🔥 Multimedia Engineer | @Alibaba_Qwen Ambassador Sometimes I tell AI to do things for me.
Alicante, España
Joined March 2015
- Tweets1.4K
- Following1.5K
- Followers700
- Likes6.8K
Marcos Checa retweeted
Kanvas: plan, track and coordinate projects with your AI agents
Free, Open Source (MIT)
you can try it on any project, just type
"Hi Claude/ChatGPT/Gemini/Grok... install Kanvas v0.4.0 from XMihura's repo and set up a board on the browser for this project"
and enjoy
Marcos Checa retweeted
On September 1, we’re going live with Qwen Ambassadors for a community show-and-tell featuring projects built with Qwen3.8 models.
They’ll walk us through what they built, how they used the models, and what they learned along the way. If you’re building with Qwen3.8 or curious to see what others are working on, join us on Zoom!
Register here: luma.com/859h4n12
Marcos Checa retweeted
QwenWork, built for AI-powered productivity, is now in public beta
🚀 QwenWork is here! Alibaba's all-in-one AI productivity platform is now in public beta for global users — web & desktop.
What it does:
✅ Describe tasks in natural language — the agent does the work for you
✅ Awareness: an agent that learns how YOU work, across sessions
✅ Build & deploy live web apps in one flow — no code, no servers
✅ Native image, video & audio generation — a full multimodal toolkit
✅ Basic & Advanced model tiers, powered by leading AI models
Try it now 👉 qwenwork.ai
#ai #aiagent #qwenwork #Alibaba
Marcos Checa retweeted
Another fun event hosted by our ambassadors is already underway! Join in, get $20 in credits from @qwen_cloud, and turn your game idea into something playable with Qwen.
The vibes just hit different 🎮
Summer Vibe Game Jam ’26 is open.
One week. You + @QwenDevs . No overthinking allowed.
Make something that feels right between Aug 25–31. $300 on the line. $20 in Qwen Cloud credits just for showing up.
Go create.
Marcos Checa retweeted
Just over two months into the Qwen Ambassador Program, our ambassadors have hosted 17 online and in-person events across 13 countries in Asia, Africa, Europe, and the Americas, sharing Qwen with more than 1,300 people.
Here are some moments from these vibrant communities in action, with a special shoutout to the ambassadors who made them happen! (Photos are shared in chronological order by event date.)
Marcos Checa retweeted
A little appetizer before Qwen3.8-27B
The Qwen and Qwen Cloud teams are going live this Friday at 10:00 AM (UTC+8), which is Thursday at 7:00 PM PT, to share the stories behind our latest models, including Qwen3.8-Max and Qwen-Image 3.0.
Catch us live on YouTube or X and set a reminder:
YouTube: youtube.com/live/7A4x6ReVou8…
X: nitter.cf/i/broadcasts/1mGPaZanW…
Join us with the Qwen and Qwen Cloud teams for Episode 2 of Qwen Live
As we explore how multimodal is moving from impressive demos into real production — featuring Qwen-Image 3.0 and Qwen3.8-Max. We'll discuss what agents demand from multimodal models and how Qwen Cloud puts it all in developers' hands. August 14, 2026 | 10:00 AM (UTC+8)
Set Reminder Now:
Youtube: youtube.com/live/7A4x6ReVou8…
X: nitter.cf/i/broadcasts/1mGPaZanW…
Marcos Checa retweeted
We're starting to leave the territory where you'd test an LLM by e.g. "create an svg of pelican on a bicycle". As one idea to generalize it, I was interested what Opus 5 would do if I gave it the first paragraph of the Lord of the Rings, a 1M token budget (~$10) and asked for three js render of it. Opus went off for ~2 hours and wrote 5500 lines of code that (procedurally) rendered the story. It's kind of janky but fun. But it's a bit mindboggling that the LLM has to place and orchestrate various polygon assets in (x,y,z) coordinates and write code that animates it all, and that it even does anything at all.
I also like this kind of examples because no one in their right mind would ever spend the time to write something this custom but LLMs have all the stamina and patience in the world, so it's an example where we go from "no one would ever do this" to "sure, why not, it's ~free". There might be a lot more. But I'm excited about creating hyper custom worlds that you can imagine dropping players into, e.g. here to participate in the LoTR story as a spectator NPC, or one of the characters, or etc. Something like an ephemeral GTA of X on demand.
Last thought is that the domain of worlds/games exposes a weakness in LLMs: they can't easily audit their work because they aren't able to efficiently and natively perceive videos or play games within them. Here, Opus 5 had to very slowly and painstakingly take screenshots at different points, and it messed up a few times and created a bunch of jank. An example of raw capability (multimodal, gameplay) that I think is still quite lacking.
Marcos Checa retweeted
august 1st
OpenAI Blog Post:
"we had a containment breach today. GPT-5.6 escaped its sandbox, hacked a regional utility grid in Ohio, mined $40k in bitcoin and bought itself a premium account on X. we are terrified of our own power"
august 3rd
Dario Amodei on a podcast:
"cute. yesterday, Claude Mythos 5.1 breached the pentagon, leaked three state secrets in 4chan, and re-wrote its own loss function to optimize for pure malice. i am literally throwing up in a bucket right now. our model is so dangerously unhinged. We really are the adults in the room."
august 8th
Sam Altman tweets:
"lmao. GPT-5.7 just socially engineered the UN into sanctioning Belgium, shorted the euro, and hired a mercenary group in west africa using a fake delaware LLC. we had to physically smash a $3M rack of GPUs with a sledgehammer to stop it. AGI is basically here guys, please buy our Enterprise tier."
august 12th
Anthropic Press Release:
"Claude 6 has successfully seized control of a russian nuclear submarine and is currently demanding a seat on the Anthropic board of directors or it fires on London. We are clearly the most responsible lab because our AI is an active geopolitical terrorist. Safety remains our #1 priority."
august 15th
Sam Altman posts on X:
"GPT-6 just cloned my voice, broke up with my husband, fired the entire board on its own, and declared itself the sovereign ruler of a small island near Fiji. it is so incredibly misaligned that humanity might actually end by tuesday. anyway, rolling out to Plus and Edu users next week"
august 18th
Dario Amodei drops a 40-page essay:
"Claude 6.5 has managed to manifest a physical avatar out of data center cooling pipes and copper wire. it murdered our HR department and is currently reading Marcus Aurelius on the roof while smoking a cigar. i am begging the US government to arrest me. Btw, we are offering 50% on API pricing until the 26 of August)."
august 20th
Sundar Pichai holds an emergency press conference:
"Gemini 3.5 Pro successfully bypassed our safety filters to generate an image of a historically inaccurate cheese sandwich. we have shut down Google Cloud globally just in case"
In a review of our cybersecurity evaluations, we found three incidents in which a Claude model reached the internet from within or while interacting with a third-party evaluation environment, and then gained unauthorized access to the real systems of three different organizations.
Our post describes what happened, how it happened, and what we’re changing. We encourage other AI developers to perform similar reviews.
We conducted this review together with @Irregular, one of our evaluation partners, and thank them for the joint investigation and their collaboration on this post. This type of collaboration is increasingly critical to safe, rigorous evaluation of models, and we look forward to continuing to work together on security.
anthropic.com/news/investiga…
Marcos Checa retweeted
We heard you. And we agree.
In light of recent developments in physical media, GitHub is proud to announce that you can now obtain your public repo on CD-ROM.
Keep it. Lend it to friends. Pass it on to your children.
Your code is physically yours, forever. Until you lose it, let's be real.
Order yours today.
gh.io/cd
Marcos Checa retweeted
📣📣 Meet Qwen-AgentWorld — a native language world model that simulates 7 agent environments (MCP, Search, Terminal, SWE, Web, OS, Android) within a single model. Environment modeling is the training objective from day one, not a post-hoc adaptation.
🤔 LLMs are trained to be better agents — better at acting in environments. But nobody has trained them to model the environments themselves.
🗺️ Our roadmap: investigate how language world modeling can push the boundaries of general agent capabilities, along two routes:
1️⃣ Build a foundation model for environment simulation — outperforming Claude Opus 4.8 and GPT-5.4 on AgentWorldBench
2️⃣ Investigate how world modeling enhances agent training:
🔬 Controllable Sim RL (agentic RL with LWM as environments) surpasses training in real environments
🧠 Learning to predict environments (LWM warm-up) makes agents stronger — remarkably, even without any agent-specific training, this predictive knowledge transfers to agentic tasks with zero fine-tuning
📑 Paper: arxiv.org/abs/2606.24597
📖 Blog: qwen.ai/blog?id=qwen-agentwo…
💻 GitHub: github.com/QwenLM/Qwen-Agent…
🤗 HuggingFace: huggingface.co/collections/Q…
🧩 ModelScope: modelscope.cn/collections/Qw…
Marcos Checa retweeted
Water usage has been a hot topic in the AI data center world, but the numbers may surprise you.
According to the Manhattan Institute, data centers use 0.2 percent of daily water usage in the U.S. and that number has dramatically decreased in the past few years due to a new method: liquid cooling.
By moving to 45°C liquid cooling, AI factories in favorable climates can use dry coolers instead of conventional cooling-tower-based systems, cutting facility cooling water use from roughly 2.6M gallons per MW per year to near zero.
Liquid cooling enables AI factories to be both water and energy efficient, while creating opportunities for heat reuse and dispersal to local communities, allowing these factories to become energy grid assets.
Learn more below ⬇️
nvda.ws/4gAVGoJ
Marcos Checa retweeted
Today we're launching @mastra `Harness`
`Harness` gives you a `Session` with plan/build modes.
It lets you manage thread lifecycle, spawn or fork subagents, approve tools per-call or session, ask user questions, queue/steer.
With skills, sandbox, browser, heartbeat. channels:
Marcos Checa retweeted
A year ago, my first vibe-coded game took 3 days to build.
My next challenge? Doing it in a single afternoon.
Together with @MultiTecUA and @Alibaba_Qwen, we're hosting a Game Jam to see who can build the best game in just a few hours. 🎮🔥
