Tales from the Google interview room
Friday, January 5th, 2007taste of bacon, either.)
However, as we discovered when we interviewed the creator of an “Artificial Intelligence Chat-bot”, programmers who develop algorithms tend to encode their own shortcomings into the systems they create. [see Do Artificial Intelligence Chatbots look like their programmers? ]
And the Times confirms that the job-bot’s selection criteria is based on surveys from existing staff. One of the indicators is ominously called “organizational citizenship”. No square pegs in those round holes, then.
In Douglas Coupland’s Microserfs, the company’s monoculture is enforced by obedience to the cult of personality – top down. By contrast, Google appears to be developing its monoculture from the bottom-up. But it’s still a monoculture – and one only likely to be reinforced by algorithmic rejection of “unsuitable” candidates.
As we discussed here recently, an algorithmically-minded corporation is ill-equipped likely to miss problems that can’t be solved algorithmically. No robot can wish them away.
If you have an amusing experience of Google’s recruitment practices – successful or otherwise – share it with us here. We’ll set our own robot on the replies, and pick out the ones whose opinions most closely resemble our own.
(Just kidding).
The DIY encyclopedia
Wednesday, December 20th, 2006
Who can fail to love the can-do spirit and have-a-go enthusiasm of Wikipedia? When the site found itself in need of copyright-free illustrations, one user simply generated his own.
We were alerted to this cockle-warming tale via a Something Awful forum, where member Stick_Fig, sets the scene like this:
A group of users has decided that because these promotional photos, which were previously allowed, are copyrighted, they need to be replaced with copyright-free images. Like, images taken by nerds for nerds. The argument is that, since the person is alive, by God, a photo can be taken, so we MUST remove the old, perfectly-fine-minus-a-little-copyright photo now.
Readers poured forth with heroic hand-crafted illustrations, such as the one above.
It was only when it was discovered that the site’s entry for “semen” was in need of copyright-free illustration that one member heroically rose to the challenge. Or rather the member’s member did. And what a splendid contribution it is.
So no more gags about Wiki-Fiddling, please. This is truly an example of “User Generated Content” at its most spontaneous.
As Tim Bray observed recently:
“There’s been a surge of recent editorial activity with super-energetic (and apparently well-informed) new contributors trimming and tweaking and growing the articles, often several times per day. In general, while I haven’t been convinced that 100 per cent of the changes are improvements, the quality of the articles as a whole is definitely trending up.”
Um, quite. How can Britannica possibly compete with that?
Now you know: Blogging is 'un-Christian'
Tuesday, October 10th, 2006Blathering on blogs is un-Christian, an Evangelical church has warned.
“Blogging has become a socially accepted practice – just as are dating seriously too young, underage drinking and general misbehaving,” notes the monthly of the Reformed Church of God, Ambassador Youth.
Blogging “often makes the blogger feel good or makes him feel as if his opinion counts – when it is mostly mindless blather!” notes Kevin D Denee.
“People will now do and say things that should only be done in private, or, frankly, should not be said or done at all,” rues Denee.
“Propriety, decorum and decency are not elements considered on blogs. People simply blurt things out, without considering the contents or consequences.”
(more…)
Do Artificial Intelligence Chatbots look like their programmers?
Friday, September 29th, 2006
Do pets eventually resemble their owners? Or do owners get to look like their pets? It’s heck of a conundrum – but one we might now be a little closer to solving. For the past fortnight it’s been hard to escape the animated faces of “Joan”, or “George” the graphical representations of what we’re told is a new breakthrough in Artificial Intelligence. TV and newspapers, both highbrow and lowbrow, have flocked to report on the chatterbot. You can talk to Joan (or George) – the output of the British software project Jabberwacky – and think it’s human!
Er, almost.
Addicted to antitrust, Microsoft outlines 12-Step Recovery
Friday, July 21st, 2006Antitrust addict Microsoft has outlined a 12-Step Recovery Program, which it says will help prevent it from lapsing back into anti-competitive practices in the future.
The declaration follows three major “interventions” in fifteen years. A 1991 investigation by the Federal Trade Commission resulted in a Consent Decree signed in 1995. A 1997 investigation by the Department of Justice, joined by a number of US states the following year, resulted in a conviction and settlement in 2002. And just last month, the EU rejected Microsoft’s claim that it was complying with a 2004 antitrust settlement.
(more…)
Meet the Jefferson of 'Web 2.0'
Wednesday, March 29th, 2006
If Google’s PageRank reflects the “uniquely democratic nature of the web” – and if weblogs are the most empowering technology of our age – then how can we begin to fete a humble entrepreneur based in St Paul, MN?
Very probably as the Gutenberg of the digital age. And the Jefferson. All rolled into one.
Brian Adams of Blue Diamond Enterprises has announced the newest tool that leverages both weblogs and the “collective intelligence” of Google’s algorithms. His new software, Blog Mass Installer, claims it can create 100 Blogger weblogs on your website in just 24 minutes.
It’s like voting – but voting done properly: early and often.
The idea behind tools such as this is to create a network of sites to host contextual advertisements and also to boost the prominence of material in Google’s search index. It’s only the tip of a vast twilight industry that, by the calculation of SEOs (search engine optimizers) like Adams, results in one third of Google’s index being comprised of machine-generated sites. Blog Software Installer takes a lot of the drudgery out of creating the blog network you need to pimp your reputation, plug your wares, or simply earn yourself a little extra Adsense cash. Some manual intervention is needed, according to the press release -
“Creating the blogs is easy and fun. You will get a friendly chime when it is time to enter the ‘captcha’ word verification. The BMI [Blog Mass Installer] tool also gives you a status indicator to know how many of your blogs have been created.”
BMI also gives you those all-important RSS feeds.
Need some content? Just dial up some PLAs, or ‘Private Label Articles.’ Sites like GoArticles and Article City. Or if you’re in a hurry, scrape some Wikipedia content: the keyword-rich online “encyclopedia” is a favorite with SEOs.
It’s all very Web 2.0. The power of the “Long Tail” put into the hands of the little guy – who needs only $197 to join the digital revolution.
But it’s also in breach of Google’s own Adsense program, which states that no Adsense ad may appear on a page “published specifically for the purpose of showing ads, whether or not the page is relevant”.
Um, now didn’t this – we asked Adams – leave Google in the delicate position of throttling its own cash cow?
We only asked, because the Blogger-accelerator was being promoted by – of all things – Google News today. Google owns Blogger.
Adams, who says Google has done a good job weeding out spam sites over the last year. It employs more human operators to identify these, he says,
“Google’s search index is more relevant than a year ago. It’s getting better.”
But basically it’s a Machine vs Machine war. Machines like Adams’ BMI create the blogs, and Google’s algorithms try and delete them.
“It’s getting harder to tell if a website was made by a machine or a human,” he says. “There are some really grey areas.”
(We’d noticed.)
Adams says BMI is more “stable” than rival tools – “stable” means Google is less likely to find it and delete it from its index.
Didn’t he feel morally responsible for bespoiling the utopian meadow of the World Wide Web, we wondered?
Here Adams takes issue with the suggestion that machine-generated axiomatically means junk.
“I wouldn’t say that the tools are just polluting it. It’s the responsibility of the webmaster to put up content that’s actually useful. If they don’t do that, Google will delete them.”
So it’s like the argument that guns don’t kill people – people kill people?
“That’s a good comparison,” he agrees.
So Splogs don’t kill people. And are less harmful than blogs.
Bootnotes
- Thanks to Namebase’s Daniel Brandt for spotting this – and for the coining the neologlism “Goobage”.
-
“The blog might do more for the emancipation of women than the invention of the birth control pill almost 50 years ago,” – Sylvia Paull, who hosts the Berkeley Cybersalon.
'Lightweight, high-velocity and very connected'
Friday, March 10th, 2006At ZDNet, it’s Microsoft’s “Pearl Harbor”! Forbes screams, “Google’s office invasion is on!”
Only it isn’t – and we have the founder’s word for it.
As we reported yesterday, Google has paid an undisclosed sum for a web-based document editor, Writely. It’s a product that seems as mature as the company which produced it, Upstartle.
Explaining why she decided to sell the company, whose only product has been in a limited, closed beta for just six months, co-founder Claudia Carpenter wrote -
“We like lava lamps and they’re pretty much standard decor at Google.”
Moving onto the vision thing, Carpenter explained -
“Writely is like a caterpillar that we hope to make into a beautiful butterfly at Google!”
(No blonde jokes, please.)
A measure of how mature the software is can also be gleaned from this blog post. Writely gained the feature “delete from trash” five weeks ago, a lower priority for the team than “new toolbar”. When the ability to remove your own work from a hosted web service is considered less important than cosmetics, you have a fair idea of the software designers’ values.
So far, so very “Web 2.0″.
That’s because of the kind of work people are doing now, which co-founder Sam Schillace explained to NPR recently, is -
“Lightweight, high-velocity and very connected.”
Or did he mean the people behind it are lightweight, high-velocity and very connected?
To be fair, Schillace is an experienced developer who created what later became Claris Home Page, before going on to lead teams at Intuit and Macromedia. And Schillace correctly denies what the headlines writers want to believe today – that Writely is a replacement for ‘fat client’ word processors.
But these are bubble days, and it’s discordant to hear a rational explanation – but one comes from Joe Wilcox at Jupiter Research. The Writely feature set is so poor, he points out, that Google bought the software solely to beef up its editing facilities in Gmail and Blogger.
Read more at El Reg.
'Take out a subscription to The Register. Then cancel it, and sign it Disgusted Wikipedian'
Friday, December 23rd, 2005“He who feels punctured must have been a bubble – Lao Tsu
A funny thing happened last week. Author and broadcaster – and veteran OpenOffice user – Andrew Brown wrote a piece in The Guardian a fortnight ago demolishing some of the more absurd myths around open source software projects. Frustrating bugs went unfixed for years, he noted, giving lie to the myth that simply because anyone could, in theory, make improvements, then improvements that users care most about would actually be made. Brown has written two books using OpenOffice, and performed his duty as a diligent user. If this was commercial software, he’d be a MVC, or “Most Valuable Customer”, and if OpenOffice was an airline, he’d be bumped up to First Class every time he showed up at the airport.
But what was particularly interesting was the range of responses to this critique, because they mirrored the responses received by The Register from Wikipedians. I have a theory about why these are similar, but first let’s see what people said about Brown’s piece. He published them on his blog here and here.
Here’s the typical response:
out?”
As we observed with Wikipedia, passing off the responsibility onto the user for dealing with the inadequacies of the software, or information, is a trait open projects seem to share.
Then there’s the age-old response that a deficiency is a FeatureNotABug.
‘spaces typed at the end of a line won’t show’ How is this a bug? It’s just a different way of displaying text. Is a printer in error because it doesn’t visually show you there is a space at the end of each line? No. There’s no reason why it should have to show a space at the end of the line. That’s you being very pernickety, not a bug.
Noel Slevin
Silly Mr. Brown, for not spotting that. More accurately, this response is classified as “Blaming The User For Being Stupid”. Again, that’s a Wikipedian trait too, and there were plenty more in the same vein.
Note the subtle variations. There’s the “Hypothetical Utopia” defense, which ignores the present for an imaginary future in which the FOSS processes work as they ideally should:
So, yes, there is a problem with the open-source model. But I wonder whether things will change if OO is adopted by cities that have skilled IT departments that can be directed to fix THOSE PARTICULAR bugs, or to make THOSE PARTICULAR enhancements, that are of importance to THAT PARTICULAR city? I can imagine city council directing the IT representative to get the bug fixed and to report back at the next meeting. Within a couple of meetings, either the bug will be fixed or the city will drop OO. This is a tight feedback loop that involves skilled workers.
Then there’s the “Never Mind the Quality, Feel The Price”.
[paraphrased] Any bugs in OpenOffice are counter balanced by the fact that it is free!
And that’s one of the commonest defenses of Wikipedia, which imagines a world in which the population is so starved of information (books and libraries don’t exist here, for example, nor do wise teachers), that every globbet of information that drips from a computer network must be applauded as an “information revolution”. In this world, the speed or price of information trumps all considerations of its quality. But as is so often pointed out, we’re hardly living in a world starved of information. We’re drowning under vast quantities of ropey information, and none the wiser for the experience.
Back to the onslaught on The Open Office User Who Dared Complain.
There’s the parry called “Flood The Area with Improbably Large Numbers”, in which downloads (or in Wikipedia’s case, the number of articles) are quoted. We shall spare you this.
But a significant proportion of responses take the counterattack, and question the critic’s motives, knowledge and quite possibly, moral inadequacies too.
Darryl LeCount’s lofty ticking off is typical:
I found Andrew Brown’s vitriolic attack on OpenOffice.org to be ill-informed, heavily biased against open source software and extremely inconsistent. He claims to “like” OpenOffice, initially using it out of “a mixture of perversity, stinginess, and vague anti-Microsoft sentiment”, before launching into a tirade about how buggy it is and how flawed the open source model is. The author has clearly neither had extensive experience of using Mozilla Firefox, Blender, or Linux, and it is also clear that he has had little involvement with the development of these products despite his vague claims.
So Mr. Brown’s critique of one product is invalidated because he hasn’t used enough of them. A snobbish variation on “user is stupid”.
Finally, there’s the kind of response which supposes that the only reason a critique was made was to drive up page traffic.
I think the author of the article has achieved exactly what he intended to do and that is generate traffic to his blog and article. If you were a good objective writer you would not need to resort to this tactic. It’s a bit pathetic that you feel the need to be so negative at the expense of something you get for free. Let’s face it, this article could just as easily have been positive but that just would not have generated the traffic right
![]()
We hate to see a sad face, at this time of year. But we also get the feeling that advocates of this, the Page View Whore counterattack, rarely meet advocates of the Flood The Area With Improbably Large Numbers counterattack, because if the project was as popular as the latter insist, then publishers would write only write nice things about open projects, to drive up their traffic. We’ll spare you the rest, but the entire defense is summed up at the end of a tedious “Fisking” delivered by one Dave Lister, who sums up Brown’s arguments bafflingly, so:
“I like OpenOffice.” translation: I really want Open Source to get better
![]()
Silly Andrew, for harboring such hopes. So what are we dealing with, here?
Well, in his Guardian piece, Andrew Brown pointed out that successful open source projects keep their users happy, and if the users share the same background, common goals, and level of technical knowledge as the authors, then the users can indeed contribute to a virtuous circle. bind and Apache spring to mind.
But when the skills and experience are, to steal a Rumsfeld-ism, “asymmetric”, there’s friction. Many of Andrew Brown’s OpenOffice critics have no idea of what a user really wants to do with the software, and can only cognize he’s rejecting their gift of free software. Many Wikipedia defenders have no sympathy for readers who complain about unreliable, or badly written information, and can only cognize a world mocking their careful handiwork, what one critic calls a “defective data device” with “-pedia” in the name.
(One Australian doctor wrote to describe how he’d made just one Wikipedia edit in his life, to correct an entry about a medical procedure, which if carried out, would result in death. Heck, this is an information revolution, and every revolution is going to have casualties!)
So perhaps it isn’t such a mystery. Open projects are by nature idealistic, a little gift to the world. When this gift is spurned, the rejection must feel terrible.
Why would an ungrateful world reject this gift?
Let’s find out.
Read more at El Reg.
The Internet Services Puddle
Thursday, November 10th, 2005
Ever the master of public relations, Microsoft has always been able to figure its way out of a tight spot with the use of a judiciously leaked memo.
Remember when AOL merged with Netscape back in 1998? Time to take a leak. Remember 2000, when Symbian was stealing the thunder from Microsoft’s cellphone strategy? Time to take a leak. Remember when the antitrust settlement talks had hit a sticky patch? Time to take a leak. Remember when Microsoft’s security woes finally became an issue? Time, once again, to take a leak.
The purpose of these releases is to bolster morale and focus the staff – Microsoft always seems to need a No.1 Enemy – and inform the press that it’s on the case.
(The memos Microsoft doesn’t want you to read such as this one and these two, are always more entertaining and enlightening.)
And so it goes. We know you’re very busy people, so in the spirit of the excellent 500-word “digested reads” offered by some of our better newspapers, we give you the précis of the latest Gates and Ozzie memos.
Web 2.0: It's … like your brain on LSD!
Friday, October 21st, 2005.
