kburman

5 ms·

31.
by kburman 7mo ago
Amazing work! I'd love to hear more details about your workflow with Claude Code. As a side note and this isn't a knock on your project specifically. I think the community needs to normalize disclaimers for "vibe-coded"
32.
by kburman 8mo ago
Honestly, with recent models, these types of tasks are very much possible. Now it mostly depends on whether you are using the model correctly or not.
33.
by kburman 8mo ago
Since Opus 4.5, things have changed quite a lot. I find LLMs very useful for discussing new features or ideas, and Sonnet is great for executing your plan while you grab a coffee.
34.
Let's Build Recursive Language Models (Theory and Code) [video](youtube.com)
2 points by kburman 8mo ago|0 comments
35.
by kburman 8mo ago
I don't get the whole idea of treating identity verification as a private enterprise problem. I realize it's easy to just blame LinkedIn or Microsoft here, but the core issue is architectural. We are trying to solve a public utili
36.
by kburman 8mo ago
Thanks for answering! Regarding cardinality, I was actually thinking more about high-cardinality dimensions on the NLU side, e.g., if a user asks for "Sales for [Obscure Company Name]," and you have 10M distinct customers. Does th
37.
by kburman 8mo ago
Thanks for sharing the links, the architectural overview is very insightful. I'm curious how this approach manages cardinality explosion? Also, how do you handle cases where a user asks for data that requires running multiple queries,
38.
by kburman 9mo ago
If a foreign entity came into Florida and bought up 35% of the entire retail infrastructure, you bet the US government would regulate it and demand local value capture. Case in point - US actively forced TSMC and Samsung to build $65B+ of f
39.
by kburman 9mo ago
You are describing a colonial model, extract all the wealth while investing nothing in the local economy. That era is over. If anything, Meta is the anomaly, not the role model. They should be required to invest more given their dominance,
40.
by kburman 9mo ago
I get the fear, but look at it from the investor's perspective. The US market is tapped out, Amazon is already everywhere it can be. Amazon isn't expanding in India out of love for the country or a desire to see it grow. They are
41.
by kburman 9mo ago
I realize it’s easy to pattern-match this news to 'hiring in India vs. firing in US' given the current climate, but having worked at Amazon India for 4 years, I can tell you the cuts happen there too. Amazon has a history of annua
42.
Build Your Own AI Coding Agent (Full Guide) [video](youtube.com)
2 points by kburman 9mo ago|0 comments
43.
by kburman 9mo ago
This is a recipe for model collapse/poisoning.
44.
by kburman 9mo ago
An LLM is optimized for its training data, not for newly built formats or abstractions. I don’t understand why we keep building so-called "LLM-optimized" X or Y. It’s the same story we’ve seen before with TOON.
45.
by kburman 9mo ago
It’s the standard enshittification lifecycle: subsidize usage to get adoption, then lock down the API to force users into a controlled environment where you can squeeze them. Like Reddit, they realized they can't show ads (or control t
46.
by kburman 9mo ago
You can be a controversial figure politically and still build a generation defining product. The market rewards utility, not ideological purity. The headline frames this as a paradox, as if these two things are incompatible. But they aren&#
47.
by kburman 9mo ago
Fair point on the strict terminology, I was using 'wire protocol' broadly to mean the communication standard vs. the implementation. A more precise analogy is likely LSP (Language Server Protocol). MCP is to AI agents what LSP is
48.
by kburman 9mo ago
This analysis dismisses MCP by focusing too narrowly on local file system interactions. The real value isn't just running scripts; it's interoperability. MCP allows any client (Claude, Cursor, IDEs) to dynamically discover and int
49.
by kburman 9mo ago
I'm curious about the economics of canceling a deal like this. Since the editor spent significant time reviewing the drafts, did the contract require you to reimburse the publisher for those costs or return the advance?
50.
by kburman 9mo ago
> a state-of-the-art research tool over Hacker News, arXiv, LessWrong, and dozens what makes this state of the art?
51.
The mongo db situation is crazy [video](youtube.com)
3 points by kburman 10mo ago|0 comments
52.
by kburman 10mo ago
My rule for modern TVs: 1. Never connect the TV panel itself to the internet. Keep it air-gapped. Treat it solely as a dumb monitor. 2. Use an Apple TV for the "smart" features. 3. Avoid Fire TV, Chromecast, or Roku. The logic is
53.
by kburman 10mo ago
My rule for modern TVs: 1. Never connect the TV panel itself to the internet. Keep it air-gapped. Treat it solely as a dumb monitor. 2. Use an Apple TV for the "smart" features. 3. Avoid Fire TV, Chromecast, or Roku. The logic is
54.
by kburman 10mo ago
Expecting the storage layer to fix an inefficient I/O pattern (millions of tiny network requests) is optimizing the wrong part of the stack. > are you working with a virtual tape drive or a virtual SSD. Treating a networked object s
55.
by kburman 10mo ago
- Modern DL frameworks (PyTorch DataLoader, WebDataset, NVIDIA DALI) do not require random access to disk. They stream large sequential shards into a RAM buffer and shuffle within that buffer. As long as the buffer size is significantly lar
56.
by kburman 10mo ago
I feel like this product is optimizing for an anti-pattern. The blog argues that AI workloads are bottlenecked by latency because of 'millions of small files.' But if you are training on millions of loose 4KB objects directly from
57.
by kburman 10mo ago
Anyone interested can also follow these amazing playlists: 1. Building LLMs from scratch - https://www.youtube.com/playlist?list=PLPTV0NXA_ZSgsLAr8YCgC... 2. Reasoning LLMs from Scratch - https://www.youtube.com&
58.
by kburman 10mo ago
Have a question related to this, if a photon has zero proper time between emission and absorption, how should I think about the influence of later-created photons or fields on it? In our frame, we can interact with a photon long after it&#x
59.
by kburman 10mo ago
This looks impressive. Could someone familiar with Postgres internals explain the hidden trade-offs of this approach? I understand the obvious limitations of it being embedded/single-host, but I'm curious about the engine itself.
60.
by kburman 10mo ago
This is absolutely brilliant. If you integrate Reddit and Twitter/X, you’d get a much more complete picture of how stories spread across the internet.

More ›