Scientist @AmazonScience | Prev: @brave, @Samsung AI, @Cambridge_CL, @CERN | Coffee drinker, traveller, tech/car enthusiast | Opinions are my own
London, England
Joined December 2011
- Tweets534
- Following679
- Followers488
- Likes981
With the wave of reverse engineering everything, I am looking forward to better printer software/firmware! It's been terrible for too long ... 🫠
I love the surge of tinkering with your purchased hardware and RE their S/W. It's been long overdue & gives control back to the user. Security through obscurity doesn't cut it. With covert surveillance on users, orgs risk their reputation. Banning tool usage, also reprehensible.
Why are people putting NeurIPS in their handles? It's still October ... Reminds me of Christmas decorations 🎄just after Halloween 🎃
Commendable effort! 👏
1. Still, it's a single point of failure, isn't it?
2. Also, 3.5 months is a sprint. Probably today's landscape resembles more of a marathon. Can such an effort scale in time? Maybe ...
We formally announced Mirendil less than 4 months ago. Today, I'm excited to share an important milestone for:
Pretraining at scale without a pretraining team.
Mirendil's mission is to solve Recursive Self-Improvement (RSI) and use it to democratize frontier AI development, ultimately accelerating science and technology. At Mirendil, when we think about RSI, we think holistically about the company as an intelligent entity that self-improves toward its goal. From that perspective, the most powerful notion of acceleration is reducing the time and resources (number of human experts, compute, and other costs) needed to achieve ambitious goals.
Given that pretraining is perhaps the most important yet most challenging part of frontier AI development, we decided to run an ambitious controlled experiment: Starting fully from scratch, can we significantly reduce the time, number of human experts, and compute resources needed to go all the way to successfully pretraining a strong coding model at scale?
Mirendil-Base is the result of this experiment. Our first pretrained model was built by one researcher in 3.5 months, with less than $10M in compute. This includes all the research and small-scale experiments it took to get there, as well as the final training run.
The recipe holds up as the models get bigger. For reference, Mirendil-Base is in the 100B-500B parameter range, Mirendil-Base-Large in the 500B-2T range, and Mirendil-Base-XL in the 2T-10T range.
We're beginning to see a world where a single researcher, empowered by AI, can take on work that once required an entire research organization. As these systems improve, the bottleneck shifts from the size of a team to the ambition and judgment of the people directing them. This has profound implications for who gets to build AI and, ultimately, who gets to push the frontiers of science and technology forward.
We’ll share independent evaluations, post-training benchmarks and API-access sooner than you think so stay tuned!
Cool, local and open! Can’t wait to test it. I am looking fw to the paradigm of local-to-remote handoff and vice versa. Not all tasks are equal, neither in terms of difficulty, nor in terms of deadline. Adaptive by design.
Introducing Flint, @Recursive_SI’s first model: a fast local coding model built with our automated research system, the Eureka Machine.
• Best-in-class fast local coding
• Up to 2x faster than comparable models: 125+ tokens/sec on a MacBook Pro, 150+ on RTX Spark
• Built from scratch in under 4 months
Private, unmetered intelligence on your own machine. Open weights available now on @huggingface, run it locally on your favorite framework or try it for free on @opencode or @pidotdev.
I read so many tweets today on Revolut’s outage and its repercussions. I get it, but having a single point of failure in your personal finances may not be a wise thing. Obviously no blame on users here, but anything can fail … Please diversify to mitigate such risks.
Model worms will not even need to break in and self-replicate. People are already giving them direct access. “While you’re in there …“ 🤠
Or as I would like to call ... Moving "tick" to "tock".
OpenAI, Anthropic and other leading AI labs face a growing challenge: Better models aren't necessarily attracting more usage. A 10-point intelligence advantage now earns just 20% more developer traffic, down from nearly 69% in late 2025.
Full story: thein.fo/4hGJ9jY
The X rec algo doesn't feel elegant. It adapts too quickly to OOD interests/searches/interactions, instead of gradually and implicitly. Ends up causing a negative, abrupt experience. At least, you can interpret why you're shown certain things most times.
Remember when they were throwing bricks on cybertruck windows …? This is demoing at another level. Not sure how useful, since we are mostly driving in 2D. But sure, why not?
This was unexpected, but super fascinating for local AI. At a very different scale than I expected. Barely consumer oriented, but will get there … Very strong benchmarks too. 🤘
Muse Spark Open -- weights releasing today -- runs great on new Mac Studios thanks to work with @johnternus and Apple.
We keep iterating on peer review being flooded with GenAI content. What about other establishments, like police, courts, customer service, hiring ... Any human-in-the-loop system risks DoS. Human effort + incentives have been the rate limiters. Does the treatment involve more AI?
I am not sure about such posts’ intention. To raise concern, to decelerate pace or to drop the ball to infrastructure/user? Who bares the responsibility? One should not solve intelligence in silo, but also think and prevent some of the downstream consequences in this neo world.
“Our experiments suggest that small AI models may be months away from being capable enough to self-replicate on commercial IoT devices, like smart ovens and off-brand Android TV boxes. A replicator like this would be challenging to shut down, since there’s no centralized point to shut off - though it may be slow (with a doubling time of days). Unlike non-AI malware, it could also figure out unfamiliar devices on its own. In our evals, a 9B model got inference running on 8 of the 9 hardware types we tested.
Models of 30B parameters and under can already complete many of the individual steps for replication today, though they make mistakes when composing them end to end. In our simplified evals, GLM-5.2 and stronger models are capable of end to end replication. If small models improve at their historical rate, they’ll match this capability level by the end of this year. Small weight-based replicators may be possible soon after.” -@bucketofkets @willzhou97
anthropic.com/research/self-…
What if there was an equivalent of Folding@Home or SETI@Home, where one can contribute tokens for scientific research? 🤔
Wild thought: We should not have SI being Super Intelligence and Self-Improvement. I prefer the latter.
The thing is that:
1. I see a person who made an effort, however robust or sloppy, to publish sth
2. You shouldn't need publications to start a PhD
3. The prof could see it as an opportunity for teaching this prospective student how to do research w/out full LLM cognitive offload
My PhD advisor received an email from a student who's been doing research for 8 months, and has *16 solo papers published* in 2026, almost all at real AI conferences, 1 paper at ICML.
All papers 100% flagged by Pangram (not reliable but a tell). The truth is academic AI research does not survive this. It has to reinvent itself fast.
