Hacker News: master_yoda_1

New comment by master_yoda_1 in "Study: ChatGPT outperforms physicians in quality, empathetic answers to patients"

master_yoda_1 — Sat, 29 Apr 2023 19:19:13 +0000

WTF my tax money is spent on this bull $hit

New comment by master_yoda_1 in "A GPT in 60 Lines of NumPy"

master_yoda_1 — Thu, 09 Feb 2023 17:21:08 +0000

the article DOES NOT describe GPT3 and this article title is misleading, and the author is lying.

New comment by master_yoda_1 in "A GPT in 60 Lines of NumPy"

master_yoda_1 — Thu, 09 Feb 2023 17:19:38 +0000

also for curious mind here is a more authentic tutorial on building gpt (precursor to gpt3) by andrej karpathy https://www.youtube.com/watch?v=kCc8FmEb1nY My point is if you want to spend time, spend on authentic material not some bogus material.

New comment by master_yoda_1 in "A GPT in 60 Lines of NumPy"

master_yoda_1 — Thu, 09 Feb 2023 17:16:11 +0000

I would better use pytorch api than this bogus code. https://pytorch.org/docs/stable/generated/torch.nn.Transform...

New comment by master_yoda_1 in "I don't read web articles anymore, but I read books"

master_yoda_1 — Mon, 09 Jan 2023 18:21:32 +0000

its depend on what domain you are interested in. for some domain progress is so fast books are lagging 10 years behind so you need to read article (also its tough to find good article).

New comment by master_yoda_1 in "Ask HN: What are the foundational texts for learning about AI/ML/NN?"

master_yoda_1 — Mon, 09 Jan 2023 18:16:29 +0000

This book is all you need https://probml.github.io/pml-book/book1.html

New comment by master_yoda_1 in "Ask HN: Do you recall any book or course that made a topic finally click?"

master_yoda_1 — Mon, 14 Nov 2022 18:22:16 +0000

For me re-reading some book twice/thrice or reading multiple book/courses on same topic make it click. Sometime I think it click but my understanding is totally wrong, talking to people and solving problem help me find the gap.

New comment by master_yoda_1 in "Ask HN: Boring but important tech no one is working on?"

master_yoda_1 — Sat, 27 Aug 2022 01:03:33 +0000

Totally agree, learning geometry or linear algebra through AR/VR would be so easy and intuitive

New comment by master_yoda_1 in "Bayes Rules – An Introduction to Applied Bayesian Modeling (2021)"

master_yoda_1 — Sun, 17 Jul 2022 18:09:43 +0000

I always get confused who the audience is for these kind of books?

For advanced reader this is really shallow, for beginners its really advanced (who really want to read 543 page book as beginner).

New comment by master_yoda_1 in "The case against CS master’s degrees"

master_yoda_1 — Fri, 15 Jul 2022 22:54:00 +0000

I fell for this bullshit before i read the last line "This is what we have designed the Bradfield Computer Science Intensive to be."

New comment by master_yoda_1 in "One of the ML mistakes that will hurt most: not collecting the right features"

master_yoda_1 — Wed, 22 Jun 2022 02:40:56 +0000

I think it’s too much simplifying ML. If features are so easy to find yes you can write a rule based system and it works. But most of the real life problem is not easy.

New comment by master_yoda_1 in "How to build a working AI only using synthetic data in just 5 minutes"

master_yoda_1 — Mon, 13 Jun 2022 17:41:10 +0000

First define what is a "working AI" even mean Be careful AI is not an iPhone app you will not find any funding for BS in this env.

New comment by master_yoda_1 in "Ask HN: What to do about ‘Good at programming Bad at Leetcode’"

master_yoda_1 — Fri, 20 May 2022 21:33:03 +0000

I don't understand OP, so you have problem with leetcode, and you also have a problem with hire and fire culture (most of these companies don't ask leetcode) so what do you want free money? And how you decided just after meeting "people who are really good programmers"?

New comment by master_yoda_1 in "The best engineering interview question I've ever gotten"

master_yoda_1 — Mon, 18 Apr 2022 02:18:14 +0000

What frustrate me about these interviews is one need to talk on phone and code at the same time. Who the fuck do that in real life.I am so fucked up with this thinking out loud shit.

New comment by master_yoda_1 in "Microservices: Why Are We Doing This?"

master_yoda_1 — Mon, 21 Mar 2022 21:46:16 +0000

I somewhat disagree with this article. Microservice is great for complex software modules. But some jokers in industry starts writing miscroservice for every small function/method in the code. It's better that microservice should be designed by experts.

New comment by master_yoda_1 in "A non-standard book list for software developers"

master_yoda_1 — Sun, 13 Mar 2022 00:48:17 +0000

I don’t understand why old programmers are so much fascinated by lisp. If it would have been good it would have survived today.

New comment by master_yoda_1 in "Machine learning is still too hard for software engineers"

master_yoda_1 — Tue, 22 Feb 2022 22:31:11 +0000

OP first define which software engineers you are talking about. Otherwise stop bullshitting just to market your company?

New comment by master_yoda_1 in "Deep Learning Interviews book: Hundreds of fully solved job interview questions"

master_yoda_1 — Mon, 10 Jan 2022 20:44:10 +0000

just a clarification I think you are confused between RL and robotics. RL algorithm could be used anywhere either in ads, nlp, computer vision etc.

New comment by master_yoda_1 in "Deep Learning Interviews book: Hundreds of fully solved job interview questions"

master_yoda_1 — Mon, 10 Jan 2022 18:33:35 +0000

I know one good reference.

https://www.deeplearningbook.org/

Also there are various courses and lectures but that needs time and effort. There is no short cuts like the book posted by OP.

New comment by master_yoda_1 in "Deep Learning Interviews book: Hundreds of fully solved job interview questions"

master_yoda_1 — Mon, 10 Jan 2022 18:21:23 +0000

My problem with these line of numerous shallow books and courses are

  1) Written by people who has no experience in industry or they are not working on "real" machine learning jobs

  2) They think the standard in industry is pretty low and any BS works. For example the concept of "lagrange multiplier" is missing from the book. One need this concept to understand training convergence guarantee.