Live data from Hacker News

Jobs at Apple-Search

jobs.apple.com

61–68 of 68 posts

Re: Jobs at Apple-Search

#61
post #42

Earlier quoted context omitted.

There is no way they would make enough from iPhones/iPads to pay for the costs of building and running a competitive web search engine. (At least not one built like Google and Bing are) There is a reason they take a cut of profits from App developers; it's to help pay for the cost of running the App store.

They're generating profits at a rate of $200 Million per day [sic]. It's fair to say they make enough to pay for the costs of almost anything, including this. You've identified the one requirement they can in fact satisfy; the hard part is... everything else. Edit: Clarity

Actually about half of that:

"For the full fiscal year, Apple generated $182.8 billion in sales, a new record for the firm, with $38.5 billion in net income."

$105MM, not $200MM.

Re: Jobs at Apple-Search

#62
post #55

Earlier quoted context omitted.

You may be right. Assuming you're right about their profits, that's 18B per quarter. Operating cost have been said to be around 5B per quarter for Google and 1.25B per quarter for Bing.

Source please? Bear in mind they're already running their own competitive map service for exactly the same reasons I say they want to run their own web search. And that's probably not cheap either.

My source for the costs of Google: http://www.datacenterknowledge.com/archives/2014/07/23/from-...

"Google data centers cost the company more than $5 billion in the second quarter of 2014"

My source for the costs of Bing: http://www.businessinsider.com/microsoft-bing-losing-billion...

"Put differently, Bing is spending $5.5 billion a year to generate $3 billion of revenue." (I divided by four to get quarterly)

GPS data is not really on the same level as Web Crawl data. How often is a street added or changed versus a web page?

Re: Jobs at Apple-Search

#63
post #42

Earlier quoted context omitted.

They're generating profits at a rate of $200 Million per day [sic]. It's fair to say they make enough to pay for the costs of almost anything, including this. You've identified the one requirement they can in fact satisfy; the hard part is... everything else. Edit: Clarity

Actually about half of that: "For the full fiscal year, Apple generated $182.8 billion in sales, a new record for the firm, with $38.5 billion in net income." $105MM, not $200MM.

You're right with those numbers, but it's only half if you take the year's average per day. I was referring to their current rate (as of the most recent quarter, during which they earned $18B). $18B gives the nice, round, shocking $200M/day, but even $100M per day isn't half bad.

Re: Jobs at Apple-Search

#64
post #55

Earlier quoted context omitted.

Source please? Bear in mind they're already running their own competitive map service for exactly the same reasons I say they want to run their own web search. And that's probably not cheap either.

They started running their map service because google was about to start charging for Google Maps. iMaps already had a rollout that felt improvised and pretty crappy, if the Apple people were suicidal enough to get in a huge business they have absolutely no idea about, such as Web Search, I wouldn't expect a barely usable product in years.

I don't remember anything about a fee, but I do specifically recall something about Google withholding from the iOS version two greatly desired features that had already been deployed on Android: vector tiles and turn-by-turn directions. The lack of the latter, at least, was beginning to rise to the level of consumer awareness as an iOS weakness.

Re: Jobs at Apple-Search

#65
post #55

Earlier quoted context omitted.

Source please? Bear in mind they're already running their own competitive map service for exactly the same reasons I say they want to run their own web search. And that's probably not cheap either.

My source for the costs of Google: http://www.datacenterknowledge.com/archives/2014/07/23/from-... "Google data centers cost the company more than $5 billion in the second quarter of 2014" My source for the costs of Bing: http://www.businessinsider.com/microsoft-bing-losing-billion... "Put differently, Bing is spending $5.5 billion a year to generate $3 billion of revenue." (I divided by four to get quarterly) GPS da…

1. $5B is super way off. First of all you are lumping one-time construction costs in with the incremental costs of running the data centers. The article clearly states the bucket in which data center operational costs are included is $2.82B.

2. Even that bucket is a grab bag of other costs as well such as amortization of acquired assets (not necessarily for data centers). Additionally while web search is major it is certainly not the only thing that Google data centers do; for example they run Gmail and an AWS competitor on it. So it's very hard to determine just what the operational cost of web search is, but it's certainly well under $2.82B.

3. The Bing article clearly states its major cost increases are "costs associated with the Yahoo! search agreement and increased traffic acquisition costs", i.e. promotional not operational costs. This is not applicable to Apple, who already gets paid for traffic.

4. As for maps it's not just GPS data, it's road and signage and building data including visual imagery. That's many orders of magnitude more expensive to gather than loading a web page (even loading it 1000 times). You have to roll trucks and fly planes. In fact there was just a story yesterday about Apple rolling trucks to start gathering competing "streetview" data.. all from scratch.

Re: Jobs at Apple-Search

#66
Apple was at Lucene Revolution in DC this past fall. It's a little silly IMO to extrapolate too much from a job posting -- Apple has had a large search infrastructure for a long time, as do most tech companies.

Re: Jobs at Apple-Search

#67
post #65

Earlier quoted context omitted.

My source for the costs of Google: http://www.datacenterknowledge.com/archives/2014/07/23/from-... "Google data centers cost the company more than $5 billion in the second quarter of 2014" My source for the costs of Bing: http://www.businessinsider.com/microsoft-bing-losing-billion... "Put differently, Bing is spending $5.5 billion a year to generate $3 billion of revenue." (I divided by four to get quarterly) GPS da…

1. $5B is super way off. First of all you are lumping one-time construction costs in with the incremental costs of running the data centers. The article clearly states the bucket in which data center operational costs are included is $2.82B. 2. Even that bucket is a grab bag of other costs as well such as amortization of acquired assets (not necessarily for data centers). Additionally while web search is major it is…

Oh, my mistake. You may be right about Maps data being more expensive, if so then it does make sense for apple to make a Web search engine.

Re: Jobs at Apple-Search

#68
post #65

Earlier quoted context omitted.

My source for the costs of Google: http://www.datacenterknowledge.com/archives/2014/07/23/from-... "Google data centers cost the company more than $5 billion in the second quarter of 2014" My source for the costs of Bing: http://www.businessinsider.com/microsoft-bing-losing-billion... "Put differently, Bing is spending $5.5 billion a year to generate $3 billion of revenue." (I divided by four to get quarterly) GPS da…

1. $5B is super way off. First of all you are lumping one-time construction costs in with the incremental costs of running the data centers. The article clearly states the bucket in which data center operational costs are included is $2.82B. 2. Even that bucket is a grab bag of other costs as well such as amortization of acquired assets (not necessarily for data centers). Additionally while web search is major it is…

Chatting about this with you made me challenge my knowledge of how web search works. Many articles I've found suggest that the hardest problem to solve isn't competing on the size of Google but the relevance of Google. Which is partly to do with the size and power of their datacenters. But also to do with smart engineering of the servers, database, and algorithms. The algorithms not only determine the how to order the results but which pages to crawl and how often.

What this all makes me think of is all the work being put into AI research lately. IBM, Google, Facebook, Baidu etc. There is a race to AI and the winner could very well make a better Google.

Post reply on HN