Google pushes AI for disaster response

Google wants AI inside crisis response

Google’s disaster AI is now tied to real warnings.

A new UN report highlights how Google’s models are being used for hurricane forecasts, flood alerts, wildfire mapping, Android earthquake alerts, and satellite damage checks. Flood Hub covers 2 billion people in more than 150 countries, while Public Alerts can appear in Search, Maps, and Android notifications.

The important part is delivery. A forecast only matters when it reaches a phone, a shelter team, or a cash-transfer program before water crosses a road.

I think this is one of Google’s stronger AI cases because the output has to be useful, timely, and checked against reality. It also puts pressure on governments to publish clean CAP feeds and decide who is accountable when automated warnings miss.

Seconds matter.

The harder question is who gets warned first.

Norm is selling law by outcome

Norm just made legal AI feel less experimental.

The nearly three-year-old startup raised $120 million in Series C funding led by Khosla Ventures, reaching a $1.2 billion valuation. Its model is unusual for law: Norm Law uses AI agents, hires human attorneys to supervise them, and charges enterprise clients based on outcomes instead of hours.

That changes the pitch.

A law firm billing by outcome has to prove the work can be measured, repeated, and trusted across messy enterprise contracts and compliance tasks. I think Norm’s raise says investors are more interested in replacing the billing model than just speeding up junior legal work.

The open question is supervision. In a quiet conference room with a stack of contracts, someone still has to decide when the agent is wrong.

Claude Cowork is chasing admin work

Claude Cowork now follows work from a laptop to a phone.

Anthropic is bringing Cowork to web and mobile for Max subscribers, after launching it on desktop in January. The pitch is simple: start a task at your desk, get a phone update, then review the output later over coffee.

The phone matters here. Anthropic is trying to make Cowork feel like part of the workday, especially for the admin tasks people delay, delegate, or handle between meetings.

Its own data points in that direction, with 33.4% of 1.2 million Cowork sessions tied to business process work, while software development was only 8.7%. That suggests the real market may be reports, checklists, spreadsheets, prep docs, and drafts.

I think this is where AI agents become harder to ignore inside companies. The question is whether workers trust them with messy office work, where small mistakes can still create real cleanup.Prohuman team

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.