Hacker News Reader

Comment by m0nastic | Hacker News Reader

I don't have any kind of knack for predicting things, but I'd argue that it's at least as likely for the "Star Trek computer" to be what things like Google Now turn into (which as far as I can tell, isn't open source at all).

I would be very surprised if the aggregation of technology required for this type of interface doesn't require a company with a lot of services know-how to champion (I could see Apple, for instance trying to go down that path, but I don't think they've ever shown any aptitude for online services).

Maybe after someone makes it, a shitty, open-source knockoff will show up; but I don't think what's laid out in this article is a foregone conclusion.

View on HN

Maybe after someone makes it, a shitty, open-source knockoff will show up; but I don't think what's laid out in this article is a foregone conclusion.

Perhaps not. But nobody wants to read a headline like:

"Some reasons why I think that it's possible that maybe, just maybe, something like the Star Trek Computer might come along someday and it might, with some luck, be open source, and it could be Apache Licensed, but maybe not".

At some point, you have to say something, and while I'm not big on headlines that are outright "linkbait" if you write stuff, you typically want people to read and comment on it, so the headline has to be somewhat catchy and maybe even a little controversial.

The key point here, isn't really the posit that the Star Trek computer will be Open Source and ALv2 licensed, it's that a lot of awesome work is going on in various semi-related ASF projects right now, some or all of which could well become part of something like the Star Trek Computer. But, again, that's too long and wordy for a headline.

So you're suggesting that articles without compelling content should still have compelling titles?

The goal is certainly to gain readership, but adding controversy to a title on an article that doesn't contain any isn't a good idea. It's a misrepresentation of your article, and what extra traffic it does bring you isn't likely to result in happy new readers.

Some articles just don't appeal to a lot of people because of their content. That doesn't mean a title that reflects what's in the article is a bad idea; it just means the author should either have realistic expectations regarding the exposure their article can reasonably expect, they should be writing something that is controversial, or they should be writing something else entirely.

If your goal is to communicate specific information and your title doesn't reflect it accurately, all you've done is fail at titling your work.

So you're suggesting that articles without compelling content should still have compelling titles?

Not at all. Obviously this exists on a continuum and is somewhat subjective, however. I would never advocate posting low quality content with a flat-out link-baity title like "Learn About Bill Gates and Ada Lovelace's Secret Lovechild: Mark Zuckerberg" or something. But I think you have to create a title which is as compelling as you can, while being faithful to the content.

But, as in all cases, there will always be people who agree and some who disagree about whether you've accomplished that or not.

All I'm getting at is that a headline shouldn't dissemble and be wishy-washy and say nothing. It's an opinion, that I'm asserting (that the "Star Trek Computer Will Be..."), but even I won't go quite as far as saying it's "a foregone conclusion". There is evidence to suggest that such a thing may be the case, and that's what this post was about.

In this case, I'm perfectly happy with the congruence between the headline and the content. If others aren't, then I'll be curious to hear their POV on it.

Google has, internally, a lot of super secret services that are absolutely critical to this sort of functionality. "Simple" things like the language model necessary for their understanding of "Apple" vs "Apples" are actually astonishingly complex, and expensive to both build and maintain. Currently, there is virtually zero chance of a merry band of open source hackers being the ones to create these numerous services and stitch them together in a meaningful way.

fake it till you make it

i'm working on a "iron man" suit that really works. It's easy to do performances where I talk to "Jarvis". He's really been programed to be an actor who plays Jarvis and that's good enough.

Reply on news.ycombinator.com