HNHacker News
TopNewBestAskShowJobs

dheatov

77 karma · joined April 27, 2018

submissionscomments
dheatov··on Comma 3X: Initial Impressions
Owh that's kinda cool, Ive never thought of that. Perhaps Im constrained by my thoughts, but I think for it to be practical it requires detecting driver's intention, which seems difficult to achieve by using only vehicular sensors data.
dheatov··on Comma 3X: Initial Impressions
Sounds like a neat idea and potential fit with certain niche market. But I would be surprise if the answer is not just drive slower in most of the cases. In that case do you still want more specifics like brake timing? Optimal speed given current traffic condition, visibility, and curvature?
dheatov··on Linux in Excel
I support you to one up them with `Lightning adapter on Excel`
dheatov··on As 'Bot' Students Continue to Flood In, Community Colleges Struggle to Respond
Seems like the fake students are automation scripts written to mimic human interaction in online class, pretty much the common definition of "bot"? For now, all bot is still in someway human-operated anyway.
dheatov··on Tencent's 'Hunyuan-T1'–The First Mamba-Powered Ultra-Large Model
I don't think your analogy works for the tailwind situation, and there is no whole idea to give up on anyway. People will still be researching this hyper-complicated matrix multiplication thing, i.e. LLM, for a very long time.

Personally, the tailwind example is an argument against one specific use case: LLM-assisted/driven coding, which I also believe is the best shot of LLM being actually productive in a non-academic setting.

If I have a super-nice RL-ed (or even RLHF-ed) coding model & weights that's working for me (in whatever sense the word "working" means), and changing some function names will actually f* it up badly, then it is very not good. I hope I will never ever have to work with "programmer" that is super-reluctant to reorganize the code just to protect their pet LLM.

dheatov··on Windows 11 Insider Preview Build
And yet it still can't reliably stay up for more than a few days without being forcefully rebooted.
dheatov··on Does current AI represent a dead end?
I have never personally met any malicious actor that knowingly dump unverified shit straight from GPT. However, I have met people IRL who gave way too much authority to those quantized model weights, got genuinely confused when the generated text doesn't agree with human written technical information.

To them, chatgpt IS the verification.

I am not optimistic about the future. But also perhaps some amazing people will deal with the error for the rest of us, like how most people don't go and worry about floating point error, and I'm just not smart enough to see how it looks like.

dheatov··on NetBSD bans use of Copilot-generated code
Anecdotally my experience with generative AI is that the valid output space is very tiny especially under a restrictive prompt, i.e. homework question, even smaller when the output also has to be a correct answer. Using a riskier decoding scheme only generates output that is paraphrastic to each other. So lexical similarity should catch most of the naive prompter.

Subjectively, if a student can knowledge-guide an LLM to generate response from a specific perspective, I think it is very debatable whether such usecase should be outright punished.

dheatov··on Show HN: I'm 17 and wrote a guide on how to build your own programming language
For some parts of the landscape like mobile app development, yes. Its hard to say if they are doing it with hidden intention, or if they sincerely believe enforcing their "best practices" on everyone yields greater good. Probably a bit of both.

The part where you build a demo/prototype showcasing if an idea works, I believe it should be much easier than before as long as you postpone understanding/deciphering the implementation. In those major ecosystems there are huge collections of single-purpose libraries doing a subset of what you want to do, all you need might just be a reasonable amount of python to string them together.

However, ensuring the correctness and actually deploying them for planet-scale mandates following a reasonable subset of "best practices" which is very overwhelming, but doable. People doesn't usually build planet-scale product 10-20 years ago.

dheatov··on Google contacted me directly to help restore FairEmail on the Play Store
Me too. But it will be 397 pages long and most of us still can't stop using it because we live in a society.
dheatov··on Show HN: Morning Cup of Coding – A curated newsletter of programming articles
So agree with your design! It's simple, not boring and retro-like. Amazed!
← PreviousPage 2 of 2