I'm doing research for a book about Google, and blogging some of my thoughts at http://richardbrandt.blogs.com/. The problem is that after two years of researching the company, I've come to the conclusion that Sergey and Larry are basically nice guys and are in fact trying to do no evil. I bought stock in the IPO, and I think the company has a much bigger future than its present implies.

I'm used to being highly skeptical of companies (although I was accused of having too much faith in Microsoft when I was covering the company for Business Week in the late 1980s.) People I respect have accused me of being too soft on Google.

The biggest criticism I have about Google is its paranoid secrecy. It's the cause of all the suspicion and dislike of the company. No, its products are not yet perfect, but give it time.

So I'd like your opinions. What do you like or dislike about Google? Is it doing great things or is it the new Evil Empire? Please post any comments here or send me email at RLBrandtgmail.com.

'I bought stock in the IPO'

So you are a very rich man now?

No, I couldn't afford to put in much money. But it's been a nice profit and I get to go to shareholder meetings. I'd buy more if I could afford it, even at its current valuation. I'm looking forward to the stock price doubling in another year.

I like Google, but the biggest problems there are IMO Chinese censorship and privacy issues. I know that Google doesn't want to abuse our privacy right now, but they know abut us almost everything, and they keep this data for ever. And you never know what will Google (or US government) do with this information in 20 years. Companies turn evil when they starts to have financial problems.

See also "Google Love/ Hate Relationships":
http://blogoscoped.com/archive/2006-03-14-n21.html
http://blogoscoped.com/forum/23355.html

Like:
Suport for Firefox
Simple layout
Wide range of web services/searches

Dislike:
Chinese Censorship
Amount of information they have about me
(http://www.google.com/notebook/public/02921283618201356110/BDRRESwoQuZ3mrbkh)

Philipp – Thanks for the Icon

1) Google needs to open its doors onprivacy and let the users know what their data rentention policies are. We trust Google so far (at least in the North Americas). However, we have seen other getting thrown under the bus in certain ccTLD zones. This could very well happen in north america too

2) customer SErvice- yikes.. their Auto responders are an irratant. Can't they just have somebody who responds to to valid question and issues.

3) Larry/Serge were one of the first C level folks in the industry which 'fesesd up to the fact that they may have gone down the wrong pathway in china. Accepted --- but they still need to advocate what they will do to ensure that this does not happen again. Just opposing the DOJ case is not enough. Every user in every country needs to feel happy using google. Right now, I really dont think that Google has a firm up strategy to ensure that all users data will be treated in "fair process". That is at least inform the user that ABC government or org is asking about their data. Trust given -should be trust returned.

Like :

Different.
It's simple.
It's layout is easy going on the eye.
Supports Open-Source
Like a Yahoo Employee once said Restores Competition in areas that have been abandoned (like Calendar and Web Based Email)

Dislike :

Secrey. (About Current and Future Products) Aah it's driving me nuts
It's not too aggressive in marketing its products (orkut anyone)
Lacks Integration between services (however it has improved recently)
Doesn't really create any brand new "Web 2.0" thing.
The Privacy stuff.

> Doesn't really create any brand new "Web 2.0" thing.

Gmail with it's AJAX technology was the first Web 2.0 application.

Privacy, they should be tellling us exactly what information they have about us. Better yet they should delete it.

>> Gmail with it's AJAX technology was the first Web 2.0 application.

That's not true. If there ever was such thing as Web 2.0, it was Google Maps.

For me, I like the way Google employees spend 20% of their time creating new kinds of applications, and have managed to create such a varied collection of products over just a couple of years. And that's only the beginning.

The thing I hate is that Google don't push these products hard enough. There is heavy competition out there, with for example Google Maps, and they can't rely on word-of-mouth alone, even though it worked for their web search.

Google is undemocratic! In Germany where Google has 90% market share it is very dangerous. You have one closed entity that determines what informatios will be found and what not. I talk about censorship but also about punishments for certain "SEO techniques" that are not even disclosed officially as harmful. There is no way the people that are harmed by not being listed at Google can really take part in the desicion making process.

The biggest German Blog, the Bildblog almost dropped out of Google some weeks ago after the algorithm was altered. Google often hurts decent sites.

We need an open source search engine that is controlled by the people.

What I do not understand is the complaints that Google has too much info:
Hello? YOU GAVE IT TO THEM!

Google is undemocratic? Maybe its just the best? The people DO control which search engine wins, they could all use Yahoo if they wanted. They also to an extent control rankings.

open source search engine?

open source for spammers too

Dislike

Lack of support for Opera in all their apps that I would like to use : GMail, Calendar, Analytics, and now Writely. This sucks donkey balls and, what's more, it seems they not only won't ever do anything about it, they simply don't care. I understand it from a practical point of view but It makes me feel they don't want me.

Love

The search. And honeslty it outweights everything else because it's the only thing that I'd miss. I use my Gmail account through a POP client so I don't even know what goes on on gmail.com (though it saved my ass to find all those archives in a recent hard drive crash; thanks Big Brother). I can't be bothered to switch to Firefox to use Calendar or Writely, so I can't get addicted to that stuff either.
But the search!

About privacy and monopoly issues I tend to agree with Bob Jones above.
And about the company I know next to nothing so...

Dislike: Customer Support. Nearly impossible to get 'good' tech support, even for (as an example) webmasters who have thousands of adwords clicks each day (not me, but as an example). I believe this is completely unacceptable for a company as large as or as profitable as Google.

Google seems to be 'trust us'. As of right now, I mostly 'trust' Google, but that is based off their past practices. Which is not a guarantee of future trustworthiness.

Minor annoyance: What's with the really long Betas?

Like: They are very focused on providing excellent services for users. Most of their products are excellent. Probably the largest US company who generally puts customers before profits.

Huge supporter of open source software. Very few companies, in my opinion, are not only supporters but are as dedicated to open source as much as Google as.

Willing to make jokes. I value the ability to lighten up and make jokes greatly in companies (and people). I would say the day that Google no longer puts up a good Fool's Day joke is the day Google becomes the (or a) evil empire. Until then, I wouldn't panic.

The only true challenge to Microsoft. About time one appeared.

Notes:
As for issues involving China, I don't like what Google has done. However, I am not too angry at Google since they are only doing censoring as required by the Chinese government and is providing their services to China for the benefit of Chinese users (unlike companies such as Yahoo and Cisco, who are going beyond what is required and, I believe, clearly in China for the money). I personally believe that entering the Chinese market was Google's only true choice (to be more precise, not entering the Chinese market wasn't a real option), though being the only choice does not make it a moral one.

I think the click fraud issue is overblown. Google is probably the best company to handle the situation (especially if they improve their customer support), and I fail to see how any type of ad service doesn't suffer from click fraud.

Overall:
Google isn't evil, and many companies could use to take notes from them.

Sébastien Guillon
Opera's Javascript is very spotty. I am guessing that for most Google Javascript-heavy pages, it would require much more time to debug the page for Opera than for both IE and Firefox.

What pisses me off about Google:

-- Lousy customer service. I mean HORRIFIC. This has been demonstrated over and over agin in how they handle inquiries to Google Answers. Customers should NEVER be given only auto-responders that don't answer their concerns. An auto-responder that says "We got your e-mail, Bob the Customer Service Guy will get you taken care of shortly" is fine, but it needs to be followed up with real communication or customers will take their money and go elsewhere. (Clearly, Googlers have never worked service, or they wouldn't need to be told this. Pay attention, Googlers. You might be smart, but that doesn't mean you don't still need to learn things.)

-- Refusal to market Google Answers. Why the hell did y'all go recruit really bright, driven people to form this pool of researching talent...and then drop them like yesterday's used toilet paper and steadfastly refuse to tell them why? It's a good and useful service, give it a little press instead of burying it. Stop treating us like the red-headed step-child.

-- Paranoid secrecy and refusal to communicate, even with people who perform services for them. When someone performing a service for you has a valid question, the answer is NOT "We're just not going to share that information with you". Play it straight instead of all this stupid cloak and dagger crap.

Things I really dig about Google:

-- I can go from zero information to oh-my-god-I'm-drowning-in-it in less than three seconds, every single time. Google search is the best I've used, ever. I even use the calculator function to scale recipes.

-- Google Calender keeps my ridiculously busy life organized.

-- Google supports women in technology and encourges women and girls to get involved in the industry.

-- YAY FIREFOX SUPPORT! IE makes me want to spork my eyeballs out, so I'm thrilled that Google plays so nicely with Firefox, it makes my work much easier.

-- GMail's searchability. I can't count how many times that has kept me out of hot water.

Dislike : 2006 April Fools Day Joke was way lame

Just to clear it up, when I made the Web 2.0 comment. I meant something entirely entirely new. The Concept Of Gmail (or Web Based Email) wasn't new.

My comment meant Something entirely different. Something that no one has created.

Richard, isn't there some journalistic conflict of interest in writing a book about Google suggesting (by implication or more) that you think it's undervalued, whilst being a shareholder?

Oh, and just because their paranoid secrecy is the cause of suspicion and dislike, doesn't mean they don't have something to hide ;)

Anyway, to avoid repeating comments about the big issues, I'll mention a specific dislike: Google's widespread use of visual CAPTCHAs.

They are evil http://www.w3.org/TR/turingtest/ :-

"A common method of limiting access to services made available over the Web is visual verification of a bitmapped image. This presents a major problem to users who are blind, have low vision, or have a learning disability such as dyslexia. This document examines a number of potential solutions that allow systems to test for human users while preserving access by users with disabilities. [...]"

And stupid :-

"External projects such as [BREAKING],[AICAPTCHA]and[PWNTCHA] have shown methodologies and results indicating that many of the systems can be defeated by computers with between 88% and 100% accuracy, using optical character recognition. [BREAKINGOCR] outlines a CAPTCHA defeat on PHP- and ASP-based systems, in which known-valid session IDs are cached and reused to circumvent several popular CAPTCHA schemes. The "Screen Scraper" attack reported by the Anti-Phishing Working Group [ANTIPHISHING] defeats the [PINGUARD] technique by capturing the screen when the user clicks to enter their secret code.

It is a logical fallacy, then, to hail CAPTCHA as a spam-busting panacea. Even 10% accuracy by a computer amounts to system failure, just at a slower rate. It is also faulty logic to believe that the adoption of CAPTCHA in large sites is evidence of its supremacy in fighting spam. Indeed, a number of techniques are as effective as CAPTCHA, without causing the human interaction step that causes usability and accessibility issues. [...]"

Google has known that for *at least* a year http://www.internetnews.com/ent-news/print.php/3518741 :-

"But captchas are gotchas for the blind.

Many vision-impaired computer users employ screen readers, software applications that convert text and graphics – as long as the graphics have descriptive "alt" tags – into audio. The speech simulator can read menus and the names or descriptions of navigational elements such as buttons and links. But screen readers are stymied by captchas.

Google is working on it, said Marissa Mayer, director of consumer products. "We are planning on releasing some alternatives in the next one to two months that make our current captchas more compatible with screen readers, and we're looking into audio captchas," Mayer said."

Yet their supposed "working" and "planning" has come to naught, not in "one or two months", and not in more than a year. (Maybe that's another dislike: Google are already a supertanker, unable to change course quickly).

And take a look at this lulu: http://www.google.com/addurl/?continue=/addurl . The CAPTCHA image is “To help us distinguish between sites submitted by individuals and those automatically entered by software robots, please type the squiggly letters shown here into the box below”.

But the ALT text says “If you can read this, you do not have images enabled. No problem; just leave the textbox below empty”.

So Google is seemingly content (if only by neglect) to consider submissions from sight-impaired individuals to be as suspect as those from spambots?

Like:
Quick load times
One account for multiple services
Slick Design
Easy learning curve

Dislike:
Nothing

Don't Care:
Chinese censorship; what were they gonna do? Say, "OK, we'll just let some other search engine take our market in the largest country in the world"? No way.
How much info they have; they've already stood up to the gov't, and I have nothing to hide. Do you?

I was scrolling down rapidly, to write what Missy so aptly already posted.

PROMOTE Google Answers! Hands down, it is the best answering service on the net.

http://answers.google.com

I don't dislike google at all. But the one thing that leaves a bad taste in my mouth is their participation in censoring in China.

It's wrong, they know it (despite their b.s. trying to rationalize it), yet the participate in oppressing people.

Maintaining your market share is not an excuse for doing anything that you want. Jeezus, Busterbot. Have you no sense of right and wrong? Apparently not.

Customer service. They have none. Simple as that. Get a phone. Hire a few hundred people to answer said phone. Problem solved.

Like:
1) Google services are easy to use, user friendly, fast and modern.
2) They support open source projects.
3) Google search brings me the best search results.

Google is neraly a perfect company but this is what I dislike about them:
1) Google services are extremely disorganized. They own thousands of domain names, the localized versions of Google's interface are never up-to-date. Some services are are really hidden, so people usually don't find them unless they know their names.
2) Google pages are not web standards compliant.

Like:
1) Google "never ever gets in the way of the user". This is their main design philosophy for web pages.
2) Seem to come up with a lot of new ideas (both technical & business).
3) Giving employees the freedom to spend 20% of their time on R&D.
4) Their support for open-source s/w.
Dislike:
1) Google Talk: It has been a huge flop and it doesn't look like they want to improve the app & make it compete with MSN & Yahoo.

Like Google!

I'am a chinese!There are a lot of Google fan like me in China! Google is good and I use it in the daily grind.

especially, Gmail give me a good feel when I surfing in the net!

but sometimes Google and Gmail can't work normally .The stability of Google in China is not good!

   eflyingxpgmail.com
[Signature URLs removed. -Philipp]

Buster, Russia is the largest country unless you're talking about Population. USA is in Online Population.

They start software barren. (Release early, Release often). Then build on it based on user requests (with the majority). This way most of what they build in is useful/based on function and will satisfy a need(time not wasted).

They have exploited very well(as many others have) distribution of applications over the internet. Where in the 80s software was basically shipped out on disk or with the computer and was much tougher to get into peoples hands. Nothing to install. Bug fixes happen on their servers (no need to push out patches).

Alot of their business model is automated. They code it and let it run (adsense/adwords). Upkeep costs much less. And there is no actual commodity to sell like dell which will incur cost so they can reap profit from the same/similar algorithm time and time again. (Checkout is an exception here where they have to deal with fraud/end users directly , along with the click fraud department).

They hit a sweet spot after the bust where they could maximize their infrastructure and have freedom to deploy all sorts of apps. Broadband became cheap and more ubiquitous in the US(15 dollars a month for DSL from verizon. 100KB down. Hard to not get) so users can handle more bandwidth intensive apps. Disk space became much cheaper the last few years (gmail and all sorts of redunancy with gfs). I remember i bought a 1 gig drive for 100 dollars in 98. Now you get a 300 gig drive for 100 dollars. Purchasing unlit fiber for really cheap and using data centers that were needing their business.

Small teams 5 to 10 people, even if they are competing on producing the same product separately won't have the attrition of huge employee bases (Recent MS Reorganization?)

Saving millions of dollars in licensing fees using open source and developing their own versions inhouse (gfs/mapreduce/bigtable/scheduler/email). Then they push it out to the users if they find out what they use in-house would be useful(Gmail)!
They are dabbling in hardware now (maybe power saving asic's for their servers)

Hopefully their wealth will rise enough to start doing more "blue-sky research" and work with the universities to become a research institution:

Rough explanation of why this type of research is less common now: http://arstechnica.com/news.ars/post/20060724-7340.html

Dislike:
The Google Print/Books scenario. The opt-out b.s.

What I like about Google:
- Good search results
- Great email client
- Overall great and pragmatic usability/ interface design (this includes the lack of clutter and the focus on the "tool" side of their apps). Windows Live for example I consider to be creative but not pragmatic in terms of interface.
- Smart and creative engineers
- Apps are constantly improved
- Their tools behave much better than those of MSN or Yahoo (e.g. they link to competitors tools, they don't try to push their own products if these aren't relevant to the task at hand – compare to the Yahoo search for [google.com]!)
- They don't announce years in advance, they just release. Everyone hates to wait, but everyone loves a surprises... Microsoft fares much worse in this regards.

What I dislike about Google:
- They compromised their mission, so it's much harder to trust them when they claim something in the future
- Many new services don't scale with the demand and get slow or crash. Google Calendar is slow for me till this day. I quit Google's Blogger a while ago because it was constantly slow.
- Some Google services aren't very innovative, like Picasa Web Albums.
- Google's "Beta", "Test" or "Labs" symbols, which mostly lost any meaning (some Labs products work perfectly, some non-Beta products have glitches, other products stay in eternal Beta)
- They often don't reply to emails
- They are often too secretive about anything
- Their blog reads too press-release-ish
- The way they structure their blog network (it's hard to "trust" a new subdomain on blogspot as everyone can register them, and they don't have a meta-blog covering all posts)

<< The way they structure their blog network (it's hard to "trust" a new subdomain on blogspot as everyone can register them, and they don't have a meta-blog covering all posts)

::cough::

Why are you coughing? :)

hmmmmmmmm..maybe Google will *now* send some cough syrup to Ionut :)-

Good stuff, keep it coming. Milly, check out
https://www.google.com/accounts/SmsMailSignup1

Gmail now allows audio captchas (started in April, I think?). The captcha service at Google is code that programmers invoke to get service. So new/important services should be able to use the audio captcha going forward without much/any extra code. The bad news is that existing Google services have to tweak their existing code to allow for the audio captcha. For something like the addurl form where it's nice-to-have but not required, I don't know if they'll go back and add the audio captcha. But new services at Google can offer an audio captcha, which is good.

As far as the meta-issue of captchas, I'd agree that they're not perfect, but they're still quite helpful to people that run a service for humans, not bots.

>>>And take a look at this lulu: google.com/addurl/?continue=/a ... . The CAPTCHA image is “To help us distinguish between sites submitted by individuals and those automatically entered by software robots, please type the squiggly letters shown here into the box below”.

But the ALT text says “If you can read this, you do not have images enabled. No problem; just leave the textbox below empty”.>>>

wow.. i never knew that.. it works regardless if you use the captcha or not..

yeah blogger had audio captcha's too.. but the squiggly letters shown here into the box are just too squiggly..has anyone in google ever conducted a UX'er test to see hwo readable these random squiggly char's are -before they were rolled out of production.. IMHO, that would be a good thing before Goog's goes all out with the captcha strategy..

yeah and hopfully they will have std [squiggly's] across all services :)

Google's captchas are often hard to read, and always annoying to enter. Isn't it possible for Google to modify the Blogger comment form code everyday, based on some semi-automated templating system, to avoid captchas? After all, this manually created forum doesn't have a huge spam problem even without captchas, and the one difference I could imagine is that spammers don't target "uncommon" comment forms as much. So maybe the big systems need to present an "uncommon" comment form everyday...

I've got an idea:

most spam comments are repeated
a duplicate content checker would help

I dont think so Ionut, I have seen vectors where the spam content has been parsed together from sets of random sentences and words

The chances that the content will be 'duplicated' is slim...

>>So maybe the big systems need to present an "uncommon" comment form everyday...

But that would completely destroy usability and user expectations of a consistent interface. What I do find interesting is what some online banking sites have switched to: You have a number pad with corresponding letters on it. The letters are different everytime, and your actual password is a number so you type in the corresponding letters for each digit.

It seems like adding some kind of clicking (also an idea in use by some online banks) would also add a layer of complexity. The problem is none of these are truly foolproof, and all add burden on the user.

Another idea:

most spam comments are irrelevant to the topic of the post or the previous comments
if the intersection of topics is (almost) void, the comment is spam

That would also remove off-topic comments and brilliant comments like "nice" or "lol".

Wow, some great stuff here, folks. There seem to be a couple themes running here.

Google tech support sucks. The company really has to start learning to interact with its customers better.

I'm not sure what alternatives there are to CAPTCHA. Google has announced that it is working on new technology for the blind.

It has a terrible tendency to be really slow to upgrade existing products. It sounds as though it's giving up on some of them without informing people what's going on. The old communications problem.

As for Chinese censorship, I think the comment from Eric Shang is telling: "I'am a chinese!There are a lot of Google fan like me in China! Google is good and I use it in the daily grind... but sometimes Google and Gmail can't work normally .The stability of Google in China is not good!"

The message Google fails to get across:

Ror the last nine years it has had a Chinese language version running offshore. Google does not censor it. But the Chinese government censors it through its firewall. That makes it slow and unreliable, probably what Mr. Shang is experiencing.

That problem has kept Google's market share in China down. So they decided to go onshore, where they have to comply with China's censorship rules, in order to make it faster and more reliable. Every search engine in China does the same thing. I had lunch with Eric Schmidt a couple months ago, and he said, "If we do not obey the laws in China, our people there will be arrested and tortured. I have a problem with that."

So what's the solution? Not go onshore at all? That doesn't help anything. Even though Google is offering a new search engine in China it STILL keeps the offshore version running! Over 90 percent of Chinese use the offshore version. So what is lost by going to China? Now there are two choices, one censored by the Chinese and one censored internally, with the hope that any new access to information will help open up China.

> But that would completely destroy usability and
> user expectations of a consistent interface.

No no, I meant that the frontend would stay exactly the same. Only things in the backend would change. Example: on day 1 the crucial text input form is named "foobar" and on day 2 it's named "barfoo", and on day 1 it's the first input box (in the source) and on day 2 it's the second input, and on day 1 a hidden input field of the name "bla44" is required to have the value "4004" and so on. The layout can use CSS to hide certain stuff so it won't appear in the frontend... but the bot will only see the backend. But probably this is too complicated, or impossible to program and has several implicatios on certain browsers... and probably, spammers could just use some tool that simulates (visual) user input patterns (click here, enter there, etc.).

did anyone catch this article ??

http://money.cnn.com/magazines/fsb/fsb_archive/2006/09/01/8384907/index.htm?source=yahoo_quote

"Keiter had fallen victim to what's known as the dreaded Google dance: sudden, seismic shifts in search results that occur whenever Google's engineers decide, without warning or explanation, to tweak the software algorithms that determine how the mighty search engine processes keywords."

Ah, Philipp...I see what you meant now.

It seems like one of the "big search problems" that remains very difficult is to search images for an object. But humans can do it quite well and easily. Therefore, I think these will become the next popular captcha, since text is broken fairly easily and the current captchas are images, but really just text.

The only problem would be people describing objects in different ways. But this would also let you do something like have a prompt that says: "Describe the shape below." and show a picture of a triangle. The next time it might say "Name the animal pictured." and show a picture of a monkey. This would be much more difficult to defeat through software.

Of course, generating all of those prompts and also being lenient on the user's response would prove challenging and time-consuming.

I think Niraj is onto something.. name the picture captcha idea is nice..!!

have an image [monkey] and buttons which sez [monkey, donkey, cat, dog, whatever] and the user just points and choose the right verbiage for that image.

Thats fun way for captchas implementation...

Oh, and I forgot to address Missy's issue about conflict of interest. In my entire career as a journalist, I have never bought stock in a company in order to avoid a conflict of interest. Now that I'm writing books, I couldn't resist the urge to participate in the IPO. My experience there will be in the book. And I will disclose prominently in the book that I bought stock, so people can decide on their own whether to take my opinions with a shaker full of salt.

And as for books, I think the publishing industry has it wrong. Google will only publish excerpts from books still under copyright. That will probably serve to increase sales. Who wants to read an entire book online anyway? If I can find a publisher who will let me use the Creative Commons
http://creativecommons.org/about/licenses icense from Lawrence Lessig at Stanford I will do so. I'm not afraid of online versions or sharing my stuff with others.

You should also look at Lessig's talk http://www.oreillynet.com/pub/a/policy/2002/08/15/lessig.html
   about how ridiculous copyrights have become. Did you know the tune "Happy Birthday" is still under copyright?

"If I can find a publisher who will let me use the Creative Commons" – Huh ??

Oh well, you can self publish on Lulu.com under CC!!

I've got a publisher. I'd just like to find one that would let me use the Creative Common license instead of the overly-restrictive copyrights.

Oh, thanks Matt. And sorry, clearly I was too harsh with the "come to naught" bit. (I didn't notice that the old accessibility.gif is now sometimes connected to a new JavaScript 'play sound' link: but those using screen readers would have realised, because it has new ALT text too).

Though ... whilst an audio alternative for sight-impaired users is an improvement, it's not much use for those with both sight and hearing impairments. And as the W3C say http://www.w3.org/TR/turingtest/#sound :-

"However, according to a CNet article [NEWSCOM], Hotmail's sound output, which is itself distorted to avoid the same programmatic abuse, was unintelligible to all four test subjects, all of whom had "good hearing". Users who are deaf-blind, don't have or use a sound card, work in noisy environments, or don't have required sound plugins are likewise left in the lurch. Since this content is auditory in nature, users often have to write down the code before entering it, which is very inconvenient. Worst of all, some implementations of this technique are JavaScript-based, or designed in such a way that some blind users may not be able to access them. Machines, on the other hand, may even have greater success with voice recognition software than they do with OCR on visual CAPTCHAs."

As for CAPTCHAs generally, I appreciate their utility, but it's the current fashion for *visual* (or audio/visual) CAPTCHAs which suck (and not just for impaired users, as this thread shows). Why not use logic puzzles, which merely aim at human brains, rather than human senses? And then consider raising the bar if, and only if, the bots actually (not theoretically) defeat them?

Hey, you could start with your blog (which has no audio alternative), but it would be nice to see Google tackling this issue in an accessible, scalable, innovative, industry-leading way ...

Niraj, the misdescription of images by users wouldn't be the only problem: what of users who can't see the pictures (because of technological or human impairment)?

Why not have it in plain (screen reader friendly) text, saying "Monkey, triangle, dog, cat: type in which is not an animal?" Or "Oblong, square, dolphin, triangle: type in which is not a shape"? Or "Type in which of these is a colour: book, yellow, chair, clock". Or "Which animal barks? Type in either cat or dog". Or one of a million, randomly presented alternatives?

Richard, I wouldn't want to lumber Missy with my opinions, so it was me who raised the topic of conflict of interest. I didn't doubt that you'd make full disclosure (you made it here, after all). Is that sufficient, as a journalist (and as a financial analyst, in which capacity you'll partly be writing, from what you say)? Is it less appropriate to avoid as conflict (or the appearance of a conflict) in a book than, say, a magazine article? Your IPO experience will still be there, even if you sell prior to publication, after all. I'll leave those as rhetorical questions, since it seems impudent for me to push it further (and because I don't know what the answers ought to be!)

As for the China thing, I think Google has got those messages across pretty well. It's just that many of us think they don't fly.

But your's (and Eric's, as quoted) are straw man arguments. No one seriously suggests that Google should disobey Chinese law: the choice is not obey or torture, it's obey or withdraw.

(Btw, doesn't it give you pause that doing what Google does best, essentially *being* Google – y'know all that "organize the world's information and make it universally accessible" and "do one thing really, really well" stuff – would get it's people tortured in China? Shouldn't Eric be wondering if the prospect of torture for his people doing what Google people elsewhere are proud to do, might itself be too high a price for doing that sort of business there?)

Similarly, it's simply not a given that not going onshore (or, now, withdrawing) "doesn't help anything". It helps maintain/regain Google's self-proclaimed principles and ethics. It shows principled leadership (as a search company; as an international company; as an American company). It adds to the existing international pressure on China to reduce repression. It reduces competitive pressure on other companies to collaborate with objectionable Chinese policies in order to do business there (and increases the ethical pressure on them not to do so). It draws a line between engagement with objectionable regimes (as we all are), and collaboration with objectionable regimes; between suffering censorship and becoming the censor (or at least the censor's stooge).

Now reasonable people may disagree about the right balance or compromise, and the likely effectiveness of either strategy. But the choice isn't between providing censored search or not 'helping anything', it's between providing censored search or taking a principled stand.

I hope you'll research and present both sides of the true arguments (including in favour of being onshore: there are others you didn't mention), even if you then come down on Google's side. Straw man arguments do both sides a disservice.

You asked "so what is lost by going to China?" Google's soul, I'd say. Not in one giant suck, as Jeremy Zawodny put it (apparently without irony, as an employee of a company long since without one). But in a slow corrosive death rattle. To paraphrase Philipp, it's harder to trust a company which compromised its mission in such a fundamental way.

> have an image [monkey] and buttons which sez
> [monkey, donkey, cat, dog, whatever] and the
> user just points and choose the right verbiage
> for that image.

I think the problem with that kind of Captcha is that it lets through 25% of all spam. Something similar is true for Milly's idea. That's likely to be enough for a spammer. They don't care if they hit the wrong button most of the time, as long as their spam ends up on a fairly wide amount of blog comments all in all.

I'd probably like Captchas that let me "find Fido" or something. But I guess it's easier to crack than distorted text.

I wonder what would happen if every comment you post costs you 1 cent? If you post 100 comments a month, OK, that's a dollar, but you can afford it. But a spammer may not be able to afford posting 1 billion comments a month :)

Milly, I suggested images because it's a much more difficult problem to create something automated that could get it right. With text, for instance, in some of your examples the answers could be pumped into Google Sets to determine which one did not belong. Similarly techniques could be used to determine a correct answer to a text question.

Of course accessibility is a major issue which this kind of captcha does not (and maybe cannot?) take into account.

And Philipp's absolutely right in the fact that with a limited number of choices this is not effective. That's why I originally suggested a system where the user provides a response without being provided choices. They just identify objects, and the type of object varies. In this way, you couldn't have a simple program that did some kind of object recognition, because a picture of a slice of pizza could be asking "What is the shape?" or "What is the object?" and a person would know the difference and provide 2 different answers for the same image.

I think biometric authentication is much better.

We won't need passwords, questions, tests – our fingerprints or iris will be enough to tell who we are, and implicitly that we're human. There are some privacy issues, but I'm sure a Google Biometric Account will solve that :)

<< I wonder what would happen if every comment you post costs you 1 cent? >>

Or as we would say, "Put your money where you mouth is!"

Google could do this with Blogger. They could link a micro-payments system into Google Checkout for anyone wanting to comment. (Of course, the blog owner could switch this option on or off and perhaps allow them moderate "free" comments on their own blog – even accepting some people as non-paying commentors.)

Of course, that idea's got its flaws but I think it's a good one. Why don't you try it on this forum Philipp? That would cut down on the spam and some of the pointless comments! It might also persuade people to check for duplicates before posting... ;-) Afterall, you are paying to host other people's rants – surely it's time to make us pay for it ourselves!

Ionut: I have seen both RFID and biometric (thumb access) on certain mission critical machines. Its there, but not implicitly main stream :)-

but thats included in the TCO's in certain env's!

15 Annoying Things About Google
http://blogoscoped.com/archive/2005-12-03.html

1. Search still has a long way to go
ex.: date ranges don't work well and, in any case, can't be sufficiently fine-tuned (like only pages updated in the last month); basic boolean logic needs extension. I want to find words/phrases "within" one word, one sentence, one paragraph, one page.

2. Gmail
grossly overrated dependence on search. I LIKE to categorize (file) my mail. There's no scrollbar at the bottom, so I always have to expand my Firefox window. I can't right click and open up mail in another window, so I'm constantly going backwards and forwards – which on dialup is a REAL pain. It may be hidden somewhere, but there is no easy way to collect a bunch of mail (all with subject x between certain dates) and do something with it: like delete it or file it (which, of course, one can't do because Google opposes file folders). There's no way to sort mail because, of course, with search it isn't needed. Quite frankly, I find both the current email versions of Excite and Yahoo to be easier to use – and they have search functions, too.

Don't get me wrong. I google multiple times/day. I have tried the current versions of other search engines, but google is ALWAYS my first shot.

I like the innovation – during the World Cup, it was great to discover I could enter the names of two teams and get the last results. But I think the Firefox extension writers are even more innovative. (I can't remember having so much fun with software since the 1980's – before Microsoft killed software development.)

LCaution: For search in gmail, you can do what you described. You can type into the search box:

"subject:x after:2006/2/24 before:2006/8/24"

to get all emails with subject "x" in the last 6 months. Once you do that, you can delete the results, or label them which is essentially putting the emails into folders.

While the search syntax above may be hidden, you can find it by clicking on the "Show search options" link to the right of the "Search the Web" button.

By the way, the labels are essentially folders, but better. That's because you can put the same email into multiple "folders" by applying multiple labels. If you archive the emails from your inbox after labeling them, they will only appear under their appropriate label (and under All Mail).

> I think the problem with that kind of Captcha is that it lets
> through 25% of all spam. Something similar is true for Milly's
> idea. That's likely to be enough for a spammer. They don't care
> if they hit the wrong button most of the time, [...]

Oh, I just adopted similarly basic ones as examples, but it should be easy enough to exponentially raise the bar higher than a multiple choice of four. Such as :-

"Type in the blanks: Both a ___ and a watch, tick. Bananas are ___ not blue."

"Type in the word Mississippi after removing one each of the repeated letters ___ ”

"Type in what remains after removing the capitals from pHiLIpp ___ "

"Type in the third letter of the fourth word in this sentence ___ "

"Type in the first animal and the last fruit and the happiest colour: dog, apple, cat, black, banana, monkey, yellow, grape, lizard ___ ___ ___"

"Type in only the letters missing from Gogle Blogocopd ___ ”
  
Each wrong answer could refresh to a different question, and a timeout (or block) could kick in for multiple rapid errors from the same IP.

Sure, they can be defeated, just as visual and audio CAPTCHAs can be. But in practice, generally? What we already *know* is that visual/audio CAPTCHAs aren't Accessible, and they trip up and/or annoy many people. And (back on topic) surely Google, if anyone, could devise a system which scaled and translated, if they put their minds to it ...

<< "Type in the word Mississippi after removing one each of the repeated letters ___ ”

It's too complicated. Why not asking for the tenth number in the Fibonacci sequence?

Or "Type the first and last letters of the word hyperbole ___ " ;)

But the Mississippi question is too ambigous, I grant you.

The first rule of usability is "Don't make me think". I think Milly those questions, while trivial, take a second or two to compute, and that might be even more of a barrier than entering a bunch of numbers and letters (which is also annoying, but at least a kind of "monkey" work). Hmmm.

Milly, sorry for calling you Missy. You raise some good points about China. My issue was that the Chinese have nothing to lose with Google going to China. Google can argue that by adding a second, faster search engine in China, which may be able to slip some stuff past the censors, there is only net gain.

But even Sergey has conceded your point that from an ethical point of view, China may not have been the company's best decision. Can I use your comments in my book?

Philipp, I agree it's hard to find the right balance. I suppose we first have to decide whether (depending on the circumstances) we'd prefer our humans-only barrier (if we really must have one) to block some people who won't bother to hurdle it, or some people who can't hurdle it. Perhaps that'll be a decision based on numbers; perhaps on fairness and non-discrimination; perhaps (though not nearly often enough, in my view) on both.

Richard, you're welcome to use my comments (er, if it's also okay with Philipp's non-commercial Creative Commons licence – see link below ;).

Btw, there's more on the same Google likes/dislikes theme (from me, but more importantly some great links to smarter peoples' thoughts) here: http://www.imilly.com/google-cookie.htm#goodguys

I’m one of several Googlers who answer questions online and sometimes for the press. I usually handle questions about webmasters or SEO, so in those areas I’m more likely to make sense and less likely to say something stupid. If I post something here that you find helpful as you build or manage your web presence, that’s wonderful. But when push comes to shove: This is my personal blog. The views expressed on these pages are mine alone and not those of my employer.
http://www.enterbet.com

Personally, my only beef with Google is the sector 5 error in Gmail, and the way Google has been so remiss in correcting it. In the Google Help Center, I've found absolutely nothing helpful.

the problem with google is the same problem with microsoft – whom you heartily endorsed. google is trying to be everything to everbody and will eventually "corner the market" on email, searches, etc. what happened to microsoft? basically they now suck donkey balls but we're all stuck with this substandard product and forced to "upgrade" crap like word, excel, etc. none of which have seen a meaningful improvement in 10 years...but hey, if you don't buy, good luck installing on your new computer a few years from now.

eventually, google will find itself in the same place: owning so much of market share that they couldn't give a flying flip about what works and what doesn't, just as long as they keep collecting coin. and, like microsoft, they'll gobble up any competitors who have a superior product and keep their competition out of the marketplace.

not to mention that we should all be running s**t scared about how much information google collects and STORES FOREVER regarding your browsing, searching, etc. habits. pretty soon our neo-fascist government won't even need to bother illegally spying on folks. just a quick call to google – who will NOT have the stones to say "no" – and your entire history is available to any dim-witted dullard who works for the government...(sorry for being redundant – "dim-witted dullard" and "works for the government" are pretty much synonymous.)

of course, you and your fellow shareholders will continue to profit and that's really that matters, right? so who cares what happens?

for the record, i'm hardly some "socialist" who despises shareholders and profits. i've made my living for better than 20 years trading u.s. markets...but i know a rat when i smell one...and google is one.

Interesting, but how do you know that Google is a rat, and that it will ignore customers once it has too much market share? It's definitely an issue, but not only is the jury out, the crime has not yet been committed.

One thing I like about Google is the thing that stock analysts used to complain about two years ago: it has no "lock-in" the way Microsoft has. People can easily switch to another search engine if a better one comes along. That means it has to satisfy customers' need to stay ahead, keeping it competitive. Stock analysts don't seem to like that vulnerability.

On the other hand, I was at a talk Schmidt gave at Stanford a couple years ago, in which he advised budding entrepreneurs to find some sort of lock-in. I asked him why he advised that, when Google seems to have no such advantage. His answer was, basically, wait and see. That really disturbed me.

Email could provide such lock-in. It think it's the only thing that keeps Yahoo mail and Outlook going.

Historically, at least, Google has shown determination to keep search results accurate and to give more weight to customers' needs than those of advertisers, investors, board members, etc.

However, I concede that Google may be slipping. Based on some of the comments here, I posted an opinion on my blog, "Microsoft CAN beat Google," http://richardbrandt.blogs.com raising the question of whether Google will be able to keep the customer satisfied. Microsoft, now the underdog, may actually do a better job.