attentive

claude code VSS - 70Gb

codex VSS - 600Mb

somehow Terra really struggles here even compare to Luna.

They were, until they decided to commit suicide for the service.

Iran calls notwithstanding, Mearsheimer is a complete clown with regard to Ukraine calls

2014 - Claimed Putin had no interest in annexing large parts of Ukraine, conquering it, or pursuing regime change

Feb 2022 - Putin had “no intention of invading Ukraine” in the sense of large-scale conquest (days before the invasion)

June 2022 - “There is no evidence in the public record that Putin was contemplating, much less intending to put an end to Ukraine as an independent state… Putin was not interested in making Ukraine a part of Russia… the Russians pursued a limited aims strategy”

Repeated claim (2014–2022): Putin is a rational actor who understands the costs and would not pursue maximalist territorial or imperial goals in Ukraine

missing gemini-3.1-flash-lite and gemini-3.5-flash

and both companies now have measures to lock their enterprise customers (who tend to sign year-long deals) at those API prices

why would enterprises do that if they can just use bedrock or vertex?

congress made h1b dual-intent in 1990

Tell us more.

"banana big but banana skin is bigger" - supposedly Orwell

Or you could use gcp Vertex or aws Bedrock and still have access to a bunch of FMs without a markup.

it's still somewhat gated behind "trusted access" for cyber, see https://chatgpt.com/cyber

You can pick the same browsers i.e. brave or chrome.

Also, let's not forget about unified memory impact. Raw cpu benchmarks are only one side of the complex system.

It's not like they don't own API's that those cli's are hitting. They have all the stats they need.

I look at those numbers and think that A18 pro is 25% faster as single thread is what matters for UX.

browserbench speedometer 3.0 on A18 pro - 33, Intel Core i5-1235U - 22

i9-13900KS gets about 33

M4 Pro - 44-50

Also I think it is telling that they are calling them SpaceX not X. The X brand is absolutely toxic, especially in enterprise.

it might not help all that much once it turns into "grok" harness or otherwise associated with elon

No need to leak. If we don't see a very public flow of various CVE's across the industry then it was a nothing-burger.

better question - will zettaflop ever own me?

Is there timeline mentioned anywhere on when any of this will be available for unprivileged public as in soon, not soon, never?

just do vibe performance optimization (I am not even kidding)

https://xcancel.com/altryne/status/2039005970865516955

  > Someone inside Anthropic, got switched to Adaptive reasoning mode
  > Their Claude Code switched to Sonnet 
  > Committed the .map file of Claude Code
  > Effectively leaking the ENTIRE CC Source Code
  > @realsigridjin was tired after running 2 south korean hackathons in SF, saw the leak
  > Rules in Korea are different, he cloned the repo, went to sleep
  > Wakes up to 25K stars, and his GF begging him to take it down (she's a copyright lawyer)
  > Their team decided - how about we have agents rewrite this in Python!? Surely... this is more legal
  > Rewrite in Py
  > Board a plane to SK
  > One of the guys decides python is slow, is now rewriting  ALL OF CLAUDE CODE into Rust. 
  > Anthropic cannot take down, cannot sue
  > Is this "fair use?" 
  > TL;DR - we're about to have open source Claude Code in Rust

bubblewrap is in many modern distros standard packages.

With all the supply chain issues these days onboarding new tools carries extra risks. So, question is if it's worth it.

it seems smart until you accidently type >input.json and nuke the file

I dream of a world in which all CLI tools produce and consume JSON and we use jq to glue them together.

that world exists and mature (powershell)

or better yet 3.1 Flash-Lite at $0.25/1M input

That was annoying to read because there is no easy to see impact of each change. It's FL2 + Gen 13 combined.

I.e. what's the FL2 benchmark on Gen 12 compared to FL1?

"Repositioning the taskbar is one of the top asks we’ve heard from you".

Tell us why it was removed in the first place, why it takes years to put it back and it's still future promises as of March 2026. That's just a clown show.

Please post it here. I'd also like to know if 5.4 mini is better than Flash 3. Include reasoning and timing, if possible.

token/sec is meaningless without thinking level. If it's fast but keeps rambling about instead of jumping on it then it can take a very long time vs low token/sec but low/none thinking.

What other alternatives exist in place of MCP? Arbitrary code via curl/sdks to call a remote endpoint?

cli?

for example aws cli. It's a full interface to aws API. Why would you need mcp for that?

and if you have any doubts, agents use it with a great effect even without any relevant skill. "aws help" is fully discoverable.

you’ll need to install the Windows app store from an enthusiast repo on Github

wsreset -i should do it