Hacker News new | ask | show | jobs
by fhunt 1923 days ago
OT: Imagine "experts" from our industry explain the mainstream our world: e.g. what's the right programming language, or database technology, why types matter, or not?

Why do we trust "experts" from other industries?

3 comments

Because software development is a skilled trade, not a science. The consortium discussed in the article is made up of actual scientists who do actual science.

"I am surrounded by bullshitters therefore all fields are equally deep in it" is not a correct view.

Then take data science or deep learning which is a science as sophisticated, same there.

We trust experts from other industries so much that we do not tolerate any other view without having a clue what is going on ourselves. The reason is simple, we have our views, they are political and we refer to "experts" to make our views scientific. Just my opinion and I bet that a huge number will disagree again with this comment. But—could I be right? Why not admit that I might be right? Because it's a political topic and your opinion is set.

I think parent is talking about professions where the scientific method is followed. It's not about how "sophisticated" something is.
Do you anything about data science or deep learning? If this is not about scientific methods then what else?

And even if not? Why not question existing methodology, people just follow some news outlets believing one and only thing.

Data science is a science in a name only.
To put that another way: most people who call themselves data scientists do not have a PhD and have zero research publications.
False claims. There're as many who do and do you know what it does not matter for this discussion because data science and deep learning is a science independently of what some people do or not. This was the point.
Haha sure, read some papers about Bert and its successors. Come back and summarize what you have learned. But your comments shows so clearly that your opinion is set and politely you should check out data science and deep learning before you again write such shallow dismissals (which is against the guidelines btw).
Because no two fields are alike? We don't do anything even remotely scientifically in this industry because it's one of the few where we don't need to for various reasons
Basically compare how software is developed by all of us to how software was written for the space shuttle.

There is just absolutely no comparison in process or methodology. Space shuttle programming is what true software engineering looks like.

I've always thought that true engineering is anything that involves human lives. Bridges, pipes, roads, pace makers, space ships, rockets. Websites don't kill people, it ain't engineering.
That doesn't sound right. Most applications of electronic engineering are not safety-critical, unless we count requirements such as avoiding improper use of toxic chemicals, but setting the bar that low would mean just about any physical undertaking counts as safety-critical.

Projects like the Linux kernel, or the HotSpot and LLVM compiler systems, presumably count as engineering. Building a new GPU presumably counts as engineering. These systems aren't intended for safety-critical use, though.

Another example: it still counts as aeronautical engineering even if you're building an unmanned drone that only ever flies in a lab.

Fear and hope.

But do you mean: so-called experts? As in, not experts? Well, then that's just a question of finding the real ones, if they exist.

Or do you dislike how they oversimplify when talking to the public? I do too. All nuance is erased. The trick, then, is find out what (true, non-politized) experts say to each other when the public isn't listening.

Problem is that most people don't listen and turned off their brain. Look at this thread. They rather believe some random experts and news outlets calling some random dudes with random degrees experts because they have themselves no clue, never lead always followed others. What do we expect?