1,866 karma · joined June 10, 2009
https://benson-guides.blogspot.com/
fung [dot] benson [at] gmail [dot] com
The trade-off ends up being that metaprogramming becomes harder while syntax gets changed
shrug, do it long enough and the parans disappear:
Q: How can you tell when you've reached Lisp Enlightenment?
A: The parentheses disappear.
~ AnonymousWith Remote, many things have to be put in writing, organized, and referenced back to.
Way less room for misinterpretation than the occasional drive-by chat.
This pattern just happens over and over in VC/startup land. See IoT and cryptocurrency/NFT companies as previous example in previous tech hype cycles.
The kid at least asks questions!
Also, CZ (pleads guilty to money laundering charges): https://news.ycombinator.com/item?id=38366729
Only the basics that are close to everyday programming work: write a for-loop, know what a Map/dict is, and Google for “how to read a file”. If a candidate can’t do that, they can’t really program. ChatGPT can probably code this answer.
Just commit and don’t change plans.
Tell everything else this time is already booked and paid for.
https://youtu.be/2LyzBoHo5EI?si=ezHx2TfJa4-llCzX
He was popular at TED and built gapminder data viz tool when web2 tech started becoming popular.
(the article author got a better deal at the end anyways, imo)
For example, “speed work running”, were you looking to improve sprinting? (I’m not a runner, so it almost reads to me like “working fast while on a treadmill”). Reading the rest of your example, maybe “sprint workouts” would give better results?
(40+yr old guy here, Google had always worked for me, as many people had described it working for 20yr olds. Finding the right, specific enough set of keywords had always been my “trick”. Little did I realize I subconsciously generate embeddings like a LLM to Google search)
The longest lasting and easiest to extend software I’ve cobbled together (and some still running in production after 20yrs) were me designing for 18+hrs straight, share general idea w/team, then furiously writing for 36hrs. Those spurts were always the most productive and setup future work to be easy to add on.
The guise of the academic paper def throws things off and it’s poorly communicated in the paper what the model benefits are.
Posting a quick write up of the talk on youtube would’ve set the right amount of rigor and expectation. (Needs marketing help, lol)
Anyone who work a lot in time series forecasting can explain this in some more details?
I’ve def used ARIMA, but only for simple things. Not sure why this would be more expensive to train and run than a Transformer model, and even if true, ARIMA is so ubiquitous that comparing resources & time would be enlightening. Otherwise it just sounds like a sales pitch and throw more obscure acronyms for a bit of “I’m the expert, abc xyz industry letters” marketing.
Should an AI created to make paper clips consider other values than making paper clips?
https://www.lesswrong.com/tag/squiggle-maximizer-formerly-pa...
Diving all the way down ad absurdum: What's the point of maximizing profit if there are no humans to utilize those profits?
The thin/thick client cycle starts again...
Do something that takes a good amount of concentration/coordination physically, so there's no way to think about work.
Ex: go to the gym, change, circuit train and focus on good form, breathing.
Make it a routine, so that the motivation barrier won't stop you from going.
I see & hear you. ;)
Def. seen jq thrown into sed/awk scripts where a readable programming language was the right move. People spend hrs finding the right syntax to these things ~ not always well spent.