
Cooking @byro_ai @agentsimdev
San Francisco, CA
Joined December 2020
- Tweets687
- Following633
- Followers95
- Likes4.3K
Fathin retweeted
Where does AI intuition come from?
And could machines ever get it?
Taste is the hard part of AI.
It's what we're working on for content distribution at @byro_ai
Fathin retweeted
My GPUs have been cooking for 48 hours straight.
I've been testing whether DNA and protein AI models can find rare-disease mutations they've never seen.
190k ClinVar variants, scored only on ones labelled after each model was trained. Some open models beat the tools clinics use today.
Most results were what I expected. One wasn't, and it changes what I'm building.
Full breakdown Monday.
@NVIDIAHealth @GoogleDeepMind
Fathin retweeted
brev i slept for 2 hours today. 2027 is my YEAR.
I mean 2024, 2025 and 2026 were mine too.
just saying..
Fathin retweeted
shipped a claude code plugin for @byro_ai that refuses to write your linkedin post if your workspace has no confirmed facts.
no facts, no post. that's the feature.
every AI ghostwriter will happily invent your "journey." mine tells you to go confirm something real first
Fathin retweeted
recursive superintelligence used to be somewhat far fetched 2 yrs ago, now it's a risk factor in an IPO filing.
when AI helps do AI research, the time-to-intelligence for the next model shows up faster and smarter, and that one speeds up the one after it. every turn of the loop gets shorter recursively
anthropic saying in an S-1 that AI is already speeding up its own R&D means the first turn is lowkey running.
the next few years abt to get very interesting.
Anthropic has confirmed in an updated S-1 today that it has determined that the Autonomy 2 capability threshold has been met, based on the “dramatic acceleration of overall AI R&D progress” condition, given that it applies where Anthropic "observes or expects" accelerated progress. Anthropic further states that it does not make this determination based on current models.
sec.gov/Archives/edgar/data/…
Fathin retweeted
Anthropic has confirmed in an updated S-1 today that it has determined that the Autonomy 2 capability threshold has been met, based on the “dramatic acceleration of overall AI R&D progress” condition, given that it applies where Anthropic "observes or expects" accelerated progress. Anthropic further states that it does not make this determination based on current models.
sec.gov/Archives/edgar/data/…
Anthropic updated its S1 this morning. Notable additions re automated AI R&D and human oversight of it:
"[...] our internal estimates and our expectations regarding future models, rather than our current models, indicate that we may imminently meet our Responsible Scaling Policy’s definition of 'dramatically accelerated progress attributable to AI.' We have not designated any of our current models as exceeding this threshold, and we continue to assess the underlying data. We are also evaluating the implications of meeting the AI research and development capability threshold and will release further reports in the future as we set forth in our Responsible Scaling Policy."
"We are also considering additional measures, including updates to our Responsible Scaling Policy, to require human review and approval of certain decisions in our research and development processes as our models perform an increasing share of that work."
"We believe our investments in safety techniques will enable us to manage these risks and responsibly realize substantially more leverage from Claude than we do today in developing AI models. However, we do not anticipate being able to safely use Claude to fully autonomously design and build successive generations of frontier models without meaningful human supervision or understanding, if and when that becomes technically possible. Accordingly, we are considering updates to our Responsible Scaling Policy to keep humans in control of frontier model development at Anthropic."
sec.gov/Archives/edgar/data/…
Fathin retweeted
Worth your time: how one of the most advanced agent memory systems works.
Memory is one of the biggest levers in your agent harness. I don't think a one-size-fits-all system exists yet.
You maintain it like code. Upgrade it at the right moments, in the right places.
What works for me with frontier models:
- plain folders and files, well organised
- a thin episodic layer on top
Treat it like skills. Keep it basic, or you get in the model's way.
Models will keep improving and memory will keep changing. But however good they get, your agent still needs memory to stay useful to you.
Memory is what allows Claude to learn from experience. In a new Engineering Blog post, we share how “dreaming” helps Claude organize what it’s learned, correct its mistakes, and find what it needs more easily the next time. anthropic.com/research/compo…
Fathin retweeted
Claude deleted 50K lines from our codebase. Tests still green.
The prompt:
/goal Draw a wiring diagram, every feature to every file it touches. Mark the worst knots in red, numbered. For each one: file, line, how to untangle it.
🦍 Introducing Memorilla: memory as a modality for LLMs
Memorilla turns everything a frozen LLM needs to remember into just 16 memory embeddings
⚡ It uses up to 4,212× fewer tokens and is 27× faster than full context inference while more accurate
🎯 Pushes the Pareto frontier over RAG and agentic memory in personalization, long-document and knowledge QA
A thread 🧵:
Fathin retweeted
my spidey senses are telling me the labs are, as we speak, using unguardrailed versions of their unreleased models to systematically reverse-engineer ALL of the internet’s software, apps, and games, then feeding the reconstructed source code straight into their training pipelines while simultaneously building guardrails to ladder-pull the rest of us



















