HN user

joelgrus

1,311 karma

http://twitter.com/joelgrus http://joelgrus.com

Posts33
Comments118
View on HN
thisnumberdoesnotexist.com 5y ago

This Number Does Not Exist

joelgrus
2pts0
leanpub.com 6y ago

Ten Essays on Fizz Buzz

joelgrus
2pts0
blog.opterview.com 7y ago

Five Bash Tricks You Might Not Know About

joelgrus
3pts0
gpt2.apps.allenai.org 7y ago

Show HN: Interactively generate text using the public OpenAI GPT-2 model

joelgrus
9pts5
www.youtube.com 8y ago

Livecoding Madness – Let's Build a Deep Learning Library [video]

joelgrus
2pts0
joelgrus.com 10y ago

Fizz Buzz in Tensorflow

joelgrus
378pts84
joelgrus.com 10y ago

A Simple Spot It Clone with PureScript and Pux

joelgrus
1pts0
joelgrus.com 10y ago

Trump Tweets on a Globe (aka Fun with D3, Socket.io, and the Twitter API)

joelgrus
2pts0
joelgrus.com 10y ago

Building a Stupid Data Product, Part 3: The Single-Page App (PureScript)

joelgrus
5pts0
joelgrus.com 10y ago

Building a Stupid Data Product, Part 2: The Web Service (Haskell)

joelgrus
3pts0
joelgrus.com 10y ago

Building a Stupid Data Product, Part 1: The Data (Python)

joelgrus
2pts0
joelgrus.com 10y ago

Polyglot Twitter Bot, Part 4: PureScript

joelgrus
4pts1
blogs.artvoice.com 10y ago

Reverse Engineering Ahmed Mohamed’s Clock… and Ourselves

joelgrus
6pts0
joelgrus.com 11y ago

Haskell-Style Fibonacci in Python

joelgrus
1pts0
joelgrus.com 11y ago

On the Mathematics of Spot It

joelgrus
9pts0
www.wsj.com 11y ago

Obama Urges Lawmakers to Pass NSA Surveillance Bill

joelgrus
14pts3
haskellbook.com 11y ago

Haskell Programming

joelgrus
70pts22
joelgrus.com 11y ago

Data Science from Scratch: First Principles with Python

joelgrus
237pts41
www.yahoo.com 11y ago

Why Airless Tires Might Be the Future of Driving

joelgrus
2pts0
joelgrus.com 12y ago

Why Programming Language X Is Unambiguously Better than Programming Language Y

joelgrus
81pts30
www.upworthygenerator.com 12y ago

Upworthy Generator - Instantly create Upworthy-style articles

joelgrus
3pts0
www.komonews.com 12y ago

College researchers eye using dead trees as biofuels

joelgrus
1pts0
www.weblaws.org 12y ago

USPS: ZIP Codes are “proprietary business information”

joelgrus
188pts138
andrewgelman.com 12y ago

Doing Data Science: What’s it all about?

joelgrus
1pts0
www.kirotv.com 12y ago

Is it food? How energy drink companies tapped taxpayers

joelgrus
1pts0
www.onthemedia.org 12y ago

How to Eavesdrop on a Former NSA Chief on a Train

joelgrus
10pts3
qz.com 12y ago

An army of robot baristas could mean the end of Starbucks as we know it

joelgrus
12pts0
theumlaut.com 12y ago

A Vindication of Juggalo Society

joelgrus
2pts0
joelgrus.github.io 12y ago

Thinking Spreadsheet: An Opinionated Guide to Problem Solving and Data Analysis

joelgrus
2pts1
joelgrus.com 12y ago

Constructive Mathematics in F# (and Clojure)

joelgrus
2pts0

Until yesterday I was using a Lenovo Thinkpad Carbon X1 running Ubuntu 22.04. It was an awful combination (microphone didn't work, nubbin stopped working, touchpad had to be replaced, keyboard had dodgy keys, battery life was bad, Chrome wouldn't start unless you added some mystery command-line options, etc, etc, etc.)

Yesterday I got a new ASUS Zenbook running Windows 11 (and I added Ubuntu in WSL2). So far so good, it seems like I can do my coding stuff on the linux subsystem but then have desktop apps and peripherals on the windows side that just work. This is the first time I've had my personal computer be a Windows machine in like 20 years, so we'll see how it goes.

The language model is GPT-2 (small).

The underlying library is called "pytorch-pretrained-BERT" because initially it just contained an implementation of BERT, but now it contains implementations of several models so they backronym-ed it to "Big-&-Extending-Repository-of-Transformers". :)

I went through this same thought process a few years ago. I basically gave myself a crash course in CS and engineering and "reinvented" / positioned myself as a software engineer.

I left my data science job and spent a couple of years as a "pure" software engineer and now am a "research engineer" which is basically a software engineer who understands machine learning and deep learning.

By and large I'm very happy with the move (other than that the company I quit as a data scientist got acquired and I would have made more $$$ if I'd stayed).

Yeah, I made it about 5 days in this year and then quit because I wasn't having fun.

Too many of the problems had the structure:

* here's Part A * submit solution for Part A * OK, now here's Part B * Part B is just different enough that you have to do a not-fun refactoring of your Part A code (that wouldn't be necessary if you'd just been asked to also design for Part B up front)

It was too much like working with an awful PM who can't decide what they want, and it mostly just made me angry.

I think it looks like a cool series of books.

I also hate that you describe it as "free to read" but then require me to give you access to my Twitter account in order to read it. I don't know you, and I don't want to give you access to update my Twitter profile or post Tweets on my behalf. (Yes, I know, then just pay for the book.)

I also hate slide decks that do that. I literally went through this one left to right and thought "wow, that was a really high level overview". Only then I realized I had to hit down some of the time.

(And the "you idiot, you weren't using the correct navigation keys" is not a particularly satisfying answer, I navigated with left and right arrows the way I have always navigated slides for years.)

Confession: I am guilty of this. Well, sort of. I actually did write a book, and it did sell more than 3 copies, and now I put "best selling author" on my resume, on account of my book being #1 in its category (sometimes) and #1 on oreilly.com (sometimes).

So far no one has gotten bent out of shape about it (or maybe no one has looked at my resume), I assume because they're smart enough to know that a "best selling" data science book is not on the NYT list outselling Stephen King and 50 Shades of Gray and Hunger Games, it's bestselling among "Books / Computers / Data & Data Analytics" or whatever.

Having put my bias on the table, I have a hard time getting worked up about what this guy did, for the following reason:

"Whoa, you're a bestselling author? What did you write?" "It's not a book, it's a picture of my foot." "Oh. That's kind of strange. How many copies did it sell?" "Three." "..."

I mean, it's not like there's some kind of special club where "best selling author" allows you to cut in line or anything. Is there? If there is, please let me know.

I "self-host" using Pelican + S3. It's super cheap (< $1/month) and pretty easy, the only real downside is that all of the Pelican themes are really ugly, and I'm not good enough at design to make a better one.

Huh, I tried a SEA-LON ticket, best price $1363 on Kayak.

TripDelta found me a $655 flight, except out of Portland (which is like a 3 hour drive away), and a $750 flight that looks like

SEA-EWR / JFK-LHR

which is a hell of a connection. I couldn't deal with the inconvenience, but even so I'm kind of impressed.

Yeah, as someone who knows a fair amount of JS but almost nothing about React, that todomvc implementation does not particularly scream "this is the framework you want to learn".

And if you leave the company they actually cost you money, since you have to pay out of pocket to exercise them (for stock that may end up never being sellable) and you also have to pay AMT taxes based on the latest company valuation.

Do those who support Curtis appearing at the talk refrain from posting out of fear of career repercussion?

No and yes. I have been pretty vocal about it on Twitter:

https://twitter.com/joelgrus/status/606474432509394944 https://twitter.com/joelgrus/status/606673594198794240

but I've also been sitting on a 75%-written blog post that I'm hesitant to actually post.

Not because of career repercussions (I have never had an employer who really gave a shit what I write about in my spare time) but because of conference repercussions.

I am promoting a book, and to that end I am scheduled to speak at various conferences and meetups. And I worry that the mob would somehow try to get me uninvited. It's not like there's an actual policy that Yarvin violated that I can be confident I'm on the right side of.

So I am torn between my natural inclination to say exactly what I think, and the more practical takeaway of "don't write or have written anything that might anger the mob".

I personally never had the motivation to work hard in college. I always assumed this was a personal flaw. Years later, after I'd completed a large number of Coursera courses, I realized it was because college was boring.

(edited to add) I should also note that the Coursera lectures themselves were pretty boring -- the innovation is that they provided slides and lecture notes so that I didn't have to sit through the lectures.

I have no non-Coursera CS training, so I use it as the "CS training" part of my resume. (I also have relevant work experience and a math degree.)

I would be comfortable hiring someone entry-level most of whose relevant experience was from Coursera, assuming they convinced me they would be a good hire.

If you are a quick learner, IMO you should feel comfortable applying for any job you think you can get hired for, even if it's a stretch based on your experience.

Honestly, if someone's resume tells me they took (say) the Coursera Natural Language Processing course, never in a million years would it occur to me to ask to see the certificate, any more than I'd ask to see their transcript to verify that they took specific classes in college.

And vice versa, I have lots of Coursera courses on my resume, no one has ever asked to see my certificates either.