Tag Archives: Eric Schmidt

Electric Cars: EPA Serves Up Green Kool-Aid To Pair With Subsidies

03 Tuesday Oct 2017

Posted by in Environment, Subsidies, Technology

≈ Leave a comment

Tags

, , , , , , , , , , , , ,

Electric cars don’t save much energy over gas-burners if at all, at least for now. Warren Meyer’s recent Coyote Blog post on this topic is aptly titled “Why Is It So Hard To Get Even Smart People To Think Clearly On Electric Vehicle Efficiency“. Meyer begins by quoting the following tweet from Google smarty Eric Schmidt, which typifies the general level of public awareness regarding the supposed energy savings from electric cars produced by Tesla and many others:

“Electric motors are the unsung hero of clean energy – the latest are 97% efficient, vs. 45% for internal combustion.“

Meyer emphasizes these major points:

  1. the efficiency with which source fuels are converted to physical work via electric and gas-burning cars is more comparable than Schmidt’s tweet suggests;
  2. differences in energy density weigh heavily in favor of fuel-burning vehicles.
  3. the so-called miles-per-gallon equivalent (MPGe) calculated by the U.S. Environmental Protection Agency (EPA) is a sham.

First, Schmidt’s tweet is accurate only if the discussion is confined to simple conversion of energy to physical work performed by the respective engines. The tweet ignores energy losses that occur prior to that conversion: electricity must be generated with far less than 100% efficiency, mainly by burning coal and natural gas. In an earlier Forbes article, Meyer compares this situation to a distorted comparison of two refrigerator installers:

“In both cases the customer lives in a fourth floor walkup. The first installer finds the refrigerator has been left on the street. He has to … haul the appliance up four flights of stairs. After that, relatively speaking, the installation is a breeze. The second installer finds his refrigerator has thoughtfully been delivered right to the customer’s door on the fourth floor. He quickly brings the unit inside and completes the installation. So who is a better installer?“

The fact is that both gas-burning and electric vehicles rely heavily on fossil fuels. And, in addition to losses in the generation process, there are other losses of energy attributable to electric cars: transmission of power involves a significant energy loss, as does charging batteries and storage itself. Meyer considers only the extra losses from production and transmission of electricity in the following comparison:

“We take 97% times 90% transmission efficiency times 50% electricity production efficiency equals 43.6%.  This is actually less than his 45% figure.  By his own numbers, the electric motor is worse….“

Meyer qualifies this comparison, as some of his assumptions are of the “best outcome” variety, but contrary to Schmidt’s assertion, gasoline and electric engines are reasonably comparable in terms of energy efficiency.

Some contend, however, that power losses in electricity transmission are much larger than the 10% Meyer assumes (see the comments on his post). Battery charging involves a loss of perhaps 20%. And a replacement for a Tesla battery, post 8-year warranty, is $8,000 – $12,000, an additional storage “cost” that is virtually non-existent for gas-powered vehicles. Beyond a certain point in its life, that cost will have an impact on a Tesla’s resale value. Moreover, some contend that the production of electric vehicles is more energy-intensive, putting them in an energy efficiency hole right from the get-go.

Meyer then takes up the notion of storage density as an explanation for why early experiments with electric cars were essentially abandoned:

“15 gallons of gasoline weighs 90 pounds and takes up 2 cubic feet. This will carry a 40 mpg car 600 miles. The Tesla Model S 85kwh battery pack weighs 1200 pounds and will carry the car 265 miles (from this article the cells themselves occupy about 4 cubic feet if packed perfectly but in this video the whole pack looks much larger). We can see that even with what Musk claims is twice the energy density of other batteries, the Tesla gets  0.22 miles per pound of fuel/battery while the regular car can get 6.7. More than an order of magnitude, that is simply an enormous difference…“

Meyer notes in the Forbes article that the EPA calculates its MPG conversion factor for electric vehicles by dividing BTU’s in a gallon of gas by the BTUs in a kilowatt hour: 33.7 KwH per gallon. Thus, the EPA multiplies an electric car’s miles per KwH by 33.7 to arrive at the so-called MPG equivalent: MPGe. But as we’ve seen above, the conversion factor ignores the generation and transmission of electricity required at the front end, and the associated energy losses that occur before a single KwH is released by a Tesla battery.

Despite what we hear from the EPA, Tesla, and other interests today, electric cars have not really overcome these disadvantages, at least not yet. The EPA’s MPGe estimates are vastly inflated. Perhaps if they were accurate, these vehicles would not have to rely so heavily on taxpayer subsidies to be competitive. By extension, the presumed environmental benefits of electric cars are nonexistent at this stage of development. I’m certain that Eric Schmidt and many other smart people are capable of understanding these nuances, but they might be too busy tripping over their politics to bother.

Does Google Dominance Threaten Choice, Free Speech and Privacy?

29 Tuesday Aug 2017

Posted by in Censorship, Free Speech, monopoly

≈ Leave a comment

Tags

, , , , , , , , , , , , , , , , , , , , , , , , , ,

I’ve long been suspicious of the objectivity of Google search results. If you’re looking for information on a particular issue or candidate for public office, it doesn’t take long to realize that Google searches lean left of center. To some extent, the bias reflects the leftward skew of the news media in general. If you sample material available online from major news organizations on any topic with a political dimension, you’ll get more left than right, and you’ll get very little libertarian. So it’s not just Google. Bing reflects a similar bias. Of course, one learns to craft searches to get the other side of a story,  but I use Bing much more than Google, partly because I bridle instinctively at Google’s dominance as a search engine. I’ve also had DuckDuckGo bookmarked for a long time. Lately, my desire to avoid tracking of personal information and searches has made DuckDuckGo more appealing.

Google is not just a large company offering internet services and an operating system: it has the power to control speech and who gets to speak. It is a provider of information services and a collector of information with the power to exert geopolitical influence, and it does. This is brought into sharp relief by Julian Assange in his account of an interview he granted in 2011 to Google’s chairman Eric Schmidt and two of Schmidt’s advisors, and by Assange’s subsequent observations about the global activities of these individuals and Google. Assange gives the strong impression that Google is an arm of the deep state, or perhaps that it engages in a form of unaccountable statecraft, one meant to transcend traditional boundaries of sovereignty. Frankly, I found Assange’s narrative somewhat disturbing.

Monopolization

These concerns are heightened by Google’s market dominance. There is no doubt that Google has the power to control speech, surveil individuals with increasing sophistication, and accumulate troves of personal data. Much the same can be said of Facebook. Certainly users are drawn to the compelling value propositions offered by these firms. The FCC calls them internet “edge providers”, not the traditional meaning of “edge”, as between interconnected internet service providers (ISPs) with different customers. But Google and Facebook are really content providers and, in significant ways, hosting services.

According to Scott Cleland, Google, Facebook, and Amazon collect the bulk of all advertising revenue on the internet. The business is highly concentrated by traditional measures and becoming more concentrated as it grows. In the second quarter of 2017, Google and Facebook controlled 96% of digital advertising growth. They have ownership interests in many of the largest firms that could conceivably offer competition, and they have acquired outright a large number of potential competitors. Cleland asserts that the Department of Justice (DOJ) and the FTC essentially turned a blind eye to the many acquisitions of nascent competitors by these firms.

The competitive environment has also been influenced by other government actions over the past few years. In particular, the FCC’s net neutrality order in 2015 essentially granted subsidies to “edge providers”, preventing broadband ISPs (so-called “common carriers” under the ruling) from charging differential rates for the high volume of traffic they generate. In addition, the agency ruled that ISPs would be subject to additional privacy restrictions:

“Specifically, broadband Internet providers were prohibited from collecting and using information about a consumer’s browsing history, app usage, or geolocation data without permission—all of which edge providers such as Google or Facebook are free to collect under FTC policies.

As Michael Horney noted in an earlier Free State Foundation Perspectives release, these restrictions create barriers for ISPs to compete in digital advertising markets. With access to consumer information, companies can provide more targeted advertising, ads that are more likely to be relevant to the consumer and therefore more valuable to the advertiser. The opt-in requirement means that ISPs will have access to less information about customers than Google, Facebook, and other edge providers that fall under the FTC’s purview—meaning ISPs cannot serve advertisers as effectively as the edge providers with whom they compete.”

Furthermore, there are allegations that Google played a role in convincing Facebook to drop Bing searches on its platform, and that Google in turn quietly deemphasized its social media presence. There is no definitive evidence that Google and Facebook have colluded, but the record is curious.

Regulation and Antitrust

Should firms like Google, Facebook, and other large internet platforms be regulated or subjected to more stringent review of past and proposed acquisitions? These companies already have great influence on the public sector. The regulatory solution is often comfortable for the regulated firm, which submits to complex rules with which compliance is difficult for smaller competitors. Thus, the regulated firm wins a more secure market position and a less risky flow of profit. The firm also gains more public sector influence through its frequent dealings with regulatory authorities.

Ryan Bourne argues that “There Is No Justification for Regulating Online Giants as If They Were Public Utilities“. He notes that these firms are not natural monopolies, despite their market positions and the existence of strong network externalities. It is true that they generally operate in contested markets, despite the dominance of a just few firms. Furthermore, it would be difficult to argue that these companies over-charge for their services in any way suggestive of monopoly behavior. Most of their online services are free or very cheap to users.

But anti-competitive behavior can be subtle. There are numerous ways it can manifest against consumers, developers, advertisers, and even political philosophies and those who espouse them. In fact, the edge providers do manage to extract something of value: data, intelligence and control. As mentioned earlier, their many acquisitions suggest an attempt to snuff out potential competition. More stringent review of proposed combinations and their competitive impact is a course of action that Cleland and others advocate.  While I generally support a free market in corporate control, many of Google’s acquisitions were firms enjoying growth rates one could hardly attribute to mismanagement or any failure to maximize value. Those combinations expanded Google’s offerings, certainly, but they also took out potential competition. However, there is no bright line to indicate when combinations of this kind are not in the public interest.

Antitrust action is no stranger to Google: In June, the European Union fined the company $2.7 billion for allegedly steering online shoppers toward its own shopping platform. Google faces continuing scrutiny of its search results by the EU, and the EU has other investigations of anticompetitive behavior underway against both Google and Facebook.

It’s also worth noting that antitrust has significant downsides: it is costly and disruptive, not only for the firms involved, but for their customers and taxpayers. Alan Reynolds has a cautionary take on the prospect of antitrust action against Amazon. Antitrust is a big business in and of itself, offering tremendous rent-seeking benefits to a host of attorneys, economists, accountants and variety of other technical specialists. As Reynolds says:

“Politics aside, the question ‘Is Amazon getting too Big?’ should have nothing to do with antitrust, which is supposedly about preventing monopolies from charging high prices. Surely no sane person would dare accuse Amazon of monopoly or high prices.“

Meanwhile, the proposed Amazon-Whole Foods combination was approved by the FTC and the deal closed Monday.

Speech, Again

Ordinarily, my views on “speech control” would be aligned with those of Scott Shackford, who defends the right of private companies to restrict speech that occurs on their platforms. But Alex Tabbarok offers a thoughtful qualification in asking whether Google and Apple should have banned Gab:

“I have no problem with Twitter or Facebook policing their sites for content they find objectionable, such as pornography or hate speech, even though these are permitted under the First Amendment. A free market in news doesn’t mean that every newspaper must cover every story. A free market in news means free entry. But free entry is exactly what is now at stake. Gab was created, in part, to combat what was seen as Facebook’s bias against conservative news and views. If Gab or services like cannot be accessed via the big platforms that is a significant barrier to entry.

When Facebook and Twitter regulate what can be said on their platforms and Google and Apple regulate who can provide a platform, we have a big problem. It’s as if the NYTimes and the Washington Post were the only major newspapers and the government regulated who could own a printing press.

In a pure libertarian world, I’d be inclined to say that Google and Apple can also police whom they allow on their platforms. But we live in a world in which Google and Apple are bound up with and in some ways beholden to the government. I worry when a lot of news travels through a handful of choke points.“

This point is amplified by Aaron M. Renn in City Journal:

“The mobile-Internet business is built on spectrum licenses granted by the federal government. Given the monopoly power that Apple and Google possess in the mobile sphere as corporate gatekeepers, First Amendment freedoms face serious challenges in the current environment. Perhaps it is time that spectrum licenses to mobile-phone companies be conditioned on their recipients providing freedoms for customers to use the apps of their choice.“

That sort of condition requires ongoing monitoring and enforcement, but the intervention is unlikely to stop there. Once the platforms are treated as common property there will be additional pressure to treat their owners as public stewards, answerable to regulators on a variety of issues in exchange for a de facto grant of monopoly.

Tyler Cowen’s reaction to the issue of private, “voluntary censorship” online is a resounding “meh”. While he makes certain qualifications, he does not believe it’s a significant issue. His perspective is worth considering:

“It remains the case that the most significant voluntary censorship issues occur every day in mainstream non-internet society, including what gets on TV, which books are promoted by major publishers, who can rent out the best physical venues, and what gets taught at Harvard or for that matter in high school.“

Cowen recognizes the potential for censorship to become a serious problem, particularly with respect to so-called “chokepoint” services like Cloudflare:

“They can in essence kick you off the entire internet through a single human decision not to offer the right services. …so far all they have done is kick off one Nazi group. Still, I think we should reexamine the overall architecture of the internet with this kind of censorship power in mind as a potential problem. And note this: the main problem with those choke points probably has more to do with national security and the ease of wrecking social coordination, not censorship. Still, this whole issue should receive much more attention and I certainly would consider serious changes to the status quo.“

There are no easy answers.

Conclusions

The so-called edge providers pose certain threats to individuals, both as internet users and as free citizens: the potential for anti-competitive behavior, eventually manifesting in higher prices and restricted choice; tightening reins on speech and free expression; and compromised privacy. All three have been a reality to one extent or another. As a firm like Google attains the status of an arm of the state, or multiple states, it could provide a mechanism whereby those authorities could manipulate behavior and coerce their citizens, making the internet into a tool of tyranny rather than liberty. “Don’t be evil” is not much of a guarantee.

What can be done? The FCC’s has already voted to reverse its net neutrality order, and that is a big step; dismantling the one-sided rules surrounding the ISPs handling of consumer data would also help, freeing some powerful firms that might be able to compete for “edge” business. I am skeptical that regulation of edge providers is an effective or wise solution, as it would not achieve competitive outcomes and it would rely on the competence and motives of government officials to protect users from the aforementioned threats to their personal sovereignty. Antitrust action may be appropriate when anti-competitive actions can be proven, but it is a rent-seeking enterprise of its own, and it is often a questionable remedy to the ills caused by market concentration. We have a more intractable problem if access cannot be obtained for particular content otherwise protected by the First Amendment. Essentially, Cowen’s suggestion is to rethink the internet, which might be the best advice for now.

Ultimately, active consumer sovereignty is the best solution to the dominance of firms like Google and Facebook. There are other search engines and there are other online communities. Users must take steps to protect their privacy online. If they value their privacy, they should seek out and utilize competitive services that protect it. Finally, perhaps consumers should consider a recalibration of their economic and social practices. They may find surprising benefits from reducing their dependence on internet services, instead availing themselves of the variety of shopping and social experiences that still exist in the physical world around us. That’s the ultimate competition to the content offered by edge providers.

Mr. Musk Often Goes To Washington

31 Monday Jul 2017

Posted by in Automation, Labor Markets, Technology

≈ 1 Comment

Tags

, , , , , , , , , , , , ,

Elon Musk says we should be very scared of artificial intelligence (AI). He believes it poses an “existential risk” to humanity and  calls for “proactive regulation” of AI to limit its destructive potential. His argument encompasses “killer robots”: “A.I. & The Art of Machine War” is a good read and is consistent with Musk’s message. Military applications already involve autonomous machine decisions to terminate human life, but the Pentagon is weighing whether decisions to kill should be made only by humans. Musk also focuses on more subtle threats from machine intelligence: It could be used to disrupt power and communication systems, to manipulate human opinion in dangerous ways, and even to sow panic via cascades of “fake robot news”, leading to a breakdown in civil order. Musk has also expressed a fear that AI could have disastrous consequences in commercial applications with runaway competition for resources. He sounds like a businessmen who really dislikes competition! After all, market competition is self-regulating and self-limiting. The most “destructive” effects occur only when competitors come crying to the state for relief!

Several prominent tech leaders and AI experts have disputed Musk’s pessimistic view of AI, including Mark Zuckerberg of Facebook and Eric Schmidt, chairman of Google’s parent company, Alphabet, Inc. Schmidt says:

“My question to you is: don’t you think the humans would notice this, and start turning off the computers? We’d have a race between humans turning off computers, and the AI relocating itself to other computers, in this mad race to the last computer, and we can’t turn it off, and that’s a movie. It’s a movie. The state of the earth currently does not support any of these scenarios.“

Along those lines, Google’s AI lab known as “DeepMind” has developed an AI off-switch, otherwise known as the “big red button“. Obviously, this is based on human supervision of AI processes and on ensuring the interruptibility of AI processes.

Another obvious point is that AI, ideally, would operate under an explicit objective function(s). This is the machine’s “reward system”, as it were. Could that reward system always be linked to human intent? To a highly likely non-negative human assessment of outcomes? Improved well-being? That’s not straightforward in a world of uncertainty, but it is at least clear that a relatively high probability of harm to humans should impose a large negative effect on any intelligent machine’s objective function.

Those kinds of steps can be regarded as regulatory recommendations, which is what Musk has advocated. Musk has outlined a role for regulators as gatekeepers who would review and ensure the safety of any new AI application. Ronald Bailey reveals the big problem with this approach:

“This may sound reasonable. But Musk is, perhaps unknowingly, recommending that AI researchers be saddled with the precautionary principle. According to one definition, that’s ‘the precept that an action should not be taken if the consequences are uncertain and potentially dangerous.’ Or as I have summarized it: ‘Never do anything for the first time.’“

Regulation is the enemy of innovation, and there are many ways in which current and future AI applications can improve human welfare. Musk knows this. He is the consummate innovator and big thinker, but he is also skilled at leveraging the power of government to bring his ideas to fruition. All of his major initiatives, from Tesla to SpaceX, to Hyperloop, battery technology and solar roofing material, have gained viability via subsidies.

But another hallmark of crony capitalists is a willingness to use regulation to their advantage. Could proposed regulation be part of a hidden agenda for Musk? For example, what does Musk mean when he says, “There’s only one AI company that worries me” in the context of dangerous AI? His own company(ies)? Or another? One he does not own?

Musk’s startup OpenAI is a non-profit engaged in developing open-source AI technology. Musk and his partners in this venture argue that widespread, free availability of AI code and applications would prevent malicious use of AI. Musk knows that his companies can use AI to good effect as well as anyone. And he also knows that open-source AI can neutralize potential advantages for competitors like Google and Facebook. Perhaps he hopes that his first-mover advantage in many new industries will lead to entrenched market positions just in time for the AI regulatory agenda to stifle competitive innovation within his business space, providing him with ongoing rents. Well played, cronyman!

Any threat that AI will have catastrophic consequences for humanity is way down the road, if ever. In the meantime, there are multiple efforts underway within the machine learning community (which is not large) to prevent or at least mitigate potential dangers from AI. This is taking place independent of any government action, and so it should remain. That will help to maximize the potential for beneficial innovation.

Musk also asserts that robots will someday be able to do “everything better than us”, thus threatening the ability of the private sector to provide income to individuals across a broad range of society. This is not at all realistic. There are many detailed and nuanced tasks to which robots will not be able to attend without human collaboration. Creativity and the “human touch” will always have value and will always compete in input markets. Even if robots can do everything better than humans someday, an absolute advantage is not determinative. Those who use robot-intensive production process will still find it advantageous to use labor, or to trade with those utilizing more labor-intensive production processes. Such are the proven outcomes of the law of comparative advantage.