AI Progresses, Brainrot IDE, & Space AI

1/ AI initially was… people taking notes

You’ve probably used Fireflies or some AI notetaker before, right? Wild fact: in the early days, it wasn’t AI at all. The founder literally joined meetings himself and manually took notes—pretending the product was fully automated.

The ultimate “do things that don’t scale” move.
This is how real AI companies actually get started.

2/ YC Funds a Brainrot AI IDE

YC just funded what might be the most unhinged but oddly brilliant idea of the year: the TikTokification of coding.

Imagine writing code, approving AI edits, and watching TikTok/Subway Surfers side-by-side…inside your IDE.

It sounds ridiculous, but honestly?
Given how people consume content today, it just might work.

3/ “Grok, is this true?”

Tagging an AI in a reply today is basically summoning a digital oracle.
People are casually calling on a superintelligence to fact-check memes, win debates, or settle arguments.

It’s tongue-in-cheek, but also very real:
We now have an always-on, hyper-smart entity we can cite on demand.

4/ Grok 5 is on the Horizon

Each Grok jump has been massive, and with training scaling aggressively, Grok 5 could mark another huge leap in capability.

5/ Jensen: “We’re not in an AI bubble.”

Jensen’s core argument is spot on:
Traditional software is static. AI is adaptive, context-aware, and capable of real-time reasoning.

We’re rewiring how software even works, and a moment like this happens only once every couple of decades.

6/ Bezos: AI Data Centers… in Space?

Bezos believes the future might involve pushing massive compute off-planet.
Launching AI data centers into space sounds sci-fi, but with falling launch costs and rising compute needs, the idea isn’t as crazy as it seems.

It’s a huge engineering challenge, but it’s also a potentially inevitable next step.