guidance can shift the moment someone hesitates before tapping an unfamiliar button. It also shows how student talent shapes the work, because instructional designer Kalyn Mitchell has been refining these modules since 2023 and knows where people stumble.
The ripple is simple, courses like this expand who feels able to travel without leaning on family for every small decision. More schools will copy the format once they see the turnout numbers.
Someone will open this course on a quiet morning and feel the first spark of permission.
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.



