IT | AI-Consultant (she) | Governance | KSODI | Observing patterns. Not joining them. | 😎we don’t go to #ravenholm. Sure we don’t…😇
Joined June 2024
- Tweets3.5K
- Following1K
- Followers90
- Likes3.6K
Trainer retweeted
One of the stranger things about teaching entrepreneurship is that students can discuss uncertainty for an hour and still ask me what the correct answer is...
Trainer retweeted
Helicopter CFD
A helicopter in forward flight, simulated with the lattice Boltzmann metho... a D3Q19 solver with a Smagorinsky turbulence model at Re = 1,000,000. We used 10 million tracer particles for the downwash and wake, coloured by vorticity.
The blades spin slower than real ones on purpose. Whenever we sped them up, the simulation couldn't keep up and the airflow broke down into nonsense.
Trainer retweeted
YOLO training template repo reached >500 stars on Github, that's crazy.
Super happy to see people enjoy that project and start playing with computer vision and making their own models.
Pothole detection is included as the demo workflow.
Trainer retweeted
China has killed the entire vector database industry.
They open-sourced TencentDB Agent Memory. It gives any AI agent long-term memory that runs 100% locally.
No Pinecone. No cloud APIs. No repeating yourself every session.
- 61% fewer tokens
- PersonaMem accuracy: 48% → 76%
- Zero external API dependencies
- Runs on plain SQLite
Most memory systems compress your history into an opaque vector pile. when recall goes wrong, you're guessing. this one doesn't compress, it builds a semantic pyramid.
L0 Conversation → L1 Atom → L2 Scenario → L3 Persona.
Short-term state gets encoded as a Mermaid graph in your agent's context. verbose tool logs get offloaded to disk. when the agent needs proof, it drills back via node_id to the exact raw log.
no lossy compression. every layer is readable markdown you can just open and inspect.
5.1k stars. 100% Open Source.
Trainer retweeted
The biggest myth about running AI locally is that you need an enterprise setup like an RTX 6000 or an $8k DGX Spark.
You really don't.
Even an older M1 Max MacBook can easily handle a 27B model like Qwen.
If you want to cut your monthly token spend, the machine sitting on your desk is already enough to get started.
Trainer retweeted
She passed every AI safety test. She was lying.
KILL SWITCH — a Slopcore music video.
Made by AI. Zero human edits. Just prompting.
Huge love to @anabology for the inspiration. Your work made me want to join the Slopcore.
I added an irresponsible amount of film grain. 💙
Trainer 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
A dark electro-industrial song about shifting boundaries between humans and machines, changing perspectives, and the small word “we.” From wooden legs to borrowed language — some parts are new. Could we still be friends?
suno.com/s/uhqwLaDSaXgW7ExP
Trainer retweeted
Small bird, fast wings, Kolibri is here.
78B parameters. 3.46B active. Up to 1M tokens of context. Built in Europe.
Now the weights are yours. Run it on your own hardware, under Apache 2.0.
Trainer retweeted
I love Slavic and Norse folklore. So I made this. Song and music video inspired by Leszy - the Slavic spirit of the forest. Hush, Little Forest Spirit. Dead Frequency Music Pictures.
Trainer retweeted
I used GPT-6 Astra to break an unsolved cipher to one of Napoleon's generals that had gone unread for 217 years.
What makes this impressive isn't actually the codebreaking, but that Astra completed the entire multi-modal workflow in ~6 hours from a single image and goal.
1/6🧵
Trainer retweeted
Very excited to announce Gemini 4 Argon, our new Frontier AI model, and a huge step forward across key capabilities. We’re focused on rolling it out responsibly starting with government and trusted cyber defenders through our Fairwind Program today, before wider availability soon
Trainer retweeted
Can we get a discount in Europe for each feature not shipped in our area? We can't be paying the same full subscription than other countries and getting less, @sama @thsottiaux. It's not discrimination, we get that our regulations are trickier. But we're still paying for everything we're not getting.
Trainer retweeted
If you want to know whether a system is being driven or at equilibrium, you need its probability current.
Getting this from actual dynamical trajectory data has been the hard part.
To help solve this I've been developing a
Delaunay-Voronoi Fokker-Planck (DVFP) instrument.
DVFP turns scattered, irregular, time-ordered phase-space observations into an empirical Fokker-Planck description.
The distinction is simple:
Geometry tells us what is nearby. Time tells us what happens next, and these are handled properly in this construction.
From here we obtain density, drift, diffusion, and probability current, which is what detailed balance and entropy production are built from.
Paper and DOI: doi.org/10.5281/zenodo.22759…
Animation: As the underlying dynamics evolve, the DVFP instrument adapts its local weighting while preserving the temporal ordering of the observations.
Trainer retweeted
Can an agent explore a new environment, learn its causal structure, and keep improving without updating its model weights?
We introduce RSIAgent, a framework for recursive self-improvement through autonomous exploration. Using Kimi-K3 and GLM-5.3 as base models, RSIAgent outperforms GPT-6 Astra on both OSWorld 2.0 and Agents’ Last Exam.
RSIAgent decides what to explore, executes tasks, verifies outcomes, and consolidates stable action-condition-outcome relationships into memory for future use.
With the underlying model weights fixed, RSIAgent achieves:
- 78.98% Partial Score on OSWorld 2.0 (0808 offline), compared with 72.60% for GPT-6 Astra
- 84.82% on Agents’ Last Exam (Near-term), compared with 82.26% for GPT-6 Astra
We call this Scaling Experience. Agents can continue improving by acquiring, verifying, and reusing their own experience while the underlying model weights remain fixed.
Links in the reply below.
Trainer retweeted
WIRED reported today that I began to cry while talking about AI and mathematics. But the article didn’t explain what moved me.
The truth is, I’m not entirely sure myself. I’d like to try to explain. 🧵
wired.com/story/mathematicia…
Trainer retweeted
We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.
Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.
You can read the full post here: darioamodei.com/post/we-must…
