Yakuza

Interesting move from Microsoft. The company just launched Copilot Cowork, built with Anthropic’s Claude, to automate complex workflows across Microsoft 365. Think assembling presentations, coordinating tasks and scheduling across apps. Alongside it, Microsoft introduced the Microsoft 365 E7 suite at $99 per user per month, bundling Copilot with a new Agent 365 platform. The key difference: Microsoft’s version runs in the cloud with enterprise data protections and its Work IQ layer. Another step toward AI agents becoming embedded in everyday enterprise software.

Gm

Nice agenda. But HK already losing 2 nil 😫😫

Happy Women’s Day to all the beautiful and smart ladies here!

Biceps day

Just a random Wednesday near Benfica Stadium of No Name Boys celebrating our anniversary https://www.instagram.com/reel/DVebyRFCEnv/?igsh=MTUzcjV2aHcxam05eQ==

Gm

Dinning with 3 Ukrainian ladies is always challenging! For them 🤣 (Sorry one decided to hide)

Wow

Today I woke up earlier than usual, around 6:30, and somehow ended up betting on weather predictions using government APIs 🤣 Turns out NOAA (the US government’s weather forecasting network) runs a $6B federal supercomputer that publishes hyper-accurate forecasts for specific weather stations across the US. And Polymarket lets you bet on whether NYC hits 42°F on a given day. So I spent some time doing this: - Find the Polymarket weather market for NYC for a specific day - Check the resolution rules, it was based on LaGuardia Airport specifically - Hit api.weather.gov/points with LaGuardia’s coordinates to find the right forecast grid - Hit the hourly forecast URL api to get the actual temperatures for that day - Compared against Polymarket’s temperature buckets Noticed the market had 40-41°F as favourite while NOAA was pointing clearly to 42°F Placed the bet in it. Now I wait 🤣

Future is here. Imagine this guy greeting you at the pub 🤣

Gm 🌞 Wishing you all a great day!

AI for the best causes ❤️ An AI tool called BRAIx can predict breast cancer risk up to four years in advance, according to a study published in The Lancet Digital Health. The model analyses mammograms and estimates future risk, even when scans initially appear clear. Among women ranked in the top 2% for risk, nearly one in ten were diagnosed within four years. Researchers say the tool could enable more personalised screening, directing higher risk patients to additional testing while reducing unnecessary checks for lower risk groups. A practical example of AI moving beyond chat and into clinical decision support.

Do you know what this means? It’s 🍻o’clock.

It’s not a convertible, but…

Hi!

More good news for Europe 🇪🇺 Amazon is committing €33.7B in Spain to expand cloud and AI infrastructure in Aragón. The largest tech investment in the country’s history. This is not just another data centre expansion. It is long term compute capacity, anchored inside the EU. At a time when AI demand is exploding and infrastructure is becoming strategic power. Europe gets criticised for “lagging” in AI. But compute is leverage. And leverage changes the narrative.

Anthropic went from being blacklisted to #1 app in the US, to crashing under its own growth. All in days. This is not just hype. It's power, positioning and principle colliding in real time. Pentagon tension. Mass adoption. Free developer certifications. Crash! The AI race is no longer just about models. Come listen to the lowdown on @pods https://pods.media/trust-me-bro/anthropics-claude-madness-days

This

Claude is Currently Experiencing a Global Outage As of today, March 2, 2026, Anthropic's Claude is facing a widespread service disruption affecting the web interface, mobile apps, and some API methods.  What is happening? • The Issue: Users are encountering "HTTP 500" and "529" errors, login/logout failures, and the message: “Claude will return soon. Claude is currently experiencing a temporary service disruption.”  • Scope: The outage is global, with significant reports coming from Europe, India, and Africa. While the core API was initially reported as functional, Anthropic later noted that some API methods are also failing.  • Status: Anthropic's engineering team identified the root cause around 13:22 UTC and is actively implementing a fix. Some users are seeing a partial recovery, but service remains intermittent. https://status.claude.com/