Loading activity…
@andrew_n_carr
co-founder leading science @getcartwheel co-founder advisor @arcade_ai Past: Codex @OpenAI, Brain @GoogleAI, world ranked Tetris player
AI
With both models at max effort, Ember matched Claude Haiku 5.5’s Intelligence Index score at roughly one-third the task cost.
AI
The model supports image inputs, with fine-tuning and inference available at launch and downloadable weights planned for later.
AI
The exabox’s refundable preorder deposit is $100,000; Tinycorp’s storefront marks preorders out of stock.
AI
Asmov-1's team reports a 0.007% error rate across all evaluation tracks and expects an open-source release in early November.
XAI
OpenAI says eligible subscribers could start claiming the credits October 28 to build apps and agents.
XAI
A post says Carbon scored 88.2% on SEC-bench Pro and is rolling out first to cybersecurity specialists.
AI
Liquid AI pitches a tri-hybrid design, single-GPU fine-tuning and on-device use without detailing hardware requirements.
AI
HazyResearch claims up to 31× faster generation on summarization and translation than similar-sized diffusion models, with comparable or better quality.
AI
In an Oct. 26, 2026, post, @qberthet described a six-month role; Google's application covers multiple research teams.
AI
A post says the model’s weights are available under Apache 2.0, with support for a 512K-token context window.
AI
A user quotes a “neolab” friend saying it could resell compute even if it fails to train the best model.
AI
A post cites an “explosion” of Microduck and Reachy Mini as the reason for the call.
AI
The Humanoid Hub says Brett sees pre-training at scale as key to winning in robotics.
AI
Moonshot says both versions are live on Kimi.ai, Kimi Work, Kimi Code and its API.
AI
NVIDIA says Vera Rubin offers Day 0 support for the model, with throughput of up to 200K TPS/GPU.
AI
The contest page lays out GPU-kernel and SGLang serving work, with different tests for each.
AI
Zhan says Sequoia, Jeff Dean, Sachin Katti and Philip Wong will be among the fund’s limited partners.
AI
The @flappyairplanes account names Index and Lux as co-leads and says the team spent a year on GPU primitives.
AI
Its developer says Flint exceeds 125 tokens per second on an M4 Pro MacBook Pro at short contexts.
XAI
The vLLM project estimates $38.96 per GPU-hour in revenue on AgentX at 196 P90 TPS, using DeepSeek’s off-peak API pricing.
