Wednesday, September 20, 2017

Bodega and the power of narrative

As mentioned before, criticizing something for being overhyped invariably to some degree makes you part of the problem. No matter how pointed and snarky your approach may be, there is simply no getting around the inherent contradiction in spending an extended period of time basically saying "we shouldn't be wasting time on this."

This is very much a danger when discussing the Bodega vending machine company (and, yes, that's what it is). This is a small and silly startup with lots of talk but few ideas, none of them good or particularly original. Though, in terms of functionality, it's generally inferior to existing vending machines, it insists that it will not only take that market but will actually replace the modern convenience store. If you need more than a couple of tweets to dismantle this business plan, you should tighten your prose.

The company is, however, of interest as a case study in the ways that coverage of business and technology often devolves into bullshit. We've already discussed the premise story, where journalists explicitly or implicitly lend credibility to otherwise laughable proposals because those proposals provide an excuse to write about a cherished topic.

Bodega also provides a highly instructive example of the power and appeal of a magical heuristics narrative. While there is no reasonable business logic that can support this model, the mystical argument is quite strong. Both the magic of association and the magic of language play a powerful role.

The narrative is so powerful that it sometimes overwhelms journalists' own experience.

Elizabeth Segran writing for Fast Company:
In most cases, Bodega doesn’t pay for the retail space, but pitches itself as an amenity or a convenience to property managers. At gyms for instance, McDonald makes the case that having a Bodega stocked with power bars and protein powder might make the facility more attractive to members.
And Mike Murphy writing for Quartz:


The cabinets are filled with items to buy that are relevant to where you are at that given moment: In the office, they might be filled with snacks and drinks; at the gym, they could have Gatorade, supplements, and knee braces; in an apartment lobby, there could be detergent, pharmaceutical goods, and perhaps some snacks, too. Whatever makes most sense.
 
Here in Los Angeles, I have a membership at LA Fitness. That's a very big chain in Southern California with lots of locations, and I have hit well over a dozen. Furthermore, since I've moved around quite a bit, I've gone to a number of gyms in Arkansas, Virginia, and Georgia. I won't claim that this constitutes a representative sample, but it is a pretty good cross-section, and in all that time I cannot recall a gym that did not have vending machines that sold items such as energy drinks, power bars, earbuds, and various protein supplements. Many also had stores/cafés/juice bars, but pretty much all had vending machines.

The fact that these establishments already have a machine that sells exactly the same things that McDonald proposes offering in his machines is definitely worth mentioning. It pretty much takes this part of the pitch out at the knees. But Segran and Murphy either ignore or possibly simply don't notice this fact because it undercuts the narrative. These are two ex-Googlers using data and artificial intelligence and machine learning and all-purpose creative destruction to solve problems and disrupt old and sclerotic industries. The fact that many people have already done what McDonald's promising to accomplish does not fit the story at all.

In fairness, there have been journalists covering this story using actual business principles. Eater and TechCrunch have both done good work in this area.

Tuesday, September 19, 2017

Scientific American's embarrassing example of the premise story, or "yes, it is just a god damn vending machine."




When criticizing the overhyped for being overhyped, one always runs the risk of becoming part of the problem. That said, the badly conceived vending machine startup Bodega is just too good an example of what's wrong with the way we discuss technology and business. It is one of those cases where multiple lines of bad journalism and bad investor logic converge.

For instance, one of the most popular and damaging tech narratives is the premise story. Some company or entrepreneur makes some ridiculous claim about how his or her (the former most of the time) new product or service will revolutionize everything. This claim will almost certainly never come to pass but it gets a tremendous amount of buzz and provides a great excuse to discuss something a writer would like to talk about. The result is a premise story.

The premise story provides a highly useful loophole for tech and business writers. An article about a proposal that is obviously fatally flawed needs to address the lack of viability, but when the proposal is framed as a premise ("if we have [Mars one/Hyperloops/Bodega boxes], what will life be like."), The author can slide by with the mildest of conditional statements.

The trouble is that the implications of an event that almost certainly won't happen might occasionally be interesting in a thought experiment sense but are, by definition, trivial. Since the authors of these pieces want to be taken seriously, they inevitably if implicitly lend an unwanted degree of credibility to these impractical schemes.

Case in point.  Krystal D'Costa, blogging for Scientific American, clearly wanted to write something about real bodegas, their history, their cultural significance, the roles they play in their communities, this is an interesting and admirable topic, but in order to make it marketable and bring in the traffic, she needed to tie it to one of the silliest startup ideas to come out of Silicon Valley in recent memory. (At this point, I am supposed to say "since Juicero," but, at the risk of praising with faint damns, Bodega's business plan really is worse.)

Thus we get these opening paragraphs:
Need deodorant? Or craving potato chips? Don’t want to run to the “corner store” to get it? Well, soon enough you may only need to go as far as a deluxe vending machine. A start-up is looking to put "pantry boxes" stocked with the non-perishable items you might find at a corner store in key locations, like your gym, or a dorm common area or apartment lobbies. The name for this venture? Bodega.

As the name implies, the creators want to replace brick-and-mortar bodegas—the name given to those ubiquitous corner stores in New York City and Los Angeles—with these pantry boxes. They believe that with time, and data, they can tailor these boxes to meet the specific needs of an immediate area by tracking purchases and restocking those items that are most commonly purchased there, which is what seems to distinguish these boxes from normal vending machines. (The logistics of this restocking has not been disclosed as it is presumably a large part of the business plan.) But the idea has drawn a public outcry—and for good reason: It presents a very real threat to smaller mom-and-pop type stores, but it also appropriates the name “bodega," without acknowledging the cultural and social capital that these spaces have.


That “very real threat” is what makes this a topical, click-worthy story; it is also complete bullshit. In terms of functionality, these pantry boxes are nothing more than fairly crappy vending machines. There is nothing you can get out of them that isn't available from existing machines and a great deal that you can get from other machines but you can't get from them. No cold sodas. No hot beverages. No frozen treats. (We won't even get into many of the more elaborate, robotic offerings in Japan and Europe.)

The pantry boxes fair even worse compared to bodegas and other convenience stores in terms of selection. In addition to all the things listed above, they offer a small fraction of the nonperishable goods available in even a tiny store, not to mention no produce, alcohol, or tobacco products.

In lieu of any real improvement in functionality, the founders offer magical heuristics. Ex-Google employees from Silicon Valley using artificial intelligence to disrupt an industry. Even "data" is used here in the magical rather than statistical sense. Large, well-established vending machine companies have tons of quality data going back forever and constantly being updated. Bodega is actually working at a huge disadvantage here, one that would take years of widescale operation to erase.

One of the most telling lines in the excerpt above is the speculation about the logistics of restocking. This is, of course, an enormously challenging business problem. It is one of the primary reasons why vending machines have not already carved out a much larger segment of the retail market.  D'Costa assumes that the company has a workable solution to the problem and is just keeping it to itself. There is absolutely no reason to believe this – – Silicon Valley startups routinely jump into press releases and funding pushes with inchoate and/or unworkable business plans – – but believing that the details have been carefully thought out is essential for preserving the premise.

The sad part is that the rest of  D'Costa's article is interesting and well worth reading. It's a shame she felt the need to prop it up.

Monday, September 18, 2017

Even by Stephens' standards, this is bad advice.

Sorry about not getting around to this piece earlier (so much stupid, so little time), but fortunately there's a timeless awfulness to Bret Stephens' “Tips for Aspiring Op-Ed Writers.” It's a genre that brings out the pretentious, the banal and the clichéd in even the best writers. With a hack like Stephens, you're pretty much guaranteed material for your next what-not-to-do chapter. The man does not disappoint… No, that doesn't sound right. He leaves potential critics with an embarrassment of riches…  No, that's not right either. Let's just say he's consistent.

While there is much to mock here, one piece of advice is not just bad and hackneyed; it's out-and-out dangerous.
4) Authority matters. Readers will look to authors who have standing, either because they have expertise in their field or unique experience of a subject. If you can offer neither on a given topic you should not write about it, however passionate your views may be. Opinion editors are often keen on writers who can provide standing-with-surprise: the well-known environmentalist who supports nuclear power; the right-wing politician who favors transgender rights; the African-American scholar who opposes affirmative action.


Putting aside the potential pitfalls of arguing from authority and ignoring the fact that Stephens' entire career has been based on getting people to print his passionate views on topics where he has no appreciable expertise, relevant experience, or discernible understanding, let's focus on the second half of the paragraph. (We'll also skip the rather curious notion that a pro-nuclear environmentalist would be that difficult to find. Coal would have been the more appropriate choice.)

The standing-with-surprise standard is virtually guaranteed to mislead readers, often on matters of vital importance. There is no concept more essential to coverage of complex issues than context and particularly consensus. There are also few concepts that journalists screw up so frequently or so badly.

Invariably, the standing-with-surprise story distorts the perceived consensus. Even the most decisively settled scientific question can be depicted as a matter of ongoing debate. What's worse, by favoring the man-bites-dog stories, editors encourage journalists to further play up the counterintuitive aspects by stretching what qualifies as relevant expertise or by taking relatively noncommittal statements and framing them as challenges to the establishment view.

This technique is also particularly useful for getting he said/she said quotes to spice up an article. A typical example reads something like this: "Climate scientists tell us that we are approaching dangerous thresholds in global warming. Others, however, are not so certain. Physicist [just make up a name] of [prestigious university] says that models of complex systems based on observational data always have a potential for error."

Bret Stephens has, of course, gotten where he is today by misinforming readers along these very lines. We probably shouldn't be surprised that he considers the practice a rule of good journalism. That doesn't mean we should let him get away with it.

Friday, September 15, 2017

That's a disturbing premise

I came across this while looking for Cohan clips for a previous post.




The Phantom President

The Phantom President tells the fictional story of American presidential candidates, based on the novel by George F. Worts. A colorless stiff candidate for President is replaced in public appearances by a charismatic medicine show pitchman.






A few notes on IBM's Watson, the battle against cancer, and what's wrong with the state of 21st century innovation

If you haven't read the Stat News article by Casey Ross and Ike Swetlitzon on IBM's project to revolutionize cancer treatment using Watson, you should do so as soon as possible. The piece raises all sorts of important points. I'll try to return and explore some of them in greater detail (perhaps even convince Joseph to join the conversation), but for now there are two or three I want to it while the topic is still fresh.

Before we go on, you should take a few minutes to listen to the following track from the Button-down Mind of Bob Newhart. It's a seminal comedy bit that also happens to be directly relevant to the conversation.





The joke here is not that using airplanes for passenger travel is absurd (it was the early 60s for Christ's sake); the joke is that the aircraft at Kitty Hawk was clearly not ready to be monetized in this way. A plan that will be viable sometime the future is not really viable.

Given the complexities of the problem and the mountainous quantities of research that need to be assessed, no one would argue that AI-based tools for diagnosis and recommending treatment wouldn't be tremendously valuable, probably even revolutionary, but IBM's Watson is looking definitely less than ready for prime time in this respect.

The stat news article also hits on a couple of long-standing threads here at the blog, starting with one of the best examples of ddulite reasoning we've seen in a long time.
IBM said in its statement that it has collaborated with the research community and presented data on Watson at industry gatherings and in peer-reviewed journals. Some doctors said they didn’t need to see more research to know that the system is valuable. “Artificial intelligence will be adopted in all medical fields in the future,” said Dr. Uhn Lee, who runs the Watson program at Gachon University Gil Medical Center in South Korea. “If that trend, that change is inevitable, then why don’t we just start early?”

Adopting technology because it supposedly will be the best choice sometime in the future is deeply flawed for at least a couple of reasons. First off, the future is, you know, in the future – – we don't know what we'll find there. While it is reasonably safe to assume that some kind of AI will play a vital role in medicine 20 or 30 years from now, it might have very little to do with the approaches currently being pursued. Furthermore, even early adopters who guess right about the direction of technology are often screwed by getting ahead of the functionality.

Another ongoing thread here at the blog is the ever increasing emphasis on hype and marketing at the expense of research and genuine innovation, particularly when combined with short-term thinking. Following speaks for itself:
“IBM ought to quit trying to cure cancer,” said Peter Greulich, a former IBM brand manager who has written several books about IBM’s history and modern challenges. “They turned the marketing engine loose without controlling how to build and construct a product.”

Greulich said IBM needs to invest more money in Watson and hire more people to make it successful. In the 1960s, he said, IBM spent about 11.5 times its annual earnings to develop its mainframe computer, a line of business that still accounts for much of its profitability today.

If it were to make an equivalent investment in Watson, it would need to spend $137 billion. “The only thing it’s spent that much money on is stock buybacks,” Greulich said.


Thursday, September 14, 2017

Equifax

This is Joseph

Via Kevin Drum, Michael Hilzik of the LA Times has a tough question:
Three Equifax executives sold shares after the discovery of the breach and before its public disclosure, according to Bloomberg. They collected $1.8 million from the sales, which weren’t part of any prearranged option-exercise programs. The sales were made on Aug. 1 and 2, the third and fourth days after the breach was discovered. An Equifax spokeswoman says the executives were unaware of the breach at the time of their sales, but that’s hardly comforting: One was John Gamble, the firm’s chief financial officer. If the firm’s No. 2 executive wasn’t immediately informed about a catastrophic security breach, why not?
This is one of those explanations that actually looks bad, either way.  If the company can't escalate news of a crisis over several days that is a bad, bad sign for their security types.  It also suggests that the senior executives are not aware of optics, as they had weeks, post-sale, to come up with a way to get in front of this news.  If there really was a moment of horror -- my unscheduled sale happened right after a breach nobody told me about -- then they had weeks to craft a thoughtful comment on the bad timing.

Other explanations are less charitable.

In general, the whole way financial stuff is validated in the internet era needs to be re-thought. The move away from personal interaction leads to efficiency, but the use of things like social security numbers as proof of identity is rather silly.  It has the same weakness as birth date -- it never really changes and so once hacked it's permanently compromised.  


Wednesday, September 13, 2017

Magical Heuristics – – you knew there'd be a New York Magazine example eventually


Earlier, we introduced the term magical heuristics (mental tools based on a fundamentally nonrational worldview that, nonetheless, often drape themselves in scientific trappings) and laid out four general categories: magic of association; magic of language; magic of will/doubt/belief; magic of destiny.

This New York Magazine piece by Benjamin Wallace provides a wealth of examples. Though it never lapses into the pseudoscience-curious writing we have come to associate with NYM recently (resonant crystals, homeopathy, Gwyneth Paltrow's goop, the West Hollywood autism epidemic), its underlying narrative is chosen ones doing the impossible through force of will.

[Emphasis added]

But Silicon Valley loved the barf ride. In this, the age of the moon shot — of bold missions to make flying cars and “end all disease” — Musk’s hyperloop met all the criteria of bet-the-ranch, future-shaping audacity: a big vision, promising a new, “fifth mode of transport” after planes, trains, automobiles, and boats; the high purpose of using renewable energy; utopian visuals; and, perhaps most important, a terrific pedigree. Even the Valley’s most peppy cheerleaders weary, occasionally, of pitches for the latest world-changing smartphone apps. Genuine moon shots stir real excitement in the hushed corridors of Sand Hill Road. But the difference between an intrepid moon shot and a misguided fantasy project often hinges entirely on the daredevil behind it.

Musk is Mr. Moon Shot — actually Mr. Mars Shot; he intends to make us a multi-planetary species. He is arguably a genius on his way to becoming a world-­historical figure. At Tesla and SpaceX, he has, through maniacal willpower, work ethic, focus, vision, and risk tolerance, repeatedly proved naysayers wrong by surmounting the seemingly insurmountable. [Not so much. See below. -- MP]  And in recent years, he has assumed the mantle, up for grabs since the death of Steve Jobs, of Silicon Valley’s reigning deity. It was only Musk’s involvement that led many observers to take the hyperloop idea at least semi-seriously.



But for Pishevar, who settled for quoting a different president at the event (Roosevelt, on “the man in the arena”), the propulsion test was a key next step in willing hyperloop into existence. As he wrote later, in a blog post: “There’s a lot of noise, hope, and hype out there about what the Hyperloop could be and will be, but this metal sled absolutely grounds the idea in much-needed reality … I had tears mixed with sand.” Most important, he told me, his parents were still alive. “For me, the fact that they saw the test and have seen the things we’re doing, that’s everything.”

A couple of points. First, the amount of skepticism Elon Musk faced over SpaceX and Tesla has been greatly exaggerated and that exaggeration has been retconned into the record. Second and more important, the primary focus of that skepticism was never on the technology (unlike the response to Musk's more recent and fanciful proposals). Almost no one questioned the viability of electric cars or the engines being developed by TRW. What the doubters took issue with was his ability to raise the capital, quickly get the organizations into place, and avoid being sued out of existence by the aforementioned TRW.

The jump from finance guy to real life Tony Stark connects back to Alon Levy's previous point about omnicompetence.
There is a belief within American media that a successful person can succeed at anything. He (and it’s invariably he) is omnicompetent, and people who question him and laugh at his outlandish ideas will invariably fail and end up working for him. If he cares about something, it’s important; if he says something can be done, it can. The people who are already doing the same thing are peons and their opinions are to be discounted, since they are biased and he never is. He doesn’t need to provide references or evidence – even supposedly scientific science fiction falls into this trope, in which the hero gets ideas from his gut, is always right, and never needs to do experiments.

Musk has been very successful and he surprised almost everyone with his ability to raise capital and set up a large, functional (if not necessarily profitable) companies. Therefore, he can do anything. He is a chosen one. Chosen ones are magical.

Tuesday, September 12, 2017

More magical heuristics -- Levy's omnicompetence

Yesterday, I introduced the term magical heuristics (still open to a better name) to describe nonrational mental tools used by many journalists and investors particularly when discussing science and technology. I laid out four general categories for these heuristics: magic of association; magic of language; magic of attitude; magic of destiny.

This post from Alon Levy (one of the most important contributors to the Hyperloop debate) perfectly fits with two of these categories, magic of association and magic of destiny (the idea that there are chosen ones among us destined for greatness). The whole thing is very much worth reading, but I've selected below the paragraphs that are most relevant to this thread and added emphasis to bring home the point:


There is a belief within American media that a successful person can succeed at anything. He (and it’s invariably he) is omnicompetent, and people who question him and laugh at his outlandish ideas will invariably fail and end up working for him. If he cares about something, it’s important; if he says something can be done, it can. The people who are already doing the same thing are peons and their opinions are to be discounted, since they are biased and he never is. He doesn’t need to provide references or evidence – even supposedly scientific science fiction falls into this trope, in which the hero gets ideas from his gut, is always right, and never needs to do experiments.

...

I write this not to help bury Musk; I’m not nearly famous enough to even hit a nail in his coffin. I write this to point out that, in the US, people will treat any crank seriously if he has enough money or enough prowess in another field. A sufficiently rich person is surrounded by sycophants and stenographers who won’t check his numbers against anything.


...

The more interesting possibility, which I am inclined toward, is that this is not fraud, or not primarily fraud. Musk is the sort of person who thinks he can wend his way from starting online companies to building cars and selling them without dealerships. I have not seen a single defense of the technical details of the proposal except for one Facebook comment that claims, doubly erroneously, that the high lateral acceleration is no problem because the tubes can be canted. Everyone, including the Facebook comment, instead gushes about Musk personally. The thinking is that he’s rich, so he must always have something interesting to say; he can’t be a huckster when venturing outside his field. It would be unthinkable to treat people as professionals in their own fields, who take years to make a successful sideways move and who need to be extremely careful not to make elementary mistakes. The superheros of American media coverage would instantly collapse, relegated to a specialized role while mere mortals take over most functions.

This culture of superstars is a major obstacle frustrating any attempt to improve existing technology. It more or less works for commercial websites, where the startup capital requirements are low, profits per employee are vast, and employee turnover is such that corporate culture is impossible. People get extremely rich for doing something first, even if in their absence their competitors would’ve done the same six months later. Valve, a video game company that recognizes this, oriented its entire structure around having no formal management at all, but for the most part what this leads to is extremely rich people like Bill Gates and Mark Zuckerberg who get treated like superstars and think they can do anything.


Monday, September 11, 2017

Magical Heuristics – we'll be coming back to this one

I read somewhere that John P. Marquand, author of serious and well-respected novels like The Late George Apley and Sincerely, Willis Wayde as well as popular entertainments such as the Mr. Moto books would use the latter as a chance to research ideas and settings for the former. Blogs and twitter can serve an analogous function. Tweets grow into posts and posts grow into articles or books. I have noticed that many of Josh Marshall's pieces show up (in very rough form) on twitter while Paul Krugman often precedes his columns with a series of blog posts on the same topic.

I've made use of both forums and can generally recommend the approach. However, I sometimes get to a piece of the argument I'm building that doesn't want to break down into manageable chunks. The result is always something that I really want to write but which nonetheless keeps getting pushed back in the queue. I've come to thee conclusion it's best to get something out – even if it's rough and incomplete – to open the thread.

One of these is the concept of magical heuristics. As I have spent more and more time critiquing individual reports about technological advances and the larger narratives they represent, I've increasingly come to the conclusion that many, perhaps even most, of these standard narratives, though they make heavy use of the language and imagery of science, are based on essentially nonscientific ways of thinking.

For lack of a better name (and, trust me, I am open to suggestions), let's call these magical heuristics. As a student of George Pólya, I tend to think of heuristics along the lines that he laid out in How to Solve It, mental tools for problem-solving, pattern recognition, and evaluating plausibility. Though these tools are widely applicable, they very much come from the world of math and science and reflect that underlying philosophy.

Over the past few years, I began to realize that many of the journalists and pundits describing and discussing science and technology (particularly the latter) were thinking about these things in a fundamentally different way than researchers, engineers, statisticians did. The final product often looked scientific on the surface but underneath was not at all.

Though I have no expertise in mysticism and supernatural belief systems (and would dearly love the input of someone who does), I tried to work out a basic framework for the mental tools being used.  So far, this is what I've come up with. I still feel like there are essential parts missing, but it's a start:



Magic of association – – properties can be transmitted through proximity (physical or otherwise). This magic is particularly strong in Silicon Valley. Almost any association with someone or something noted for great wealth, success or innovation can pass on these properties. When no direct association is possible, it may be enough to simply invoke the name of a great success which leads us to...

Magic of language – – the proper use of words can alter reality. In addition to the aforementioned example of invoking names like Apple or Google, certain words such as "disruption" are assigned special power. Mission statements actually help determine the fate of companies. Great emphasis is also put on aspirational language which tends to segue into...

Magic of will/belief/doubt – – attitude also shapes reality. Things are more likely to happen the more deeply you believe in them. Correspondingly, skepticism and negative attitudes can undermine this magic. In extreme cases, particularly if surrounded by true believers, there are those who can simply will things into existence which leads us to...

Magic of destiny – – there are chosen ones among us. Their powers are all-applicable, not tied to any specific area or based on specific skills and knowledge; they can simply make things happen. Any association with the chosen ones is unquestionably beneficial. Like messianic American Express cards, they have no preset limits, but they do have at least one weakness: doubters. To question a chosen one is to inspire great hostility.



In the next few weeks, I'm going to fill this out with more examples, many of which will not involve Elon Musk. The eventual goal is to work this into a presentable essay. In the meantime, I can use all the help I can get. Any suggestions?

Friday, September 8, 2017

The end of the networks

A few years ago, faced with not one but two existential technological threats, industry observers started talking about the possible death of network television and the highly likely death of at least one of the majors.

In response, the networks panicked and released an enormous and embarrassing slate of new shows that demonstrated they had not only lost the ability to entertain and innovate, but to pander and steal with any degree of success. It was as if they were determined to prove all of the naysayers correct.

Continued below the break...



Thursday, September 7, 2017

More on Google not not being evil

The Google/New America story we discussed last week continues to reverberate. I'll probably come back to this with a points-to-remember post on Google (and possibly something on the foundation as well). In the meantime, here are excerpts of some articles you might want to check out if you're following story.

As you might expect, Gizmodo has been all over this. Here's an account by Rhett Jones of how Google appears to have used its position to squelch a potential rival to Chrome.
While that sort of practice could be seen as fair game in the business world, Tetzchner says that he recently ran into a more problematic situation with Google. Vivaldi’s account with the Google AdWords advertising network was mysteriously suspended just a few days after he gave an interview to Wired in which he was critical of the company’s privacy practices. He writes:
When we reached out to Google to resolve the issue, we got a clarification masqueraded in the form of vague terms and conditions, some of which, they admitted themselves, were not a “hard” requirement. In exchange for being reinstated in Google’s ad network, their in-house specialists dictated how we should arrange content on our own website and how we should communicate information to our users.
He claims it took about three months to resolve all of Google’s “unreasonable demands,” and he believes that its monopolistic tendencies have reached the point “where regulation is needed.” He didn’t outline what these specific demands were so we reached out for clarification. A Vivaldi spokesperson sent us the following explanation:
They asked us to add an end-user license agreement (EULA) to our website. Which we did – only to get another pushback. Now, Google was asking us to add EULA “within the frame of every download button.” This, they later admitted, was not a “hard requirement” and yet somehow it was enough to keep our account suspended. They also asked us to add detailed information about uninstalling Vivaldi – a requirement they didn’t have direct guidelines on in their help articles. Admitting to that, they wrote to us that they “rely on the advertisers to take responsibility when advertising with us, hence reading up on and following our policy guidelines.” Fair point, if only these policies were clear and available – or if Google lead by example and added EULA “within the frame of every download button” on Chrome’s landing pages. If that wasn’t enough, it took them over a month to approve us as an “authoritative distribution site for free desktop software – Vivaldi.”
The fact that their specialists are dictating how and where we should place our content on our own website is very difficult to fathom. They do not spell out their terms and conditions in the first place and is case by case which is strange.
Three months of work to get a company reinstated in an ad program does sound a bit outrageous—so outrageous that one would probably feel inclined to no longer do business with the company that gives its customers so many headaches. But Vivaldi really doesn’t have a choice, Google controls 75.8 percent of the search ad market. To cut off Google means cutting off three-quarters of the web.


Gizmodo also ran this striking first-person account by Kashmir Hill of how Google used its dominance in search to pressure websites into helping promote Google's attempt at launching a social network, then pressured Forbes into killing a story revealing the practice.


I was working for Forbes at the time, and was new to my job. In addition to writing and reporting, I helped run social media there, so I got pulled into a meeting with Google salespeople about Google’s then-new social network, Plus.

The Google salespeople were encouraging Forbes to add Plus’s “+1" social buttons to articles on the site, alongside the Facebook Like button and the Reddit share button. They said it was important to do because the Plus recommendations would be a factor in search results—a crucial source of traffic to publishers.

This sounded like a news story to me. Google’s dominance in search and news give it tremendous power over publishers. By tying search results to the use of Plus, Google was using that muscle to force people to promote its social network.

I asked the Google people if I understood correctly: If a publisher didn’t put a +1 button on the page, its search results would suffer? The answer was yes.

After the meeting, I approached Google’s public relations team as a reporter, told them I’d been in the meeting, and asked if I understood correctly. The press office confirmed it, though they preferred to say the Plus button “influences the ranking.” They didn’t deny what their sales people told me: If you don’t feature the +1 button, your stories will be harder to find with Google.

With that, I published a story headlined, “Stick Google Plus Buttons On Your Pages, Or Your Search Traffic Suffers,” that included bits of conversation from the meeting.
The Google guys explained how the new recommendation system will be a factor in search. “Universally, or just among Google Plus friends?” I asked. ‘Universal’ was the answer. “So if Forbes doesn’t put +1 buttons on its pages, it will suffer in search rankings?” I asked. Google guy says he wouldn’t phrase it that way, but basically yes.
(An internet marketing group scraped the story after it was published and a version can still be found here.)

Google promptly flipped out. This was in 2011, around the same time that a congressional antitrust committee was looking into whether the company was abusing its powers.

Google never challenged the accuracy of the reporting. Instead, a Google spokesperson told me that I needed to unpublish the story because the meeting had been confidential, and the information discussed there had been subject to a non-disclosure agreement between Google and Forbes. (I had signed no such agreement, hadn’t been told the meeting was confidential, and had identified myself as a journalist.)

It escalated quickly from there. I was told by my higher-ups at Forbes that Google representatives called them saying that the article was problematic and had to come down. The implication was that it might have consequences for Forbes, a troubling possibility given how much traffic came through Google searches and Google News.

I thought it was an important story, but I didn’t want to cause problems for my employer. And if the other participants in the meeting had in fact been covered by an NDA, I could understand why Google would object to the story.

Given that I’d gone to the Google PR team before publishing, and it was already out in the world, I felt it made more sense to keep the story up. Ultimately, though, after continued pressure from my bosses, I took the piece down—a decision I will always regret. Forbes declined comment about this.



Finally, here is a selection from a thoughtful piece from Josh Marshall. Not surprisingly, TPM has also done a first-rate job covering this.

A Serf on Google’s Farm by Josh Marshall

It’s great that all this stuff is coming out. But what is more interesting to me than the instances of bullying are the more workaday and seemingly benign mechanisms of Google’s power. If you have extreme power, when things get dicey, you will tend to abuse that power. That’s not surprising. It’s human nature. What’s interesting and important is the nature of the power itself and what undergirds it. Don’t get me wrong. The abuses are very important. But extreme concentrations of power will almost always be abused. The temptations are too great. But what is the nature of the power itself?

...

What I’ve known for some time – but which became even more clear to me in my talk with Barry Lynn on Monday – is that few publishers really want to talk about the depths or mechanics of Google’s role in news publishing. Some of this is secrecy about proprietary information; most of it is that Google could destroy or profoundly damage most publications if it wanted to. So why rock the boat?



So let’s go down the list: 1) The system for running ads, 2) the top purchaser of ads, 3) the most pervasive audience data service, 4) all search, 5) our email.

...

But here’s where the rubber really meets the road. The publishers use DoubleClick. The big advertisers use DoubleClick. The big global advertising holding companies use Doubleclick. Everybody at every point in the industry is wired into DoubleClick. Here’s how they all play together. The adserving (Doubleclick) is like the road. (Adexchange) is the biggest car on the road. But only AdExchange gets full visibility into what’s available. (There’s lot of details here and argument about just what Google does and doesn’t know. But trust me on this. They keep the key information to themselves. This isn’t a suspicion. It’s the model.) So Google owns the road and gets first look at what’s on the road. Not only does Google own the road and makes the rules for the road, it has special privileges on the road. One of the ways it has special privileges is that it has all the data it gets from search, Google Analytics and Gmail. It also gets to make the first bid on every bit of inventory. Of course that’s critical. First dibs with more information than anyone else has access to. (Some exceptions to this. But that’s the big picture.) It’s good to be the king. It’s good to be a Google.

There’s more I’ll get to in a moment but the interplay between DoubleClick and Adexchange is so vastly important to the entirety of the web, digital publishing and the entire ad industry that it is almost impossible to overstate. Again. They own the road. They make the rules for the road. They get special privileges on the road with every new iteration of rules.



Now Google can say – and they are absolutely right – that every month they send checks for thousands and millions of dollars to countless publishers that make their journalism possible. And in general Google tends to be a relatively benign overlord. But as someone who a) knows the industry inside and out – down to the most nuts and bolts mechanics – b) someone who understands at least the rudiments of anti-trust law and monopoly economics and c) can write for a sizable audience, I can tell you this.: Google’s monopoly control is almost comically great. It’s a monopoly at every conceivable turn and consistently uses that market power to deepen its hold and increase its profits. Just the interplay between DoubleClick and Adexchange is textbook anti-competitive practices.

There’s one way that Google is better than Facebook. When Facebook is getting a bigger and bigger share of the advertising pie, that money is almost all going to Facebook. There are some small exceptions but that’s basically the case. When Google is making insane amounts of money on advertising, it’s not really the same since a huge amount of that advertising is running on websites which are getting a cut. Still, the big story is that Google and Facebook now have a dominant position in the entirety of the advertising ecosystem and are using their monopoly power to take more and more of the money for themselves.

We’re basically too small for Google to care about. So I wouldn’t say we’ve had any bad experiences with Google in the sense of Google trying to injure us or use its power against us. What we’ve experienced is a little different. Google is so big and so powerful that even when it’s trying to do something good, it can be dangerous and frightening.

Here’s an example.

With the events of recent months and years, Google is apparently now trying to weed out publishers that are using its money streams and architecture to publish hate speech. Certainly you’d probably be unhappy to hear that Stormfront was funded by ads run through Google. I’m not saying that’s happening. I’m just giving you a sense of what they are apparently trying to combat. Over the last several months we’ve gotten a few notifications from Google telling us that certain pages of ours were penalized for ‘violations’ of their ban for hate speech. When we looked at the pages they were talking about they were articles about white supremacist incidents. Most were tied to Dylann Roof’s mass murder in Charleston.

Now in practice all this meant was that two or three old stories about Dylann Roof could no longer run ads purchased through Google. I’d say it’s unlikely that loss to TPM amounted to even a cent a month. Totally meaningless. But here’s the catch. The way these warnings work and the way these particular warnings were worded, you get penalized enough times and then you’re blacklisted.

Now, certainly you’re figuring we could contact someone at Google and explain that we’re not publishing hate speech and racist violence. We’re reporting on it. Not really. We tried that. We got back a message from our rep not really understanding the distinction and cheerily telling us to try to operate within the no hate speech rules. And how many warnings until we’re blacklisted? Who knows?

If we were cut off, would that be Adexchange (the ads) or DoubleClick for Publishers (the road) or both? Who knows?

If the first stopped we’d lose a big chunk of money that wouldn’t put us out of business but would likely force us to retrench. If we were kicked off the road more than half of our total revenue would disappear instantly and would stay disappeared until we found a new road – i.e., a new ad serving service or technology. At a minimum that would be a devastating blow that would require us to find a totally different ad serving system, make major technical changes to the site to accommodate the new system and likely not be able to make as much from ads ever again. That’s not including some unknown period of time – certainly weeks at least – in which we went with literally no ad revenue.

Needless to say, the impact of this would be cataclysmic and could easily drive us out of business.



Of course, the real issue is the monopoly and how it applies to money. Is your favorite website laying off staff or ‘pivoting to video’. In most cases, the root cause is not entirely but to a significant degree driven by the platform monopolies – in this case, Google and Facebook – taking a bigger and bigger slice of the advertising dollars. It’s going to their profits and being taken away from publishers who of course are also trying to maximize their profits but do it through paying for journalism.



We could see this coming a few years ago. And we made a decisive and longterm push to restructure our business around subscriptions. So I’m confident we will be fine. But journalism is not fine right now. And journalism is only one industry the platform monopolies affect. Monopolies are bad for all the reasons people used to think they were bad. They raise costs. They stifle innovation. They lower wages. And they have perverse political effects too. Huge and entrenched concentrations of wealth create entrenched and dangerous locuses of political power.

Wednesday, September 6, 2017

"None had ever performed such wonderful things in so short a time"

One more from Mackay's Extraordinary Popular Delusions.

On the 29th of May, the stock had risen as high as five hundred, and about two-thirds of the government annuitants had exchanged the securities of the state for those of the South Sea Company. During the whole of the month of May the stock continued to rise, and on the 28th it was quoted at five hundred and fifty. In four days after this it took a prodigious leap, rising suddenly from five hundred and fifty to eight hundred and ninety. It was now the general opinion that the stock could rise no higher, and many persons took that opportunity of selling out, with a view of realising their profits. Many noblemen and persons in the train of the King, and about to accompany him to Hanover, were also anxious to sell out. So many sellers, and so few buyers, appeared in the Alley on the 3rd of June, that the stock fell at once from eight hundred and ninety to six hundred and forty. The directors were alarmed, and gave their agents orders to buy. Their efforts succeeded. Towards evening confidence was restored, and the stock advanced to seven hundred and fifty. It continued at this price, with some slight fluctuation, until the company closed their books on the 22nd of June.
 It would be needless and uninteresting to detail the various arts employed by the directors to keep up the price of stock. It will be sufficient to state that it finally rose to one thousand per cent. It was quoted at this price in the commencement of August. The bubble was then full-blown, and began to quiver and shake, preparatory to its bursting.
 Many of the government annuitants expressed dissatisfaction against the directors. They accused them of partiality in making out the lists for shares in each subscription. Further uneasiness was occasioned by its being generally known that Sir John Blunt, the chairman, and some others, had sold out. During the whole of the month of August the stock fell, and on the 2nd of September it was quoted at seven hundred only.
 The state of things now became alarming. To prevent, if possible, the utter extinction of public confidence in their proceedings, the directors summoned a general court of the whole corporation, to meet in Merchant Tailors' Hall, on the 8th of September. By nine o'clock in the morning, the room was filled to suffocation; Cheapside was blocked up by a crowd unable to gain admittance, and the greatest excitement prevailed. The directors and their friends mustered in great numbers. Sir John Fellowes, the sub-governor, was called to the chair. He acquainted the assembly with the cause of their meeting, read to them the several resolutions of the court of directors, and gave them an account of their proceedings; of the taking in the redeemable and unredeemable funds, and of the subscriptions in money. Mr. Secretary Craggs then made a short speech, wherein he commended the conduct of the directors, and urged that nothing could more effectually contribute to the bringing this scheme to perfection than union among themselves. He concluded with a motion for thanking the court of directors for their prudent and skilful management, and for desiring them to proceed in such manner as they should think most proper for the interest and advantage of the corporation. Mr. Hungerford, who had rendered himself very conspicuous in the House of Commons for his zeal in behalf of the South Sea Company, and who was shrewdly suspected to have been a considerable gainer by knowing the right time to sell out, was very magniloquent on this occasion. He said that he had seen the rise and fall, the decay and resurrection of many communities of this nature, but that, in his opinion, none had ever performed such wonderful things in so short a time as the South Sea Company. They had done more than the crown, the pulpit, or the bench could do. They had reconciled all parties in one common interest; they had laid asleep, if not wholly extinguished, all the domestic jars and animosities of the nation. By the rise of their stock, monied men had vastly increased their fortunes; country-gentlemen had seen the value of their lands doubled and trebled in their hands. They had at the same time done good to the Church, not a few of the reverend clergy having got great sums by the project. In short, they had enriched the whole nation, and he hoped they had not forgotten themselves. There was some hissing at the latter part of this speech, which for the extravagance of its eulogy was not far removed from satire; but the directors and their friends, and all the winners in the room, applauded vehemently. The Duke of Portland spoke in a similar strain, and expressed his great wonder why anybody should be dissatisfied: of course, he was a winner by his speculations, and in a condition similar to that of the fat alderman in Joe Miller's Jests, who, whenever he had eaten a good dinner, folded his hands upon his paunch, and expressed his doubts whether there could be a hungry man in the world.

Tuesday, September 5, 2017

Like we said before one more time, levels of automation, schedules of implementation

Apologies to regular readers who've been through this before, but probably the two most important concepts when discussing the potential impact of automated vehicles and certainly the most underused are SAE Autonomy Levels and the rate at which each of these levels will likely be adopted. If you follow the coverage closely you will routinely find, even in the most serious and respectable publications, arguments that mix levels of autonomy in the most careless way and, worse yet, promote as just-around-the-corner applications that require virtually every car on the road to be reliably using compatible level V technology.

Just to be clear, autonomy is coming and it is coming quickly, but when you hear promises of curing congestion by packing cars inches apart at high speeds or redesigned roads with narrower lanes and other human-driver-hostile features, consider the following article from Jalopnik.
For one thing, retail sales of fully-automated vehicles aren’t expected to begin until 2025 at the earliest, wrote Mitch Bainwol, in testimony to the Senate Committee on Commerce, Science and Transportation. Bainwol is the president and CEO of the Alliance of Automobile Manufacturers, a trade group that represents Ford, General Motors, FCA, BMW and more—automakers that produce 80 percent of cars on U.S. roads today. To put it plainly, he speaks for a significant chunk of the industry.
And in his written testimony ahead of the committee’s hearing on Wednesday, Bainwol offered a sober assessment of the situation.
Level 4 geo-fenced self-driving vehicles that can only be operated by an Automated Driving System will probably begin around 2021. But, retail sales to consumers of so-called Level 5 vehicles that can operate anywhere a person can drive a conventional vehicle today is unlikely to happen until around 2025 or after. Given how much vehicles cost and how long they last – more than 20 percent of cars on the road today were produced before 2002 – vehicles equipped with Level 5 systems will likely not be a majority of the fleet for three more decades. Ubiquity is not projected to occur for at least four decades largely due to the fact that over 260 million light duty vehicles are registered in the U.S.

Anyone who seriously researched this field, discussed the issues with researchers and made a sincere effort to find out what's going on, had a pretty good idea that this was the case. The credulous, sensationalistic, future-is-now reporting you've seen on the subject is not just misinformed it is irresponsible.

Monday, September 4, 2017

The following video clip should challenge cherished assumptions about Tesla, Silicon Valley, and self driving cars (it probably won't, but it should)

Before we go on, if you haven't, take a few moments and read this (we'll be quoting from it later but you really need to read the whole thing).




Obviously, we need to be careful about reading too much into this. There is a limit to how much you should take away from PR releases and tech reporters. There are, however, a few things we can conclude with some safety.

First, the system looks good. By most accounts it is the equal if not the superior of anything else on the road and, in one respect, it represents a significant step forward. Cadillac's system actually watches the driver, a considerably more sophisticated and  safer solution than Tesla and Mercedes Benz's nudge-the-wheel, and far more elegant than Google's just-give-up approach to the driver interface.

We can also conclude that that both the executives and the legal department of General Motors are fairly comfortable with this technology.When Tesla started rolling out its autonomous technology, there was always the possibility that Elon Musk was just betting big and playing reckless. That is certainly not the case here, which leads us to challenge one of the most cherished beliefs of the narrative.

The one thing everyone, from the trendiest tech blog to the New York Times, knew about autonomous vehicles was that the primary challenge was not technology and engineering, but the regulation and legislation.There would have to be massive and inevitably glacial changes in both of these areas before things could move forward. In other words, we couldn't get any cool stuff until those darn bureaucrats got out of our way. It was a narrative with almost archetypal appeal, but we can now say with some confidence it was never true. Multiple car companies are rolling out driverless options despite no great bureaucratic shift because bureaucrats were never a problem. Up until now, we didn't have driverless cars because the technology was not good enough.

Another cherished myth that is seriously threatened is that of the Silicon Valley Savior. It was often assumed that anything as wonderful as a car that drove itself would have to come from that most magical places. You couldn't expect anything wondrous and disruptive to come from a Detroit or a Stuttgart. Thus, the contributions of companies like Google and Tesla were given a wildly disproportionate amount of coverage. Even Apple, each managed to do nothing except prove that it wasn't up to the challenge of making cars, got listed as a major player. This was always a distortion. Silicon Valley never had a lock on the research or even a clear lead.

The deep-seated belief in a magical Silicon Valley populated by powerful chosen ones who could will great advances to being played a large role in the enormous valuation of Tesla. At the risk of oversimplifying, the reasoning went something like this: amazing things are about to happen with automobiles; this sort of thing comes out of Silicon Valley; the only Silicon Valley company/chosen one in position to do these things are Tesla and Elon Musk.

That said, there was a case to be made for Tesla reaping huge profits from autonomous vehicle technology. Here Goldfarb lays out the details:

So is Tesla the next Honda, or the next Webvan? If one believes that Tesla is worth more than Ford or GM, one better have a decent theory of how Tesla is making current capabilities obsolete, why incumbents cannot replicate these capabilities, and whether Tesla will earn good margins in a post-Detroit world. In short, when betting on disruption, one has to identify how this disruption will occur — not just intone the magic word. I identify four bets that Tesla is plausibly making: the EV bet, the autonomous vehicle bet, a software-platform-in-a-car bet, or the clean-energy-and-battery-company bet. None of these bets appear particularly promising.



The autonomous-vehicle bet

One aspect of the software strategy is so important that it deserves its own separate category. Perhaps Tesla will win the autonomous vehicle race. Alphabet (a.k.a. Google) is famously in this space, and Intel just bought the Israeli startup MobileEye for $15 billion. Uber raided Carnegie Mellon to begin to develop this capability, and every major manufacturer is investing in automation technologies. This is a very competitive space. But it’s undeniable that Tesla is at the forefront of self-driving technology that’s already on the roads. Tesla is able to learn as the technology is used — and this may be Tesla’s secret sauce: the difficult-to-get information that is necessary to perfect autonomous vehicles.

Driving is one of the most dangerous activities most people do on a regular basis. The technological problem of building software and sensors that replicate what a good driver does is difficult — and based on understanding rare events. (One Tesla driver, Joshua Brown, thought the technology was more advanced than it was and died when his auto-piloted Model S failed to recognize a semi-trailer and his car slid under it.)

Of all of the theories involving Tesla and disruption, this is the most intriguing — and convincing: Tesla is the first to roll out an autonomous driving technology. Its use allows Tesla to learn and improve, leading to a technological leadership position. Since the technology relies on machine learning, and machine learning only gets better with more data, Tesla is then able to develop a leadership position that becomes difficult to assail — analogous to Google’s in search. Tesla is then able to monetize this dominant market position.

The trouble, once again, is that at least 33 other companies are also developing autonomous vehicle technology, and also gathering a great deal of data. Any successful strategy would require a quick rollout that would prevent competitors from catching up. Mercedes’ new models already have advanced driver assistance capabilities similar to Tesla’s. Cadillac is expected to deploy within months. But perhaps the software platform is integral to this. Perhaps the software bet blends with the autonomous-vehicle bet! Tesla recently showed off this capability when it curtailed some of the autopilot features of Teslas with a software update.

The presupposition is that Tesla has and maintains an insurmountable lead. With the Cadillac generation of self driving technology, it is not clear that Tesla has any lead at all. Just to be clear, this does not mean that Tesla is out of the race or that it won't go on to a long and profitable future, but it makes it increasingly unlikely that these profits will reach the sky high level needed to justify what people have been paying for the stock.

Friday, September 1, 2017

Good science reporting can be popular if it comes with enough obscenity and cute photo-shopped memes

A while back (the queue is long), Joseph and I were discussing the post I was about to run on the treatment of autism in that godawful New York Magazine piece. I argued that the risk of throwing out irresponsible and unlikely theories about the disease was that it was likely to encourage false diagnoses. Joseph counted that the real cost was in the wasted resources that would be expended on theories that had already been effectively ruled out.


Later that evening I got around to watching this Last Week Tonight segment on Vaccines and heard John Oliver making the same case as Joseph.