Chris Dukes retweeted
Learn linear algebra. Learn to visualize matrices as spatial transformations. Learn to see calculus as the language of optimization. Learn vector spaces until neural networks stop looking like black magic and start looking like high-dimensional geometry. A deep understanding of math transforms you from an API-wrapper developer into an AI architect.
Chris Dukes retweeted
Why do I keep talking about base:0xacfe6019ed1a7dc6f7b508c02d1b04ec88cc21bf. Because on the long horizon I see two different futures -- one where big tech and AI controls us all like cattle, or one where people can exercise liberty, and venice.ai stands at the crossroad as a bridge from the former to the latter. We can call it 'Private AI', which is mostly a self-explanatory description. But what it 'IS', is the sovereignty of intelligence. Not just artificial intelligence, but your own included. It is a duty, it is a necessity, it is "support this or we're fucked".
This is the single most important time in our entire human history. A critical juncture and inflection point where we either decide to hand ultimate power to Governments and Corporations, or we hold it for ourselves - 'the people'.
Every time you use ChatGPT, Grok, Gemini, Claude, Meta... you are building your own prison. And you won't even realize it until you need to escape it. I'm not saying boycott anything, I'm just saying be conscious about the road we're travelling on. You have a better alternative without compromise, so use it. Keep your data private -- not just for your sake, but for everyone's sake.
In my view -- it is a moral obligation to use venice.ai instead of it's data harvesting competitors. As the saying goes; 'know thy enemy' & 'know thyself' -- if you don't want to be treated like an enemy to big tech and gov, then be an enigma to them, lest the authorities come knocking because they deem that your thoughts are too dangerous or unacceptable. You don't care enough until it happens to you, but it sure as shit is happening to others. It's not a matter of time, its a matter of choice.
Choose wisely, because this really is a Rubicon moment.
#AI #PrivateAI
Chris Dukes retweeted
A Chinese tech company from 🇨🇳Shenzhen, China's tech hub, just combined humanoid robots with industrial robot arms!
This is what happens when your country invests in technologies instead of funding wars!
Chris Dukes retweeted
CodeRabbit is joining the OpenAI Daybreak Defense Network.
CodeRabbit Secure now runs on GPT-6.1 Sol and OpenAI's GPT Cyber models. Early users have already found and fixed broken access controls and authentication failures with it.
Deep Scan is free until Nov 8, run it on your repo and see what it turns up.
@OpenAIDevs
coderabbit.ai/blog/one-milli…
Chris Dukes retweeted
NVIDIA just put 2,575 simulation-ready 3D assets on Hugging Face 🤯
A 472GB collection covering everyday objects, tools and industrial equipment.
It’s called NVIDIA SimReady Assets. The models include physics settings and semantic labels for use in simulated environments.
- Train robots to pick up and move objects
- Build virtual factories and warehouses
- Generate labelled training images
- Populate 3D scenes with ready-made objects
Useful for robotics research, digital twins and anyone building detailed virtual environments.
Download the collection below 👇
Chris Dukes retweeted
Today we’re releasing Muse Spark Open, the most capable US open-weight model to date.
Built for coding and agentic work, it excels at long-horizon tasks, complex software engineering, and multi-step reasoning.
Key Model Specs:
→ 816B total / 42B active parameters
→ 1M token context window
→ Natively multimodal across text, image, video, audio
Chris Dukes retweeted
As the exclusive sponsorship begins, another big thank you to @Omarchy's Omacom Foundation for sponsoring Hyprland for 3 years!!
This will allow Hyprland to smoothly continue development for the coming years and bring more and more hypr into linux.
omarchy.org
Chris Dukes retweeted
No! I'd call it the most disappointing statement from bunch of high priests of math! Math isn't any guild's property; it belongs to all humanity. Instead of cherishing this incredible moment in math history, they're mad their blessing was ignored. The days of gatekeeping are over
Chris Dukes retweeted
Nick Saban warns against the biggest enemy to continued success.
"Complacency...when you get satisfied - it creates a blatant disregard for doing the right things."
"You start to resent the very things that got you where you were."
You no longer pride yourself in hard work and earning it.
"We pride ourselves in hard work. Then it becomes: 'I resent that. I resent the hard work that you're making us do.'"
"'Why do I have to do this? I just had a great month. I just sold however many cars. We won so many games. Why do I have to do this?'"
That's the trap. You start questioning the process that built you.
Then he dropped the truth:
"Success is momentary. It's not continuous. It's right then - this is what you did - that's over and done with."
"You better be where your feet are and move on to the next one."
"Because when you are successful, you just become the target."
Success doesn't protect you - it exposes you.
It's a continuous process. You have to keep earning it.
This is why people tell you to stay hungry and stay humble because you can't resent the work that got you there.
(🎥CBT Automotive )
Chris Dukes retweeted
Meet Mistral Large 4, aka Le Chonk.
• 1T parameters, natively multimodal. 49B active.
It is the best open weights model from US or Europe on aggregated benchmarks.
• State-of-the-art on critical workloads, including cyber defense, manufacturing and finance and it surpasses closed frontier models on visual grounding.
• Forged in Europe end-to-end and is deployable from Europe via our own Mistral Cloud infrastructure.
• Available to all via API today. Working with cybersecurity partners privately.
Open weights release end of October.
🤖 Made with AI
Chris Dukes retweeted
Mark Cuban's take on the next job wave deserves more attention than it's getting.
His argument: software as we know it is dying. Generic SaaS gets replaced by AI customized to how each business actually operates.
The question then becomes — who builds and deploys that for the 33 million companies in the US that can't do it themselves?
That's the market. Not AI for enterprises with 500-person eng teams. The other 99%.
SMBs don't have CTOs. They don't have prompt engineers. They need someone to sit across the table, understand their workflow, and wire up an AI system that fits it.
That's a services business. Probably the largest one this decade.
Chris Dukes retweeted
Yann LeCun's latest talk at ETH Zürich
"You should not work on LLM. At least if you're in academia, you should absolutely not work on LLMs. There is nothing you can bring to the table.
If you're interested in making real progress in AI, in sort of grounded AI for the real world, if you want physical AI, don't work on LLMs, and don't work on generative models either.
So, as you can probably guess, this does not make me very popular in Silicon Valley."
----
From "Perfology Clips" YouTube channel, (link in comment)
Chris Dukes retweeted
.@ylecun is absolutely correct. The risk posed to humanity from Dario and his crazy cult is far greater than the "AI risk".
_
Chris Dukes retweeted
The same model, with the same weights, scores 62% in one agent harness and 33% in another.
@adithya_s_k and the @huggingface team just released the ultimate guide to multi-harness RL, and it's one of the most practical RL write-ups this year, and everything open!
The trick is simple. Don't touch the harness. Point it at a proxy instead of the model. The proxy speaks all four API formats coding agents use (OpenAI Chat Completions, OpenAI Responses, Anthropic Messages, Gemini). It records the exact token ids and logprobs vLLM sampled, and you train on that. You don't change a single line of Claude Code, Codex or OpenCode.
Results:
🔹 Trained across 4 harnesses at once, LFM2.5-2.6B by @liquidai went from 42% to 54%
🔹 31% fewer tool calls, thanks to a small bonus for solving tasks in fewer steps
🔹 Training in OpenCode alone took OpenCode from 34% to 58%, but the multi-harness model improved everywhere
They also tried the shortcut everyone reaches for: fine-tune on 3,189 successful rollouts from Qwen3.8-27B. Imitation plateaued at 47.5%, below both RL runs. Copying a bigger model doesn't get you there. Practice does.
The best part is that everything is open: the capture proxy in OpenEnv, the trainer in TRL, the tasks, the SFT data, the training code and all seven trained models.
Agents will run in dozens of harnesses.
Now open models can be trained for each of them, by anyone.
Read it here 👇
huggingface.co/spaces/FineEn…
