Month: October 2018

Late today I was sent a “press release” from “Blind: Your Anonymous Workplace Community” (“teamblind”) with the headline: 

88.4% of Google Conservatives Feel Their Political Views Not Welcome at Work

along with some response breakdowns of “liberal” – “moderate” – “conservative” and so on.

I wasn’t really familiar with Blind, but I did remember something from August where they claimed that:

65% of Google Employees Are in Favor of Censored Search

These are intriguing numbers, but as an old statistics guy from way back — ever since I read the 1954 (and still a classic) “How to Lie with Statistics” by Darrell Huff — I had to ask myself, what sort of statistically valid methodology is Blind using to gather these numbers?

Turns out — as far as I can tell at this point (and I’m certainly open to being corrected on this if I’m wrong!) — there appears to be no valid statistical methodologies in those surveys at all!

Blind’s primary model, as far as I can determine, is an app that interested users can install where various surveys are offered, and users who want to participate in particular surveys can choose to respond to them. 

To help ensure that workplace surveys are answered by actual employees of specific firms, Blind apparently verifies that users have appropriate corporate email addresses.

That serves to try keep random people out of the surveys, but doesn’t make those surveys in any way statistically valid, because they apparently remain fully “self-selected” surveys subject to the well known problems of “self-selection bias” effects.

In other words, you can’t infer any statistical information from these surveys beyond the opinions of the particular people who happened to be interested enough at any particular time to respond, and that will vary greatly depending on the nature of the questions and the types of people predisposed to install the Blind app and participate in any Blind surveys in the first place.

Your basic Statistics 101 course explains why the big polling organizations like Gallup — who do generate statistically valid surveys and polls — use carefully designed mathematical models to determine whom THEY will contact for surveys. They don’t just say “Hey, come on over and vote on this!” That’s why meticulously designed surveys of around 1000 or so people can be extremely accurate even when looking at national issues.

That’s not to say that Blind’s self-selected surveys regarding Google or other firms are worthless — they are indeed snapshots of interested users from subsets of their app’s user community. But that’s all.

It would be a tremendous error to try extrapolate from self-selected Blind surveys to any populations beyond the specific individual app users who chose to respond — so such surveys are essentially worthless for serious analysis or policy planning purposes.

This was true when Darrell Huff wrote his book in the mid-20th century, and it remains just as true today.

–Lauren–

Of late we’ve seen both leaked and open evidence of many employees at Internet tech firms in the U.S. rebelling against their firms participating in battlefield systems military contracts, mostly related to cloud services and AI systems.

Some reactions I’ve seen to this include statements like “those employees are unpatriotic and aren’t true Americans!” and “if they don’t like the projects they should just quit the firms!” (the latter as if everybody with a family was independently wealthy).

Many years ago I faced similar questions. My work at UCLA on the early ARPANET (a Department of Defense project) was funded by the military, but was research, not a battlefield system. A lot of very important positive research serving the world has come from military funding over the years and centuries.

When I was doing similar work at RAND, the calculus was a bit more complex since RAND’s primary funding back then was also DOD, but RAND provided analytical reports to decision makers, not actual weapons systems. And RAND had a well-earned reputation of speaking truth to power, even when that truth was not what the power wanted hear. I liked that.

But what’s happening now is different. The U.S. military is attempting to expand its traditional “military-industrial” complex (so named during a cautionary speech by President Eisenhower in 1961) beyond the traditional defense contractors like Boeing, Lockheed, and Raytheon.

The new battle systems procurement targets are companies like Google, Amazon, and Microsoft.

And therein lies the root of the problem.

Projects like Maven and JEDI are not simply research. They are active battlefield systems. JEDI has been specifically described by one of its top officials as a program aimed at “increasing the lethality of our department.”

When you sign on for a job at any of the traditional defense contractors, you know full well that battlefield operational systems are a major part of the firms’ work.

But when you sign on at Google, or Microsoft, or Amazon, that’s a different story.

Whether you’re a young person just beginning your career, or an old-timer long engaged in Internet work, you might quite reasonably expect to be working on search, or ads, or networking, or a thousand other areas related to the Net — but you probably did not anticipate being asked or required to work on systems that will actually be used to kill people.

The arguments in favor of these new kinds of lethal systems are well known. For example, they’re claimed to replace soldiers with AI and make individual soldiers more effective. In theory, fewer of our brave and dedicated volunteer military would be injured or killed. That would be great — if it were truly accurate and the end of the story.

But it’s not. History teaches us that with virtually every advance in operational battlefield technology, there are new calls for even more military operations, more “interventions,” more use of military power. And somehow the promised technological advantages always seem to be somehow largely cancelled out in the end.

So one shouldn’t wonder why Google won’t renew their participation in Maven, and has now announced that they will not participate in JEDI — or why many Microsoft employees are protesting their own firm’s JEDI participation.

And I predict that we’re now only seeing the beginnings of employees being unwilling to just “go along” with working on lethal systems.

The U.S. military has made no secret of the fact that they see cloud environments, AI, robotics, and an array of allied high technology fields as the future of lethal systems going forward.

It’s obvious that we need advanced military systems at least for defensive purposes in today’s world. But simply assuming that employees at firms that are not traditional defense contractors will just “go along” with work on lethal systems would be an enormous mistake. Many of these employees are making much the same sorts of personal decisions as I did long ago and have followed throughout my life, when I decided that I would not work on such systems.

The sooner that DOD actually understands these realities and recalibrates accordingly, the better.

–Lauren–

UPDATE (November 22, 2018): Save Google — but Let Facebook Die

UPDATE (November 3, 2018): After the Walkout, Google’s Moment of Truth

– – –

The Death of Google
Lauren Weinstein
8 October 2018

Blog: https://lauren.vortex.com/the-death-of-google
PDF: https://lauren.vortex.com/google-death.pdf
Google Docs: https://lauren.vortex.com/google-death.gdoc

Google is dying. It may be possible to save the patient, but it’s also quite possible that Google has already passed the point of no return, especially with the array of forces now attacking it from all sides and from within. Since this situation has been largely enabled by unforced errors committed by Google itself, the prognosis can only be described as bleak.

Unfortunately, I have strong doubts that Google is capable at this time of making the kinds of “lifestyle changes” that would be required to truly save themselves. I would love to have these doubts proven to be incorrect.

A company named Google and its parent Alphabet will continue to exist for the foreseeable future, but for all practical purposes the Google that we all know appears to be in a kind of terminal decline, even as the money continues rolling in for now.

How can this be?

Today’s announcements of a Google+ security breach and the upcoming shutdown of consumer Google+ are but immediate symptoms of a malignancy that has been creeping through Google for years. UPDATE (October 11, 2018): This turns out to be more of a bug than a breach per se, and as I note below its security impact is virtually nil. However, it still should have promptly been made public.

As a big fan of Google, spending a significant amount of my time retorting the mischaracterizations and lies of the Google haters via my written posts and radio interviews, I take no pleasure in this kind of diagnosis.

I’ve watched the death throes of other major technology firms over the years, who originally seemed nothing short of invincible. 

AT&T for one. Digital Equipment Corporation (DEC) was another. Their declines took time — these are processes rather than events. It’s actually a fairly long list if you go far enough back. DEC was assimilated into other firms and its talent siphoned off in various directions. AT&T today is still large and powerful but in many ways is but a shadow of its former self, with its gems like Bell Labs long since morphed into meaningless.

The forces that are ripping Google apart are somewhat different in kind, but all the more tortuous and painful to behold.

For at its core, Google is suffering a complex and multifaceted ethical dilemma that not only threatens to decimate the firm from the inside over time, but has opened up vast gaping wounds that legions of politically-motivated Google haters are using to further evil agendas.

I’ve traveled quite the arc when it comes to Google. In their earlier days starting some 20 years ago, I was a rather intense critic — various of their early data collection and privacy practices seemed to be driven by a cavalier attitude that I viewed as unacceptable.

My first direct physical contact with Google occurred in 2006, when I was invited to Google’s L.A. offices to give a talk that I entitled “Internet & Empires” (the video of that presentation by a significantly younger version of myself is here: https://www.youtube.com/watch?v=PGoSpmv9ZVc). 

I believe it was the first talk they’d ever recorded at that office. There was no podium yet — I just sat on the edge of a table for the presentation.

My interactions with Googlers that day — both from the Q&A and our later discussions before I headed home — yielded me an immediate epiphany of sorts.

Googlers are probably the best people I’ve ever met or worked with in tech — or anywhere else for that matter. It was an honor to consult to Google internally and work directly with them for a significant period several years ago.

They’re intelligent. They care. Many of them are pretty nerdy — but I certainly plead guilty to that myself. I’ve nearly never met a Googler that I didn’t like.

But it became immediately clear that day back in 2006 that something of a discontinuity existed between “rank and file” Googlers and some individuals in Google’s upper management. Even on that first day of contact, Googlers expressed to me their frustrations in this regard, relating to the very issues that I had discussed in my talk.

Over the years since, a wide range of issues related to Google have changed dramatically for the better. Google has become a world-class leader in privacy, security, and artificial intelligence policies. This doesn’t mean that Google is perfect in these respects, and bugs can still occur, but they have excellent people working on those teams — I know many of them personally — who put their lives into this important work. 

However, in key respects it seems that the chasm between Google’s management and other Googlers has grown from a disconnect to a gaping chasm.

Google has always had what I’d charitably call “blind spots” in various areas. Over the years I’ve written publicly about these many times, and I won’t go into detail about them again here, but we can briefly review a few.

Customer service has been an ongoing problem since day one. It has certainly made significant positive strides over time, but still is massively lacking in important respects, especially when dealing with growing populations of non-techie users who depend on Google products and services, but are increasingly left behind by Google user interface designs and available help resources.

When it comes to user interfaces, readability, and similar areas, we again see a sort of “split personality” from Google. They have excellent and rapidly evolving resources for persons with severe conditions like blindness, but continue to deploy low contrast fonts and confusing user interfaces that drive many users with common visual deficiencies absolutely nuts.

Proposals to create the kinds of roles at Google that have been so successful elsewhere — such as Ombudspersons and Consumer Advocates — have continually and routinely hit brick walls at Google whenever I’ve suggested them. I’ve probably written a hundred thousand words or more on this topic alone in my various essays about Google issues.

It has been very clear that Google’s style of public communications has became a major part of their ongoing problems — because in my experience so many common false claims about Google are easily refuted when you take the time to actually do so in a way that non-techies will appreciate.

Yet Google PR has always had a tendency to clam up when something controversial occurs — until the situation has escalated to the point that silence is no longer an option, and matters have become much worse than they would have been if dealt with publicly in a prompt fashion. Google’s deeply entrenched fear of the “Streisand Effect” — the idea that if you say anything about a bad situation you will only draw attention to it — has not served them well.

Today’s belated announcement of a security breach related to Google+, which appears to be the handy excuse for Google to shut down consumer Google+ over a period of 10 months — a process that Google also announced today — encapsulates much of what I’ve said above.

Though the practical impact of the breach seems to be negligible, Google played directly into the politically-motivated hands of the lying Google haters, who have already been screaming for Google’s blood and for its executives to be figuratively drawn and quartered. 

These kinds of Google communications strategies are giving the evil haters even more ammunition to use for false accusations of political user censorship, they give the EU additional excuses to try fine Google billions extra to enrich EU coffers, and they give massive energy to the forces who want to break up Google into smaller units to be micromanaged for political gain by politicians and those politicians’ minions and toadies. 

In the case of Google+, while I don’t have any inside information about today’s announcements, it’s pretty easy to guess what happened.

I’ve been a very active user of Google+ since the first day of beta availability in 2011. But it was obvious from the outset that Google management’s view of the platform was significantly different from its many dedicated users — and there are many millions of them despite the claims of naysayers. I have a wonderful core following of Google+ users who are absolutely great people, and the loss of Google+ will make me both sad and yes, extremely angry. It’s difficult to consider this to be anything short of loyal users being betrayed by Google itself.

Because it didn’t have to happen. Google+ has obviously been operating on very limited internal support resources for quite some time — this was apparent to anyone who used G+ routinely. And there were some terrible executive decisions made along the way — perhaps mostly notably an ultimately abandoned integration of G+ and the YouTube commenting system, which cross-contaminated completely different spheres of interest with disastrous effects. I advocated against this both publicly and internally, but even though it was ultimately rescinded the damage was already done.

Another Google self-inflicted injury is the new controversy over purported plans for Google to again provide Chinese government censored search in China, a concept that Google abandoned many years ago. I’ve written a lot about this recently — I believe it’s a terrible idea and plays into the hands of Google’s adversaries — but I won’t get into the details again here, other than to note the great distress that these moves and the ways that they were handled internally have caused many Googlers who have spoken out publicly.

And yet as I’ve also recently written, when we view that leaked Google TGIF video where Google executives discuss this matter, you won’t see any evil intents, and in fact you’ll find execs emphasizing the need to continue preventing any political bias from finding its way into Google search or other Google products. So their hearts are clearly in the right place overall.

But even the best of intentions are not enough.

With the opening words of Google’s 2004 IPO Founders Letter, Larry Page and Sergey Brin wrote:

“Google is not a conventional company. We do not intend to become one.”

I can’t help but be reminded of that classic scene in “Citizen Kane” where Charles Foster Kane takes the “Declaration of Principles” that he’d written many years earlier and rips them to pieces, declaring them to now be antique.

It is indeed possible, even likely, that Google can continue onward without the kinds of changes that I and other Google supporters have advocated over the years, and still make bushels of money.

But it won’t be the same Google. It will have become the “conventional company” kind of Google, not the firm of which so many Googlers are so rightly proud, and that so many users around the globe depend upon throughout their days.

The Google that we’ve known will be dead. And with its passing, we’ll be entering into a much darker phase of the Internet that many of us have long feared and have worked so hard to try prevent.

And that loss would be terrible for us all.

–Lauren–

UPDATE (October 6, 2018): It appears that at least some Gmail users are now getting an (apparently one-time) pop-up box giving the option to turn off “Smart Compose” when it first becomes active for them. This is definitely an improvement. However, if someone accepts that default (“Got it”) to try it out, there’s no clue provided to help the user turn it off again at some future time, without digging around in the user interface as I describe below. Many users report regretting accepting it in the first place, since they didn’t know how to turn it off afterwards.

– – –

I had sort of hoped that Google would step up to the bat on this one themselves, but my inbox is still full of queries about this — all day, every day.

Google recently deployed a feature in Gmail that tries to guess what you’re about to type, and “helpfully” fills it in for you. They activated it by default, with no information provided to users (not even a one-time pop-up information bubble) explaining how to turn it off. (Please see update above regarding this aspect.)

I’ve seen this “Smart Compose” feature described publicly with a range of adjectives, including intrusive, wonderful, invasive, creepy, accurate, loony, mistaken, helpful, misguided — well, you get the point, opinions are all over the map.

In my case, I’d say that “annoying” is the descriptor I’d sort to the top of the heap. 

With the understanding that Google has great AI and is itching to use it whenever and wherever possible, I don’t really need it analyzing my email drafts as I type them. At least in my case, its proposed wordings are nearly always — what’s the technical term? — oh yes, WRONG. Not what I intend or want to write. 

And the predictions intrusively and continuously interrupt my flow of typing as each one needs to be individually bypassed. 

More Google-enhanced “dumbing-down” I really don’t need. Luckily, like the silly little “smart reply” labels that Gmail pops up by default these days (also useless for me, but far less annoying than Smart Compose”) this feature CAN be disabled.

Of course, you have to go on the usual Google user interface scavenger hunt to figure out how to turn this new feature off, because as I noted above, Google sprung it on everyone without information about opting out from its tender mercies. (Please see update above regarding this aspect).

I would not assert that “Smart Compose” is useless. For users who do find it helpful that’s excellent, fine, and dandy. More power to them, as the saying goes. Smart Compose generally seem more acceptable and helpful for mobile use — though Google mobile voice input is so good that voice is my own preferred method to input text on mobile.

My foundational complaint here isn’t that Google deployed Smart Compose, but rather that they enabled it by default without providing users even basic related information, including the all important “How the hell do I turn this damned thing off?” — the very question filling my inbox of late! (Please see update above regarding this aspect.)

So here’s how you turn it off. It’s easy, IF you know how.

Click the desktop Gmail gear icon at the upper right. Then click Settings. You should already be on the General tab at this point. Scroll down until you find “Smart Compose” and click the “Writing suggestions off” choice. Many users assume that their changes have taken effect at this point. Nope, not yet. You next must scroll all the way to the bottom of the page and click “Save Changes” to actually cause any changes to take place.

By the way, you can also turn off the “Smart Reply” feature I mentioned above, via this same settings page. 

There are many better ways that Google could have deployed Smart Compose. Instead of enabling it by default, they could have popped an invitation to try it. Or if it had to be enabled by default, they could have popped a little box saying something like “Can be disabled on the General tab in Gmail settings” — or something along these lines. (Please see update above regarding this aspect.)

Unfortunately, the way that Google chose to launch Smart Compose is rather emblematic of continuing blind spots in Google’s attitudes toward user interface design and the needs of their very wide community of users. 

Google can easily do better, if they choose to do so by considering the needs of ALL users in these user interface decisions and designs.

–Lauren–

Just a very short note! A few days ago, in “How Google Documentation Problems Can Lead to Public Relations Nightmares” (https://lauren.vortex.com/2018/09/27/how-google-documentation-problems-can-lead-to-public-relations-nightmares), I proposed that Google make available a series of tutorial resources — “explainers” so to speak, regarding a wide range of Google services, technical issues, and policies that tend to be misunderstood by significant numbers of persons in their user community and the global community at large. I suggested that both textual and video content in this vein could well serve toward improving the understanding of many things Google.

Apparently some readers misunderstood my post — or perhaps were incompletely informed about it by third parties. Because I’ve been flooded with people asking me where to find these “Google Explainer” resources.

You can’t find them. They do not exist at this time! I was making a proposal, not an announcement.

I hope that Google will move in the direction that I’ve suggested, but there are of course no guarantees that they will do so. I appreciate the emails expressing support for the concept, but this ball is firmly in Google’s court, not mine! It would not be practical for a non-Googler to write up such docs and keep them in sync with Google to the degree that would really be necessary for such resources to be genuinely useful.

Sorry about that, Chief!

Be seeing you.

–Lauren–