inimino

48 ms·

541.
by inimino 7y ago
Though this is not particularly strong, I don't think you would get similar strength from an n-gram model. You need longer-term correlations, which is generally where transformers do well.
542.
by inimino 7y ago
Fair enough. At that point, the term excludes so much that I don't think it really has much value. At least, I don't think it's easy to argue that "compiled to machine code ahead of execution" is something that a us
543.
by inimino 7y ago
I wouldn't bet on current funding patterns lasting for the next 50 years. AI winter is probably coming soon. All your other points I agree with, however the higher-level counterargument is that all these things have been just around th
544.
by inimino 7y ago
Generally the problem with systems like this is that you need a critical mass of users to rate books for a site to have any value. If the only ratings you see are from people you trust, this becomes a lot worse. Once you add UI friction, ad
545.
by inimino 7y ago
> where is the incentive Regulation.
546.
by inimino 7y ago
Is it rational for a customer to not fly on a plane that is as safe as any other just because the same company later made another plane that was less safe? It seems like a lot to ask. In any case, even if 10% of people totally stopped flyin
547.
by inimino 7y ago
The number of publications per year per academic seems to me to have increased over the last 50 years. I don't have a citation. Regardless, my original claim was that the absolute number of papers is growing, and most of them are trash
548.
by inimino 7y ago
Well, I mean, do you have to use all of them? Where do you draw the line? Does using any third-party library make something non-native? With languages that compile to other languages, toolkits that wrap or abstract other toolkits, it's
549.
by inimino 7y ago
Fair enough! You're not responsible for using organizational clout you don't have, but can still look for opportunities to make things better as they come.
550.
by inimino 7y ago
OK, which ones?
551.
by inimino 7y ago
> the older the tech, the longer it will remain in active use You may already be aware, but this effect has a name: https://en.wikipedia.org/wiki/Lindy_effect
552.
by inimino 7y ago
The post you originally replied to listed specific technologies that purely add complexity when introduced to a project. Hadoop isn't replacing something equally complex, it's simply adding complexity. Now if you simply can't
553.
by inimino 7y ago
It's entirely possible, and I certainly hope so. But it hasn't happened yet. The history so far suggests that better technology will generally be ignored unless there's a clear way to make money from it now , which tends to
554.
by inimino 7y ago
You really missed the point. It's not "there is an older way that works fine". It's that there is a simpler way that can be built in less time, takes fewer developers to maintain, will have less bugs, performs better, an
555.
by inimino 7y ago
OK, I'll bite. What's your definition of "native" that conforms to your intuition and works equally well for all platforms (iOS, Android, Windows, Mac, Linux, ...)? Don't get me wrong, I think there are plenty of re
556.
by inimino 7y ago
Getting the job done and being employable are more closely correlated in some positions, industries, and markets than others. Finding those is a key to job satisfaction if this sort of thing matters to you. From the other side, if employmen
557.
by inimino 7y ago
I suppose this is true. I haven't had a chance to work with or teach anyone who is learning to think this way, and I don't really remember what it was like for me. However, I've noticed that when I talk to some people who are
558.
by inimino 7y ago
There have always been weaker scientists, but there hasn't always been the economic incentive to publish in order to maintain a teaching position. This is a relatively recent (few decades) thing and is due to structural factors in acad
559.
by inimino 7y ago
There is no such thing as one right answer from a search engine. Use them both, get used to how they work, and see what works for you.
560.
by inimino 7y ago
You can judge a musician in a 20-minute audition, but you can't tell if research will reproduce without doing the work. However, there are papers that are so vague and poorly written that they cannot possibly be reproducible, and this
561.
by inimino 7y ago
It's true that science doesn't value every published paper equally, but it's also true that publish or perish is creating ever-growing mountains of worthless papers. This is a real problem, and drags the quality of everything
562.
by inimino 7y ago
Ah, yes. Wishful thinking.
563.
by inimino 7y ago
Isn't it the exact opposite? Presumably the papers scientists are reading are primarily the ones in their own field.
564.
by inimino 7y ago
At least favoritism has some variability in it. The problem with metrics, besides being gamed, is that they are so uniform.
565.
by inimino 7y ago
There may be some tiny influence that the HN crowd can have through the impact of hiring influences in industry on respective incentives in academia, but overall, I agree completely. Solutions are easy to state but as always the difficulty
566.
by inimino 7y ago
Making a significant contribution to your field used to be a reasonable requirement for a Ph.D. However, we are churning out more and more Ph.Ds, and there just aren't that many new discoveries to be made. This is creating a pressure t
567.
by inimino 7y ago
Because we design it, so we have the ability to choose its characteristics. The current prioritization of 1% improvements in artificial benchmarks over all else is a disease that is actively retarding progress in the entire field. Political
568.
by inimino 7y ago
I encourage people to learn the language and avoid the ecosystem. You don't need frameworks or build systems to write good, clean, performant web apps. Unfortunately, arguments for simplicity often fall on deaf ears.
569.
by inimino 7y ago
I've heard this criticism a lot, but I don't really understand it. In comparison to Lisps, Haskell, and other languages I love, of course Go is inexpressive, but that's part of the point. In comparison to languages with which
570.
by inimino 7y ago
Assume good faith, and try to get the point. "That is a question that a junior developer would ask" is not on the same level as "that is idiotic". If you're asking after "hot new tech" it's probably b

More ›