HN user

ubixar

15 karma
Posts0
Comments12
View on HN
No posts found.

C# gets close to this with records + pattern matching, F# discriminated unions are even better for this with algebraic data types built right in. A Result<'T,'Error> makes invalid states unrepresentable without any ceremony. C# records/matching works for now, but native DUs will make it even nicer.

The most interesting thing here isn't the CVE - it's the invisible coordination. A backbone provider acted on advance knowledge of a critical flaw, implemented filtering at scale, and the rest of us didn't notice until GreyNoise's data showed the drop. The vulnerability got patched at the network layer before it ever reached the application layer. This is what mature security ecosystems look like - the boring, quiet fixes that happen before the press release.

The most interesting finding isn't that hyperbolic growth appears in "emergent capabilities" papers - it's that actual capability metrics (MMLU, tokens/$) remain stubbornly linear.

The singularity isn't in the machines. It's in human attention.

This is Kuhnian paradigm shift at digital speed. The papers aren't documenting new capabilities - they're documenting a community's gestalt switch. Once enough people believe the curve has bent, funding, talent, and compute follow. The belief becomes self-fulfilling.

Linear capability growth is the reality. Hyperbolic attention growth is the story.

Just when you thought it was safe to use Opus 4.5 at 1/3 the cost, they go and add a 6x 'bank-breaking mode' - So now accidental bankruptcy is just one toggle away.

First time I've seen Josean switch to another OS fulltime - was always on macOS since I started watching him.

He's a Vim / Terminal super user and was always surprised that he stuck it out on macOS for so long, IMO Arch Linux / Hyprland is the best choice for VIM keyboard warriors.

been playing around with world models for sim-to-real transfer lately. the waymo approach looks solid, but curious how you're handling the distribution shift between generated scenes and real sensor data. any tricks for that besides the usual domain randomization?