Marco Mascorro

RT @wllottnwldr: Marriage is scary bc what if she doesn’t agree to this floor plan https://t.co/AqsYyezycr

RT @pmddomingos: Late nights are for working. Mornings are for sleeping.

RT @cursor_ai: GPT-4.1 is enabled for all Cursor users! We're launching it for free for the time being, to let people get a feel for the m…

RT @amanrsanger: Grok 3 is available for all Cursor users! The mini model is available for free. Curious to hear what you think. In our in…

This is super neat! Computer Use Agent Arena: a way to test computer-use agents in the wild without any setup. Available to test models like OpenAI Operator, Claude 3.7, Gemini etc on Windows and Ubuntu (MacOs coming soon):

Bowen Wang @BowenWangNLP

🎮 Computer Use Agent Arena is LIVE! 🚀 🔥 Easiest way to test computer-use agents in the wild without any setup 🌟 Compare top VLMs: OpenAI Operator, Claude 3.7, Gemini 2.5 Pro, Qwen 2.5 vl and more 🕹️ Test agents on 100+ real apps & webs with one-click config 🔒 Safe & free https://t.co/Btycd6jxJZ

Llama 4 is here! with 2 models available: - Maverick a 400B multimodal model with 17B activate parameters and 128 experts and 1M context window. - Scout a 109B multimodal model with 17B active parameters, 16 experts and 10M context window https://t.co/cfTydsnBU6

Gemini 2.5 Pro is so good, it definitely feels like Gemini has caught up, and it might just become my default model.

Gemini 2.5 Pro is now on Cursor with 1M content window 😀

Cursor @cursor_ai

Gemini 2.5 Pro is available to all Cursor users! You can enable the full 1M context window if you'd like. We're curious to hear how you think it compares to Sonnet.

RT @_akhaliq: Alibaba Qwen just dropped Qwen2.5-Omni-7B on Hugging Face you can now talk to Qwen like chatgpt This is an omni model, whic…

San Fransokyo https://t.co/2CqFklqKcc

This has to be one of the coolest tools to have at your home/lab. A FDM 3D printer, a laser engraver/cutter, and a plotter, all in one for $1.8K. The Bambu FDM 3D printers have been great for me and "they just work".

Bambu Lab @BambulabGlobal

Meet the H2D & H2D Laser. 🎉 Our most advanced multifunctional manufacturing hub yet. Ready to take personal manufacturing to the next level? Order now. #BambuH2D #RethinkPersonalManufacturing https://t.co/ZpE6aPQZ8K

It's still surprising that given all the economic potential for Agents and the euphoria around them in the last ~2 years, the benchmarks and datasets out there for measuring progress are very small, limited and narrow. We haven't gone beyond WebArena, WebVoyager, OSWorld,

Deepseek V3 0324 is a great model.

The new ARC AGI 2.0 benchmark is out and we are super excited about it. It was an honor for @GEVS94 and I to have hosted the preview of it at @a16z. Congrats on all the great work @mikeknoop, @fchollet, @GregKamradt and @arcprize Team: https://t.co/bzQGGJdlri

ARC Prize @arcprize

Today we are announcing ARC-AGI-2, an unsaturated frontier AGI benchmark that challenges AI reasoning systems (same relative ease for humans). Grand Prize: 85%, ~$0.42/task efficiency Current Performance: * Base LLMs: 0% * Reasoning Systems: <4% https://t.co/k9o7t031fN

New DeepSeek V3 model just dropped today. Maybe the base for R2? https://t.co/BVOP6plR9s