Andrej Karpathy:
“A 1B model can be much more effective than most current LLMs.
90% of current models waste a ton of capacity and money on remembering stuff that doesn’t matter.”
here is his 5 core thesis about Small LLM’s:
1. the intelligence was never the knowledge. Strip the memorized internet, keep the reasoning. What’s left is the cognitive core, and it’s small.
2. the core might fit in ~1B params. Maybe less. “Plenty of room at the bottom.” We’ve been paying for a library when we needed a brain.
3. size stopped being the bottleneck. What makes an agent work is the tools it can call and how tight the feedback loop is, not parameter count.
4. intelligence lives in the system, not the weights. A raw LLM is a CPU with no OS. Small core + good harness > giant model alone.
5. you can own it. A model on your hardware is one no one can rate-limit, price-hike, or deprecate on you.
in 12 months the gap between teams shipping on-device and teams paying per-token frontier bills won’t be capability.
TwIL an 8.8B model at 90%+ of frontier isn’t a benchmark flex.
It’s the thesis Karpathy’s been repeating all year, made real.
study more on this below ↓
Modest intelligence is arriving faster than most people realize.
Today, we're open-sourcing TwIL-8B.
Just 8.8 billion parameters, yet approaching frontier-model scores on several difficult benchmarks:
• 93% of frontier performance on coding
• 92% on instruction following
• 90% on contest mathematics
All from a model small enough to run on your own hardware.
But the benchmarks aren't the most important part of this announcement.
For years, the AI industry has focused on building the most capable model, regardless of the cost required to run it.
I believe that equation is changing.
The next era of AI will be defined by the cost of completing work. Not simply the intelligence of the model, but how efficiently that intelligence can accomplish the task.
If a model with 8 billion parameters can reliably complete the same work as one with hundreds of billions, why pay for the larger model?
As smaller models become more capable, intelligence becomes something we can distribute, specialize, and run across the hardware we already own.
That's the future we're building at @thewebAI
TwIL-8B is available today. Open weights.
And this is just the first of several announcements we're making today.
HF Link: huggingface.co/webAI-Officia…
Lorenzo Farnararo retweeted
you pull out your phone to do one thing and end up doing something completely different.
agents were supposed to fix that - now we juggle chats instead of apps.
voice was supposed to fix that - now we walk around holding the phone up like a pizza slice.
so we built guv, a walkie-talkie for your agents. hold the button, think out loud, it gets done.
works with grok bot today. other personal agents and agents on your machine coming soon.
have an apple watch? reply with the first thing you'd tell your agent for access.
We want our attention back. Launching guv.sh - a walkie-talkie to your agents.
Hold the button, think out loud, it gets done.
Reply with the first thing you'd say into it, we'll send you a testflight link.
Lorenzo Farnararo retweeted
Boycott Haier! I bought two appliances because of the @home_assistant integration. They truly have shot themselves in the foot with their heavy handed approach. Haier's most loyal customers add value to @HaierAppliances at zero cost to @HaierEurope. 300k active HA users alienated
This Post is from an account that no longer exists. Learn more
Lorenzo Farnararo retweeted
Something recently clicked. Developer Experience is a part of Software Architecture.
Lorenzo Farnararo retweeted
If you value making it work over making it right then you will make it work by doing the wrong thing. As that continues, and the code degrades, it becomes harder and harder to either make it work or make it right. – at Québec, Canada
Lorenzo Farnararo retweeted
We’ve made the decision to stop all political advertising on Twitter globally. We believe political message reach should be earned, not bought. Why? A few reasons…🧵
Lorenzo Farnararo retweeted
"The sooner you start to code, the longer the program will take." - Roy Carlson
Lorenzo Farnararo retweeted
In the last three decades we have irrevocably moved software into the very foundations of our civilization.
We programmers now bear an awe inducing responsibility. Civilization is now ours to improve or destroy.
What code of ethics ought we embrace? – at Wisconsin, USA
Lorenzo Farnararo retweeted
year 1: hoping and praying no-one asks you to do anything on the big scary command line
year 5: if I have to leave the safe haven of my terminal to use some GUI nonsense, someone's gonna pay
Lorenzo Farnararo retweeted
for halloween, I'm coming as the technical debt ticket you've shoved in the backlog for the past 6 months
























