Most of their job is about people, leverage, and execution. It's not a sit and think job, it's mostly 1:1s and coalition-building
Theyre also utilizing LLMs to augment themselves. And are getting more data, from off-record convos to reports from trusted smart colleagues who used an LLM to refine their intel
yes, the quote is what I'm referring to, directing the AI is part of it, people use these to quickly brainstorm and refine ideas. I'd be more charitable and wouldn't hastily assume it was some skill issue, especially them being a principal engineer
they said 90% of it was spent on ideation and exploration
they didnt specifically mean they built a wordle clone, just a game like it. if they wanted just a wordle clone, they wouldve gotten one within a few minutes of using codegen tools.
Is there any source for first-hand specifics of what she does?
I used to argue in reddit (same username as my HN) basically calling Musk a fraud and Gwynne Shotwell being all the brains 6 years ago, but I've since changed my position after seeing engineers in spaceX give props to Musk at podcasts, twitter, and various interviews.
CP just teaches some familiarity with DSA/algorithms, and there's much more to perf than DSA. Even assessing algorithmic performance requires real benchmarks and profiling, while the complexity analysis people do in CP disregards other factors like hardware, architecture, format, and other abstractions. Squeezing perf via DSA is much more easier/straightforward, people don't need to grind CP to learn that
I guess lots of people find
him interesting (see HN guidelines)
Charlie (and Buffett) are often recommended here for their mental models as their approach to investing and finance are very transferrable in startups/engineering. At the end of the day, large scale software engineering is mostly about managing risk, strategy, and corporate finance/value. Often, Poor Charlie's Almanack is recommended by my top Staff+ colleagues.
I wouldnt say so. I think it's the same HN. Although when it gets political, I've noticed that the discourse does degrade a bit faster than it used to be for such topics, for there's a bigger pool of users who'll get provoked by the opinion
A book that has a similar style to this one (teaching algorithm design) and my personal favourite is Algorithmic Problem Solving by Roland Backhouse.
It just teaches algorithmic problem solving via math, but unlike mathy algorithmic books like CLRS, you only need algebra for this one.
I like how the book, IIRC, starts with invariants, providing the reader the foundational skill needed to notice underlying patterns and to decompose problems.
I think it's a close book. The book details the decision process from both sides. Japan's plans to continue the war. The way the high command coped with the first atomic bomb, etc. It does not sway you into thinking any side was justified, but it expands one's perspective to help you understand the other's side.
I was born in Southeast Asia, lived in China for a while, everyone I've met views it as some kind of karmic justice. It's the idea you get from school (we were living peacefully, then we got attacked and they did all these brutal crimes) and from our grandparents.
I've only seen the contrarian opinion coming from westerners. Perhaps because there's barely any media from Hollywood about the pacific theater, and if there was (eg. HBO's The Pacific), it does not show the horrors of it.
If one is more interested in an objective take on history, especially those using the argument that the Japanese were already "defeated" before the bombs came, then the book Downfall by Richard B Frank is a good book to start with for those who hold a firm anti-nuke stance.
Making them compulsory would be good, but replacing the "intro" courses with a paradigm-survey course would probably be too much for students to handle.
Students would also likely not be technically mature enough to be able to appreciate these early in their journey. I certainly didnt when I worked through SICP in my late freshman/early sophomore year, eventually I ended up forgetting everything about it.
I like the PLAI lectures as the prof (Shriram) was able to relate the concepts to those in the industry and various languages like haskell, as well as made it practical by discussing program design -- perhaps HtDP was also a prerequisite?
I'm not sure about Elon's proficiency with software, but based on his comments on twitter I'd lean on him being out of touch there. He does seem to be knowledgeable on the AI/DL side though, see Andrej Karpathy's comment:
This post quotes top SpaceX engineers like Tom Mueller, Kevin Watson, Garrett Reisman, Josh Boehm, and others like John Carmack about Musk's involvement in engineering:
A series of distributed systems courses from UIUC at coursera where you learn about systems design, distributed algorithms, and has you build projects like a distributed DB, implementing algorithms like gossip protocols, etc. in C++:
Most of their job is about people, leverage, and execution. It's not a sit and think job, it's mostly 1:1s and coalition-building
Theyre also utilizing LLMs to augment themselves. And are getting more data, from off-record convos to reports from trusted smart colleagues who used an LLM to refine their intel