I think what's sad is not that they shut down the project soon after acquiring Titan Aerospace, but that they kept it quiet for almost a year. That sort of PR practice makes me worry about their internal culture.
Edit: It concerns me because I wonder how many other projects they've announced and shut down without telling us, and what that must be like for the employees who are no doubt under NDA who are shuffled around and unable to talk about it. Imagine if your company were acquired by X, and then you get fired, and you can't even tell your friends or potential employers it's because X axed the project.
In a broad interpretation, such a practice could be interpreted as innovation stifling. Buy a solid player and covertly shut it down -- (1) no expenses for you outside the buying and (2) others might/will be discouraged to progress in the area because you, Alphabet/Google sent the message that you will do things in the area.
Also Microsoft's unreleased booklet PC, Microsoft Courier. Never officially announced as anything other than a concept, but they did offer an official notice that the project was canceled.
Owning up to mistakes is the fastest way to make progress. Waiting so long to admit that a project failed worries me. X is supposed to be a moonshot factory, where failures are celebrated and reality dictates that the majority of their undertakings will fail. To me, the delay is indicative of a culture more concerned with appearances than substance.
Maybe X had good reasons for staying quiet for so long, but unless they say why it's hard to give them the benefit of the doubt on this one.
> Owning up to mistakes is the fastest way to make progress. Pretending they never happened is indicative of an opaque culture that is more concerned with appearances than substance.
What makes you think this didn't happen internally?
I don't know what happened internally. But if that happened internally and they weren't willing to announce it externally, that shows they are concerned with appearances. And if that didn't happen internally then they have a cultural issue where they are afraid to embrace failure and learn from it.
Obviously I have limited information and I am not trying to define X based on this one event. But I do think it says something about the culture there.
A large part of why the original Skunk Works was so successful is because Kelly Johnson, and later Ben Rich, did not care about outward appearances. They had a job and they did it (and they made it profitable). Here's a great book on the subject.
They did the closest thing they could - when a project failed, they killed it and refunded money to the government. In the defense industry, this is pretty rare (the F35 project is widely criticized as a failure by almost everybody that's ever used it but it's still going to get hundreds of billions of dollars in future funding). They could have said nothing, pretended everything was fine, and continued to siphon money from their customers but they decided to cut their losses early and maintain their reputation as an organization that gets things done on time and under budget. This got them a lot more contracts moving forward.
First, skunk works never refunded money to the government. There is at least one anecdote that they tried but the government accounting office had no way to receive the money back.
jayjay's initial comment was that the lack of PR was concerning, but then brings up a super secret facility that had no PR whatsoever as a counter example. The comparison just doesn't make sense.
But let's say that your claim is true, that skunk works communicated with their customer on a failed project. That still has no relation to a public announcement about a failed project. This was a self-funded project by Google, they are their own customer. The lack of PR about internal projects says nothing about their internal culture.
I think the comparison between Skunk Works and X is excellent. They're both moonshot branches of much larger companies that are meant to be free from bureaucracy. There won't be a perfect comparison, and obviously there won't be PR announcements for classified projects, but the point remains - focus on substance instead of appearances and you'll be much more successful. There's lots of evidence that Skunk Works had a great culture, but there seems to be little of that for X. In fact, there are numerous reports of people leaving X in the past year.
Must have reached the nesting limit, can't reply to your other comment.
> They're both moonshot branches of much larger companies
Not at all! Skunk Works started with one single focused mission. To design a jet to compete with the German Messerschmitt! Yes, they did a lot of things differently to clear bureaucracy out of the way and get stuff done. But to call Skunk Works a "Moonshoot" branch is not correct at all. The definition of moonshoot includes not having an expectation of profitability. Lockheed saw a threat, saw a need, and saw huge potential for profits!
It could be the case that not caring about appearances is one of the best ways to optimize for appearance. Somewhat like men who don't care (or do not give the appearance of caring) what women think tend to get more women than men overly concerned with appearance, even though women judge men by appearance (including what non-physical traits they appear to possess).
That seems to have happened to some of the robot companies Google bought. Bot and Dolly [1], which did robot-assisted special effects for Hollywood, had actual products, customers, and film credits in major movies. Now they're just gone.
That's what I don't understand. If they pushed this technology further (high-altitude, solar powered drones) it seems it could take over a big chunk of satellite industry. If you don't actually need space, then solar drones seem like so much better option - significantly cheaper to make, significantly cheaper to launch, etc.
> no problem with keeping the batteries charged overnight while flying
Satellites in LEO have night and day cycles too, albeit shorter. But you have more problems with keeping batteries cool during the day and warm during the night, and also each additional gram you want to add to a satellite costs a lot of money thanks to the magic^Wtyranny of rocket equation - so you can't just spam solar cells and batteries like you could on an aircraft.
> no FAA/other countries' counterparts issues, etc.
Space is pretty regulated too; you can't just lob a piece of junk into whatever orbit you want and call it a day.
> just spam solar cells and batteries like you could on an aircraft.
Things like the Titan drone are mostly solar cells and batteries. It's not like you can increase the % of mass allocated to that because it's pretty much maxed out.
It's easier to keep things in space, but it's much more difficult to get them there because you have to use a huge rocket to make the thing go fast enough to stay on orbit. If you don't need that speed and the ability to circle the whole globe, high-altitude planes seem like much cheaper options.
You'd also need a much larger number of the solar powered drones to cover the same area as a satellite. They don't make it clear but I'd guess one of the main issues was getting the radio and networking gear light, small, and low power enough to work in their drone design with enough throughput to make the network vaguely acceptable while covering a large enough area to make the number of drones required reasonable.
OTOH having to have more drones gives you benefits of economies of scale when you move past testing phase; you can mass-produce a platform and drop the price significantly. Satellites, on the other hand, are mostly one-of-a-kind constructions, produced in single or double-digit numbers.
(Also with a fleet of drones, failures are cheaper and less dangerous.)
There are lots of factors affecting economics of both solutions. I guess one of these days I'll have to make a spreadsheet with back-of-the-envelope calculations to sort this all out, but my gut feeling for now is that solar drones are a viable replacement for a lot of current satellite uses.
Making more would help but there's still limits on how far down it could possibly drive down the price on this large complex drone.
To me the largest issue is simply power, these drones need to store a lot of energy just to fly overnight. Add in the power requirements for all the radio gear, it'll need to operate overnight too to make the service particularly useful, and battery weight can quickly spiral outside of what can reasonably be done by a small company trying to build a small plane. There's also 2 hard competing design goals solar panel area and aerodynamics. The first wants broad flat surfaces to place as many panels as possible on while the second is pushing you towards long thing wings and bodies to reduce drag.
I'm not saying these are impossible things to build but I do think it's more in the realm of large aeronautic companies rather than a small division of google.
Satellites built to accomplish the same task as these drones would not be higher volume than that. Neither volume, though, would getuch benefit from scale. Specialty parts that required significant development would be expensive even if the production run was in the thousands.
It's a shame (non-military segments of) governments don't do this kind of projects. Instead we're spending tons of money on rusting tanks, flying killing machines and what not.
Let's pretend the US ended all military spending. Would you think that wars and defense would just become irrelevant? A strong military prevents wars. Would Taiwan be free of it weren't for all those "flying killing machines?" Would South Korea or Japan even exist if it weren't for the military strength of the US? The main reason we didn't go to war in Europe against the Soviets was because of all of those 'rusting tanks.' Communism didn't fall because of luck; the US outspent and out produced the Soviets and bankrupted them. Threats to freedom still exist and they aren't stopped without diplomacy backed by the threat of force.
Most of the most significant innovations of modern history came about because of free-market capitalism, not because of governments -- with very few exceptions.
The role of government isn't to innovate -- it's to maintain an environment that makes innovation possible.
I would say that many significant innovations have been driven by the military applications. Rocket technology and drones are 2 that come immediately to mind but I am sure there are many others. Not defending the military industrial complex though.
Would you really trust government-controlled internet provider? I'm not usually one to spout conspiracy theories, but even I'd be unable to prevent doubts about the content coming through it.
I don't really think there's a difference between what we have right now and a government-controlled provider. The major backbones are already working hand-in-hand with the government, major news sources are working with the government, social news sites have government-published stories. There's no way to validate that anything isn't from the government on the Internet.
The only difference would be in perception, not in reality.
I don't know that I agree with your first paragraph, but validation of what's from the government isn't the only problem. In the internet age, freedom of speech effectively requires internet access. If the government can boot you off the internet or ban your content, that's a major blow to freedom of speech.
My impression is that Titan and Loon are meant to work with phones, whereas a receiver for SpaceX's thing would be the size of a pizza box. No way it's gonna be free.
Interesting. Though I wish someone pursued this particular tech (solar-powered drones) further - high-altitude, long-duration drones could be a good replacement for many things we use satellites for, and drones would be much cheaper to launch...
There are lots of companies working on light solar-powered drones, and in fact some of them call them pseudo-satellites. Airbus, Boeing, NASA, DARPA... the list goes on and on. This really isn't a setback for the field in general.
A few nights ago, over a liberal quantity of beers, my friends and I came up with our latest nonsensical conspiracy theory. Google was started by some of a group of time travelers who were sent back in time from an apocalyptic future bringing advanced technology with them. They would use just enough of the tech to make vast sums of money, then use that money to invest in the tech that destroyed their future world.
By investing in tech they already know would work, they would vacuum up all the resources and talent around that technology so that nobody else would bother (because who could come in cold and compete against a titan like google?) and then some time later kill all of the projects very publicly in order to make us all think that these technologies were nonviable.
Except instead with every piece of world changing technology they killed, they were buying their future Earths freedom from the apocalypse they were fleeing from, even if that future they were creating destroyed their own and left humanity in a just post-agrarian stage of development.
This group has been infiltrating our timeline for decades, first with blimps, then Atari, then Xerox, now Google.
Experience is subjective. If you teleport to another location you're not the same person you were before you left, you're a similar set of molecules reconstructed in the same fashion. In all actuality, "you" died and another "you" took your place. But to you and everyone around you, it seems like nothing has changed, so you carry on with your life.
Likewise, if you travel back in time and change the future, you're not helping the people in the future you left, you're helping the people in the past avoid that future. But to you, it looks the same. It looks just like the past you're familiar with, and it develops into some semblance of the future you were expecting. So for all intents and purposes, from a subjective viewpoint you're avoiding the paradox and helping your ancestors. It doesn't matter if the people you left are still suffering, subjectively they don't exist anymore.
This is the exact explanation used in the anime Steins;Gate. Okabe (the main character) is capable of 'moving timelines' which is treated as jumping between different parallel worlds.
He leaves his friends behind in a world where WW3 breaks out over world powers fighting over research for a working time machine. In a way, he is leaving his friends to die in a post-apocalyptic WW3...but that doesn't matter because he's saving them in his 'current' timeline. The other worlds, from his perspective, "don't exist" anymore.
>In all actuality, "you" died and another "you" took your place. But to you and everyone around you, it seems like nothing has changed, so you carry on with your life.
In the show they call this power "Reading Steiner" when he "replaces himself" in a new timeline. Since he is capable of remembering what happened in the other timelines (nobody else does - as from their perspective these other timelines never existed).
In the Star Trek universe, where the matter is converted to information, transmitted and then converted to matter, then your statement may be true. But, in other universes where the matter is actually translated in a dimension then it will not be. Also, we don't actually know what "you" is and whether it would be included in the Star Trek mechanism.
My point really, is that there are varying fictional universes but in ours, we don't currently have a way to do teleportation. So, as you say, experience is subjective.
If you stay in the "new" past, you can just enjoy the fruits of your labor. If you jump back to the future by conventional means, like hibernation/cryogenic sleep, you will wake up in the same changed universe, including altered versions of the loved ones you left behind (if you are lucky [or unlucky if they change in a bad way]).
If you jump back into the old universe, nothing will have changed – so that is pretty much the only place you should stay away from.
There won't be a future where people knowingly will have experienced the results of your heroic deeds, which might be sad to some, but you can just tell/try to convince them of what good you did. Maybe take some videos or other proof.
They must be maintaining some form of communication with their future and results are being reported back to our time. So there must have been some sort of flying drone rebellion forming so they had to be stopped from existing.
If you assume the Many Worlds interpretation of quantum mechanics, no paradox would be created. Merely time-travelling at all forks off a brand new universe with it's own future from that point on.
That said, you do lose the drama surrounding the possibility of "destroying the future" via making gross (killing Hitler) or subtle ("butterfly effect") changes, as none of the consequences ever affect the universe you travelled from (other than the fact that you've left, never to return).
On a sidenote, what exactly did Google Reader do that was so great? I'm not knocking it, it's a real question.
I ask because a lot of things Google has killed is hard for us to replace. Eg, i don't want to try and build Wave, or continue with Apache Wave, etc. It's a big, complex problem. But a personal content aggregator like what i thought Google Reader was seems quite easy to implement.
In other words, why haven't we (public or startups) built the core functionality of Google Reader that so many appear to love? It seems like an eager and easy to reach market. What am i missing?
It's a cloud-based RSS and ATOM reader that frequently updates, has unlimited storage duration, is cloud-based, and free. The UI is also a plus - GMail-style, you can read a thread, read by time, see all unread fairly easily.
There are already quite a few replacements around that are better in my opinion (like inoreader), but people seem to like complaining about Reader as if Google killed off their firstborn for some reason.
I'm one of the people who really enjoyed it. I enjoyed the hassle-free nature of it, and it basically did what I expected to in a decent interface.
Most other solutions that I've found make some assumptions about what I'll be reading or how I'll read it and try to force me into that, claiming it's for my own benefit.
In the end, I ended up self-hosting "Selfoss". I have to keep it up to date and occasionally fix a bug when it won't read a stream properly, but it basically does what I want otherwise.
It was a one-stop shop that had all your content in an easy to read format sans ads and tracking. I think Google realized that if everyone went with Reader their cash cows of ads and tracking would disappear, so they killed it.
Its probably trivial to replicate, but RSS/Atom wasn't very popular to begin with and this move just motivated others to give up on it. Webmasters didn't want to have that weird RSS icon that no one understood and designers probably didn't like that their hard would went away with RSS.
On the flip-side ad blockers and things like Firefox's and Safari's reader mode help a lot. I think its not a big coincidence Chrome is late to the reader mode game either. Google wants you to load that ad tracking .js file because that's their bread and butter.
You go and add a feed, and then you don't just get the latest 20 posts. Instead you get all the posts since the beginning of the feed, since google saw it then during a crawl and decided to watch it ever since.
That is something that is a fair bit harder to provide by any competitors
Couldn't you (gently) use the wayback machine for this? Or just start crawling feeds with something like scrapy? There isn't enough RSS, I feel, to make this prohibitive.
> As feeds move beyond merely announcing new content on somebody's blog and into organizing data, you can easily find situations where you don't want your feed to include all of the available data. This tip shows you how to create an Atom feed that lets users page through it using "next" and "previous" links or button.
[..]
> Creating this functionality in PHP doesn't require any special skills or modules, just a methodical approach.
Nuff said?
edit: why downvote? It'd be trivial for most site to make ALL past items available, without having to offer one huge feed with all those items in it, and any reader would be free to read all those items when subscribing to a feed. There, done.
In the 5 years or so since Google Reader's demise I'm yet to come across a decent feed aggregator. I'm currently using Feedly but it sucks compared to GR.
I use Feedly too, but the lack of other decent feed readers or aggregators suggests that there really isn't a user demand for RSS. OTOH, many sites and news organizations post to Twitter as a rudimentary RSS feed. Google probably could have adapted Reader into the social platform they wanted.
RSS (in the sense of using a reader to track a personally curated set of RSS feeds) is pretty niche these days. One audience is journalists who really do want to discover news from a variety of sources faster than someone else does.
Most people are fine though with just using social. If it's noteworthy enough they figure they'll learn about it soon enough. Personally, I do have Feedly configured but I confess I don't use it systematically any longer.
I'm still a bit puzzled about Google shooting Reader. Keeping it would have been a tiny effort and it doesn't seem much removed from a mission of selling ads (excuse me, organizing information). But, then, Google has largely walked back/de-emphasized pretty much all their information curation projects other than straight search.
They nearly collapsed under the load a few years back when Google cut GR, nearly shutting down. But now it's solid, and they seems to have got financial backing.
There are few limitations (100 feeds max) with a free account, but for me, it's enough.
Plain decentral RSS is being used and has been all the time since like 1999-2001.
If you're missing a big-time "disruptive" announcement on HN, let's run a story about RSS so that it becomes a thing again in people's minds. We can pepper it up with JSON support and call it RSS 3.0 (but please let's not go down the Atom rabbit hole).
1. Social aspect of it - I would find "human filters" that shared very high quality content on different topics - art, programming, medicine, architecture - and that was very transparent for them - they just kept interesting articles for themselves, but I also had access to them.
2. Search
3. Archives
As of now I think HN/reddit/twitter serve the same purpose (poorly - twitter more and more tailored for person with thousands of subscriptions and not for someone who's carefully organizes his "human filter" collection; and none of services provides subscription for content - hell, I have dilbert as email subscription now)
I think my conspiracy theory is more likely to be true:
>I've speculated for a long time that basically anything interesting Google says they're doing is essentially meant to be a jobs program to keep employees from leaving, PR for external stakeholders like investors, media, being attractive to potential employees, etc. They seem to have lots of formal ways to keep employees from leaving/close as well including investments off of Google's balance sheet (not GV or Google Capital) into ex-employee startups and just flat out paying people not to leave (which is the arrangement I'm guessing that Matt Cutts is under). It all seems very Microsoft of old.
Tornadoboy's source is good, but primarily focuses on the amount of heavy communication between Google and government, so I'd prefer to add this source specifically, on the strong revolving door effect:
Campaign for Accountability's page, above, has an interactive chart where you can see which employees, by name, moved between Google and the government during Obama's term, and what positions they held on each side. Beyond the "Read more" link is also a significant amount of specific detail about some of the more notable transitions.
Oh yeah, Loon is going to be so much more feasible. Loon makes the spatial economics of providing internet so much worse since it has to cover the 3/4 of the world that is ocean plus all of the places that are deeply uninhabited.
To get fiber into rural places you just have to parallel the electric lines. Except there is something ritually unclean about wires and we can't do it.
This article has a few details that are very different than the way I heard it. What follows are rumors, but I work in this industry so I have a few sources that I trust.
Google did not win the bidding over Facebook. Facebook actually won the bidding with a bid of about 60 million dollars. Facebook pulled out during the due diligence phase. Google swooped in and picked it up for about 20 million after Facebook pulled out. (Rumors remember, I can't find sources for these numbers right now)
Internally, Google was very quickly disappointed in their purchase. They found out that they had pretty much bought a nice PowerPoint slide deck and not much else. Titan still had a long way to go before they had a viable drone.
Anyway, that is the rumor that I heard.
I'm only somewhat envious of the drone companies that are selling out for tens of millions of dollars, while I sit here and work my butt off to keep making a product.
https://youtu.be/b7SjOOuTct0
Much like aircraft-delivered nuclear weapons, are these being quickly replaced by impending low orbit micro-satellite constellations under development now?
These days upstanding citizens is hard to believe anything from these corporations. All news about anything seems to be fake.
reply