Kunal (@therealkmans) profile picture
Joined March 2023
Photo by Kunal (@therealkmans)Photo by Kunal (@therealkmans)Photo by Kunal (@therealkmans)Photo by Kunal (@therealkmans)Photo by Kunal (@therealkmans)Photo by Kunal (@therealkmans)Photo by Kunal (@therealkmans)Photo by Kunal (@therealkmans)Photo by Kunal (@therealkmans)Photo by Kunal (@therealkmans)
Intelligence is about learning - from few examples and from interaction with the world. Current models are not intelligent. They are very capable and useful, but not intelligent in the way humans are because they need so many examples to reach human level. LLMs are not bitter lesson pilled because they still rely on human data, and lots of it. An intelligent agent- a truly intelligent agent- can just learn from the world by continual exploration. I think this is what he means.
Brute force computation is not enough for intelligence. Clever new methods are needed. The science of AI is far from finished. IMHO
4
5
42
3,029
I shipped more LeetGPU features & bug fixes with Grok @bot in 1 day than I have in the last month
2
5
266
we’re doing cool things. join us in delivering science fiction tesla.com/careers/search/job…
Replying to @teslayoda
We’re making interesting progress here. Digital Optimus can play about halfway through the Diablo campaign so far just by looking at the screen like a human. Skill at Counter-Strike and other fast games is good too. League is also in training. Goal is to generalize across all games.
11
4
1
129
5,973
so bullish on jev
1
3
340
Kunal retweeted
i often think about this…
264
8,323
227
74,034
1,485,946
Kunal retweeted
We just released a new challenge today ❗ GPT-2 (120M) Transformer Block Compose multiple kernels into a full transformer block. The first of many upcoming challenges focused on real-world inference optimization. Write your solution in CUDA, Triton, PyTorch, JAX, Mojo, or CuTe DSL and benchmark it on state-of-the-art GPUs like the H100, H200, and B200, and more.
7
51
1
570
30,987
Coporate america hates people who eat lunch after 1:30
513
We now have claude code on a daily cron making PRs for new challenges on @LeetGPU 👀
1
423
Super excited to launch these features! Let us know what you think
We just launched a couple of new features on LeetGPU - PyTorch Profiler Traces for every submission - AI Chat to help explain, debug, and optimize your code Go try them out!
2
787
Wise words from claude
1
197
How it feels trying to keep up with feature requests for @LeetGPU
Oh, you're writing CUDA kernels? Everyone's on Triton now. Just kidding, we're all on Mojo. We're using cuTile. We're using ROCm. We have an in-house DSL compiler targeting the NVGPU MLIR dialect but wait, Tile IR just dropped so we're going to target that instead. Our PM is on TileLang. The team lead was on CuTe but now she's back to handwriting PTX. If you're not on Pallas, you're ngmi. Our intern is building on TT-Metalium for our Wormholes. Our CFO approved an order for some big chungus wafer-scale chips so now we're porting our kernels to CSL. Our CTO is working on a kernel-less graph compiler so we won't need to write kernels anymore. Our CEO thinks we're talking about the Linux kernel. We're building Claude for dogs.
5
385
We just shipped JAX support on all challenges 🚀 Try it out!
1
8
635
Subway can be so good but its way too overpriced
2
152
How you code is how you think
1
3
221
Chess player to tech bro pipeline is wide
I noticed that the tech community is really friendly and welcoming. No judgement, just pure encouragement. Couldn't reply to all msgs, but thanks to everyone who reached out and extended help on various posts..
2
377
this is how you know you wrote good code
Fun fact: I just write branchless code because I'm afraid of indentations. By accident it's also fast.
1
323
Doing my monthly re-read
3
135
Replying to @mysticwillz
API gets 10x traffic overnight with no page? I'm checking revenue! 😆
3
1
108
10,852
This is so sick
Moving on to Matrix Multiplication and branching within the kernel. Discovered @LeetGPU courtesy @sadernoheart They have a nifty CLI that can invoke a GPU directly from the terminal. 4/n
2
3
476