Ricardo Vázquez (@ric2z) header
Ricardo Vázquez (@ric2z) profile picture

3x founder + cto. building something new Here + now.

the internet
Joined April 2009
Photo by Ricardo Vázquez (@ric2z)Photo by Ricardo Vázquez (@ric2z)Photo by Ricardo Vázquez (@ric2z)Photo by Ricardo Vázquez (@ric2z)Photo by Ricardo Vázquez (@ric2z)Photo by Ricardo Vázquez (@ric2z)Photo by Ricardo Vázquez (@ric2z)Photo by Ricardo Vázquez (@ric2z)Photo by Ricardo Vázquez (@ric2z)Photo by Ricardo Vázquez (@ric2z)
Ricardo Vázquez retweeted
Low-agency and high-agency people have different priors about the malleability of reality. Low-agency people have a prior that reality is fixed and the world has rules. Things are the way they are and your job is to navigate the world as it exists. When you encounter a barrier, that's the world telling you "this path is closed." You respect that. You find an open path. If all paths are closed, you wait for someone to open one. High-agency people have a prior that reality is negotiable. The world has defaults, not rules. Things are the way they are because no one has changed them yet. Your job is to shape the world toward what you want. When you encounter a barrier, that's just the world's opening offer. You make a counteroffer. You negotiate. You find or create leverage. The path isn't open? Make it open.
28
299
30
2,408
82,296
no they are not, all flights are grounded in the US
Your email might be down, but the skies are always open. ✉️✈️
127
i love this app
Now you can really see everything
1
39
Ricardo Vázquez retweeted
Anyone can build now. Few dare to aim big enough. That's why we're launching Southbound W27 by @thelabmiami & @focal_vc today. Code is now cheap, abundant, and written by machines. What’s scarce is real ambition, intelligence, compute, and distribution. It's what Southbound is built for: To help the best technical founders think bigger and move faster from day 1. We're now opening applications to the second cohort of our winter residency. If you want to: - Escape the cold - Work on an absurdly ambitious problem, surrounded by your best technical peers - Get weekly mentorship by unicorn founders - Funding on an uncapped MFN SAFE (vs giving up 7%) - $1M in perks and compute - and much more ... then join us in Miami from Jan 11 - Apr 11 2027. Southbound is: - Limited to 10 spots (teams or solo founders) - Fully in person in Miami - 7 days / week in office. The cohort is small by design: Instead of being a number in a big batch, you get a hand-picked group of your best peers, weekly dinners with unicorn founders, and real time with focal’s GPs. Rolling admission until all 10 spots are filled / applications close Dec 11. Apply asap (link in comments!) @ayalstern | @DanielDarling | @amiine_afia | @wynwoodceo | @SydneyWasi
29
33
20
98
41,691
wild
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math. It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training. We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵 Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math. rikiyat.github.io/blog/weight-sp… github.com/RikiyaT/lm_und… huggingface.co/Qwen/Qwen3.5-0… huggingface.co/RikiyaT/Qwen3.…
34
Ricardo Vázquez retweeted
do not forget to set your clocks on fire tonight because time is irrelevant
20
398
10
2,761
42,521
if they made good will hunting today, it would make no sense. matt damon’s character would have no future in mathematics
good will hunting but matt damon is just a regular janitor with a claude subscription
34
I am no longer afraid of AI taking people’s jobs.
We are literally doomed American High School English Teacher reads real responses from her student’s assignment after reading The Great Gatsby We have a serious education issue in America and if we don’t address it, we have a horrible future ahead of us These kind of responses are to be expected because 12th grade reading results in 2024 in America showed that reading levels right now the lowest they’ve ever been since the assessment began in 1992
27
Ricardo Vázquez retweeted
I gotta be honest, the whole qwen neuron thing, I don't like it. normies shouldn't know about this. I would vastly prefer they go back to talking about abliterated models or pumping memecoins or whatever it is they do
73
36
13
1,481
225,478
obviously these models are sleeper cells. what did you expect? 😩
While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected. Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions. Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent. But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models? We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem. Read the full report: cohere.com/blog/concerning-b…
34
Ricardo Vázquez retweeted
we're cooked
86
386
44
14,377
461,573
Ricardo Vázquez retweeted
343
1,782
436
20,618
2,991,219
incredible, lfg
I promise this will be the best 20 min you spend today! Robotics: Endgame, the sequel to my last year's Sequoia AI Ascent talk, "Physical Turing Test". I laid out the roadmap for solving Physical AGI as a simple parallel to the LLM success story. Be a good scientist, copy homework ;) And stay till the end, more easter eggs and predictions for your polymarket! 00:30 DGX-1 origin story at OpenAI, I was there in 2016 signing with Jensen and Elon. Heading to the Computer History Museum! 01:42 The Great Parallel 03:31 Robotics, the Endgame 03:39 Why VLAs fall short 04:32 Video world models as the 2nd pretraining paradigm 06:09 World Action Models (WAM) 07:46 Strategies for robot data collection and the FSD equivalent to physical data flywheel for robot manipulation 11:06 EgoScale and the Dexterity Scaling Law we discovered recently 14:00 Physical RL: bridging the last mile 15:39 DreamDojo: an end-to-end neural physics engine for scaling RL in silico 17:00 Civilizational Technology Tree and my predictions for the near future. Spoiler: it's closer than you think. Thanks to my friends at Sequoia for inviting me back to AI Ascent this year! I had a blast! Last year's talk is attached in the thread if you missed it.
77
Your life is as you are
39
Told you
Origin, our code hosting platform, is now live. It's fast, easy to use, and deeply integrated with Cursor. Get started by syncing your repos from GitHub.
1
1
151
Ricardo Vázquez retweeted
if you watch youtube shorts or any short form video content you are a slave
122
64
20
1,475
47,847
To understand how to build a community: shut down all short-form algorithmic consumption in your life
1
75
Overthinking is troublesome for people and AI alike. We might have to teach models how to relax
New Anthropic Research: “Inverse Scaling in Test-Time Compute” We found cases where longer reasoning leads to lower accuracy. Our findings suggest that naïve scaling of test-time compute may inadvertently reinforce problematic reasoning patterns. 🧵
92
Has anyone tried making an LLM play RollerCoaster Tycoon?
1
44
How you do anything is how you do everything
37