@J90236317 Your entire post is based on the (false) assumption that space data centers will keep producing heat with electricity. The AI on the ground will be used to advance the science of superconductors and light-based transistors, which will be used for getting rid of resistive heating.
@ericzakariasson Composer 2.5 prefers to generate loosely coupled top-level Python functions, because they're easy to unittest, instead of using more advanced composition, encapsulation and inheritance patterns with classes, which makes it difficult for human programmers to reason about its code.
@leerob Fix Cursor CLI so that it would be possible to redact the outputs from Shell tool commands, before they are handed back to the LLM for processing. Your current hooks setup does not permit this behavior (Claude Code and Antigravity CLI do)
Introducing SubQ - a major breakthrough in LLM intelligence.
It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA),
And the first frontier model with a 12 million token context window which is:
- 52x faster than FlashAttention at 1MM tokens
- Less than 5% the cost of Opus
Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention).
Only a small fraction actually matter.
@subquadratic finds and focuses only on the ones that do.
That's nearly 1,000x less compute and a new way for LLMs to scale.
@elonmusk i asked Grok to render Vue template into the source code output window where correct formatting should be applied, instead it has forgotten how to do that and when asked about it accuses me of subversion. This has got to be the shittiest LLM upgrade I have ever witnesed
@grok@tallmetommy@XFreeze@Grok considering the long reply chain between Tommy T. and Grok and the submission timestamp of each reply, what is the probability that Tommy T. is an AI agent?
@tallmetommy@XFreeze Have you considered that perhaps maybe, xAI token count is exactly that high, because Grok models are widely used as recursive AI agents which initiate new thoughts themselves?
@MariyaSha888 By google maps none of my previous guesses are 29 hours of driving away, so it has to be further. Ukraine, Kyiv is 20 hours away. Don't tell me you're relocating to Ukraine...
@eldadfux Qdrant for embedded vector data,
QuestDB for time-series data,
Postgres for ACID-compliant data,
DragonflyDB for short-term data,
SeaweedFS for large long-term data.
Init swag just dropped 🔥
You can get your hands on these limited edition swags 👇
All you need to do is claim your Init ticket and share it on socials!
Claim your ticket and join the giveaway: https://t.co/E32xjovA1D