anthlax

Yes - decorators are syntactically nicer :)

Atlas shrugged

Thanks for this comment. I’m going to spend some time looking into whitepapers on dropout - this is a new technique to me. Thanks for the in depth comment, I appreciate it a bunch.

Could you further elaborate on “intellectual engineers”? What mistakes were made? Esoteric languages? Obfuscating design patterns?

Partially so I could learn from mistakes and partially since I’m a sucker for post-mortems :)

> “Modern machine learning problems often involve fitting a huge number of parameters with a massive amount of data. GPT-4, the next iteration of the engine underlying ChatGPT, is rumored to have 1 trillion to 100 trillion parameters. No computer in existence could handle these parameters all at once. So in each step, algorithms pick a small random subset of parameters (whatever the computer can handle) and just work with those instead.”

I can’t wrap my head around this - is this an oversimplification? Does an input take some n runs and then the outputs are summed? Since each hidden layer has a connection to every neuron in the previous hidden layer, how are the results not severely warped?

(I’m having trouble articulating my non-understanding, sorry for the low quality question. Perhaps if someone has a link to a paper on this technique that could be helpful)

Yep - I can’t help but wonder if this is a work of fiction since it is absolutely incorrect

What I am not understanding is if all paths have an equal amplitude, and there are infinite paths, shouldn’t all paths cancel out? I’ve heard an example of throwing infinite stones in a pool, and the ripple that forms is what we perceive as reality. But how does our reality not have a counter-reality that cancels it out?

Yes, it is not for IAM policies

Good to hear about your family! Is this supported by macro trends?

I think this guy is a bot… comments seem unnatural.

This is a phenomenal answer, thanks for taking the time! Many more variables in the equation than I initially considered, thanks for all the points.

Haha - I really began to learn nix when I started to read the docs before bed like a novel. I’ve started to do this for more libs / frameworks and it’s going quite well - It seems to be the fast track to power user status!

This is a great take! I agree placebo is a powerful tool. Thanks for the conversation

This is an great take! I agree placebo is a powerful tool :). Thanks for the conversation

Interesting - how does one determine if they are “affected by placebo” or not?

And how did you control for placebo?

Where the hell are all these “bloat jobs”? Perhaps I’m too optimistic about other people, but I couldn’t name a single person that wasn’t pulling their weight. Perhaps people that were operating more slowly than others, but no one who was outright costing more than they provided.

At Google, for reference

Can the method of generating MacOS serial tooens be used for iMessage on android??

My current best use-case for chatGPT is to shorten emails. I’m the kind of person that will add a lot of prose to emails - I plug much of my writing into chatGPT saying “please shorten this email” and it works well. Not perfect (sometimes cuts out important info) but well! This is to respect my teammates times.

Nix to replace / supplement docker

Do you have sources for the claim that “western kitchens are built incorrectly”? Or are you referring to the same studies as in the OP? It feels intuitive but I would like to see the extent of damage. As someone looking to build a home this could be extremely relevant.

Is there a link to this discussion? This seems interesting. If I can’t make a call to a downstream service, or a file I’m trying to read doesn’t exist, or a s3 bucket 404’s, or almost any other “real world” error I can think of, the only (sane) way I can think of handling this is propagating the error down to the caller (and perhaps logging?) Do you mean to say that it is idiomatic in Go to handle errors by… doing something else?

On mobile so I can’t put in a code block, but here’s how I thought Go was written: value, err := some_fn()

If err != nil { Return err }

(? Operator works here because you could do value := some_fn()? And remove the if statement boilerplate)

Do you mean that instead of “return err” Go idiomatically does something else?

Does anyone else think this reads like a horoscope? As in, it’s quite likely that this applies to most people reading it to some degree. Not many jobs (at least in my SWE space) are directly linear and require a variety of tasks.

I was excited about this because I use docker-compose to run a service written in Rust. The final command of the container is “cargo watch -x run” and the compiles take 1-2 minutes per hot reload. Porter solved this by completely sidestepping images, but I’m still considering if this is worth having the user install cargo on their host machine (in the case of porter) and give up hermetic development. Not sure yet, but thanks for this tool!

Disagree - “Musk or any family member” vs “Musk” are fundamentally different - the first implies none of his family was present. The second could be seen as “we discovered a family member of musk was present, but musk was not”

This firmly puts you in the first category

Metrics are religiously collected and any sizable performance improvement will have a clear impact on one metric or another no?

This has got to be elon’s HN account. Insane amounts of pro Elon, anti union comments.

Correlation or causation? Perhaps if you didn’t skip classes, you’d still struggle through adult life but also struggle through childhood.

Perhaps you’d just be more numb.

What do you think about GPAs being non-standard? Test scores calibrate between schools - GPAs only kind of do.

I remember many classes that were taught by two separate teachers: one teacher consistently gave students higher grades. The other teachers students consistently did better on the AP test - much more strict with homework and test scores, but it ended up positively impacting the students.