How Not to Manage Artificial Intelligence

You might want to take a look at Michael O’Hanlon’s post at The National Interest on the importance of artificial intelligence to U. S. defense:

A case in point is what our colleague at Brookings, retired Gen. John Allen, calls “hyperwar.” He develops the idea in a new article in the journal Proceedings, coauthored with Amir Husain. They imagine swarms of self-propelled munitions that, in attacking a given target, deduce patterns of behavior of the target’s defenses and find ways to circumvent them, aware all along of the capabilities and coordinates of their teammates in the attack (the other self-propelled munitions). This is indeed about the place where the word “robotics” seems no longer to do justice to what is happening, since that term implies a largely prescripted process or series of actions. What happens in hyperwar is not only fundamentally adaptive, but also so fast that it far supercedes what could be accomplished by any weapons system with humans in the loop. Other authors, such as former Brookings scholar Peter Singer, have written about related technologies, in a partly fictional sense. Now, Allen and Husain are not just seeing into the future, but laying out a near-term agenda for defense innovation.

The United States needs to move expeditiously down this path. People have reasons to fear fully autonomous weaponry, but if a Terminator-like entity is what they are thinking of, their worries are premature. That software technology is still decades away, at the earliest, along with the required hardware. However, what will be available sooner is technology that will be able to decide what or who is a target—based on the specific rules laid out by the programmer of the software, which could be highly conservative and restrictive—and fire upon that target without any human input.

To see why outright bans on AI activities would not make sense, consider a simple analogy. Despite many states having signed the Non-Proliferation Treaty, a ban on the use and further development of nuclear weapons, the treaty has not prevented North Korea from building a nuclear arsenal. But at least we have our own nuclear arsenal with which we can attempt to deter other such countries, a tactic that has been generally successful to date. A preemptive ban on AI development would not be in the United States’ best interest because non-state actors and noncompliant states could still develop it, leaving the United States and its allies behind. The ban would not be verifiable and it could therefore amount to unilateral disarmament. If Western countries decided to ban fully autonomous weaponry and a North Korea fielded it in battle, it would create a highly fraught and dangerous situation.

IMO if we continue down the path we’ve followed for cyberwar, we’ve already lost this race. The United States does not produce the largest number of solitary geniuses (that would be Russia), it doesn’t do bureaucratic science better than anyone else (that would be Germany or Japan), and it can’t deploy enormous numbers of AI workers as China has in its approach to cyberwar.

For more than a century the distinct genius of the United States has been based on culture and economics. The A-bomb shortened WWII it didn’t end it. What won WWII was our farms and the car culture—all of the kids grew up using firearms and who knew how machines worked. Similarly, the U. S.’s failure to use our main strength, the large numbers of young people who’ve invested so much time in video games, and instead tried a top-down approach to cyberwar is a vision of things to come.

2 comments

What Is Ireland?

The United States isn’t the only country suffering from anomie. Check out Fintan O’Toole’s column at the Irish Times on how all of the narratives that have guided the Republic of Ireland have collapsed:

Nations tell themselves stories. They are not fully true, they are often bitterly contested and they change over time. But they are powerful: they underlie the necessary fiction that is “us”. And at the moment it is not quite clear what the Irish story is. What is the state of “us”?

The majority story, the narrative of modern Catholic Ireland, has had six different elements – components that have sometimes complemented, sometimes competed with, each other. The striking thing is that none of them really works any more.

If they were novels, we could call these stories, Survival, MOPE, The Scattering, Thoroughly Modern, Top O’ the World, Ma! and Who’s Sorry Now?

And we’re not alone. All of the ethnic states of Europe other than, as I noted yesterday, the newly-recreated ethnic states in the Baltic and the Balkans are going through similar identity crises.

I don’t know what the consequences of all of this will be. Largely, other than in the United States I think it’s none of my business. I’ll just pass along an anecdote.

My best high school buddy is of 100% Irish ancestry. All four of his grandparents were born in Ireland. In her 80s his paternal grandmother, who had left Ireland when in her teens, returned to Ireland to visit relatives. She returned in some disgust declaiming that all of the Irishmen like those she remembered were in the United States.

2 comments

How Far Back?

At Salon Paul Rosenberg argues for rolling back the “mistakes of the Clinton era”:

Much of the internal strife inside the Democratic Party today hearkens back to policy choices made in the 1990s, as the Clinton administration firmly wedded the party to a neoliberal policy agenda — on trade, deregulation, welfare reform, mass incarceration, and so on — much of which is finally getting the sort of sustained critical scrutiny it escaped at the time. While neoliberal ideas from the 1990s have been the source of significant strife, some problems facing us have solutions rooted in that decade as well.

The same intellectually restricted environment that made those Clinton-era compromises seem sensible also excluded insights, ideas and perspectives that could have served us far better at the time, and — more importantly — can still help us get a handle on some of the most challenging problems we face today. They are well worth re-examining and integrating into any progressive to-do-list for the years ahead.

The changes he would like to see include either abrogating or reworking NAFTA (in agreement, apparently, with Donald Trump), revising energy and environmental policies, and electoral reforms that I don’t recall having changed during Bill Clinton’s term of office. Conspicuous by their absence are reforms to the tax code (outsized executive compensation was an unforeseen consequence of Clinton era reforms to the corporate income tax), granting China “Most Favored Nation” trading status, and sponsoring China’s admission to the World Trade Organization.

His position leads me to the question how far do progressives wish to turn back the clock? You don’t have to search very hard to find Paul Krugman’s nostalgia for the pre-Kennedy tax code. Democrats have been pretty successful in turning back the clock to the Eisenhower era when Adlai Stevenson lost two consecutive presidential bids.

The Truman era when Harry Truman nationalized the steel industry? FDR’s first term when he instituted the National Recovery Administration, Works Project Administration, and Civilian Conservation Corps?

All of which underscores a point I’ve made before: Democrats have become conservatives while Republicans have become radicals.

0 comments

The End of the Peace

I think I can summarize Richard Haass of the Council on Foreign Relations’s post at Project Syndicate in a single sentence. Can the Asian economic “miracle” survive the end of the Pax Americana? Not only do I think the answer is “No” but IMO it’s no miracle. China in particular has followed the trail blazed by the Soviet Union with a few improvements but has run into the same problems as the Soviets did, too. Unless the authorities are willing to release control of the reins, I suspect they’re headed for problems. China presently has excess productive capacity enough to satisfy the entirety of world demand in multiple sectors of the economy for years to come.

0 comments

The War Between Post-Modernism and Liberal Democracy

In his remarks about the aftermath of the events in Charlottesville at The American Interest, Damir Marusic notes:

The 1960s, however, saw identity politics gradually emerge as a revolutionary force on the Left. Starting with Civil Rights, through feminism, and on to LGBTQ activism today, with each successive breakthrough the logic of the movement has become embedded in the thinking of an ever-wider segment of the Left, to the extent that today it is taken for granted by many. These movements have sought justice for oppressed groups by increasingly relying on mechanisms gleaned from a radical postwar political philosophy explicitly intended to serve as a critique and rejection of the Enlightenment in the shadow of the Holocaust: postmodernism.

One can easily get lost in the minutiae of these philosophies and forget the bigger picture, which is that the politics of postmodernism are ultimately incompatible with liberal democracy. Since it got its start as a radical form of literary criticism, postmodernism is a philosophy of competing “narratives” that sees dominant ones violently suppressing weaker “others” as part of an endless zero-sum competition that leaves no room for meaningful political compromise. The struggle ends up being not between ideas, but between groups that have to varying degrees been repressed, each with its own set of contingent “truths.” To challenge any of these truths on objective grounds represents a mortal threat, an attempt by “hegemony”/”patriarchy”/”capital” (take your pick) to “silence” the weak, to deprive them of their very ability to exist. Even at its least violent—when it is not calling for the overthrow of the dominant “narrative” but rather asking for the space to have a thousand (identity) flowers bloom—postmodernism doesn’t allow for any kind of positive, constructive politics. Everything boils down to an absolute struggle between oppressor and oppressed. There is no room for a common positive vision in such a Manichean world.

Violence is not speech. The proper response to speech including speech you despise is either ignoring it or persuasion. That persuasion requires knowledge, skill, and patience is no excuse for violence.

The proper response to violence is enforcement of the law. That enforcing the law is difficult and expensive and requires skill, courage, and determination is no excuse for failing to enforce the law.

Which side are you on?

15 comments

The Awful Truth About North Korea

Maybe I’m seeing things that aren’t really there but in a recent post at the RAND Blog I see hints of an explanation of why so much tripe has been said and written by so many otherwise intelligent people about North Korea. For example, quoting North Korea scholar Bruce Bennett:

Bennett has a story he likes to tell as a stage-setter in his briefings. Tensions were running high in the early 1990s. The leader of North Korea at the time, Kim Il Sung, called together his senior military officers and asked if they could win a war against the United States.

“Now remember,” Bennett says, “these were North Korean military people. What would they say? ‘We’re gonna win, yeah!’ But then he asked them: ‘If we lost, what do we do?’

“The North Korean military guys were all smart enough to know that was a really good time to keep your mouth shut. But his son, Kim Jong Il, the father of the current leader, spoke up and said, ‘If we lose, I will be sure to destroy the earth. What good is the earth without North Korea?’”

which provides a hint as to North Korean grand strategy. Their interests aren’t limited to regime survival. The United States hasn’t threatened the existence of the Kim regime in more than 60 years other than by our very existence. They want Korean reunification, presumably by force, and a sort of aggressive “Samson option” to prevent the United States (or China) from intervening against them.

The present North Korean god-king, Kim Jong Un, is no Kim Il Sung. He isn’t even a Kim Jong Il. He’s crazier and less competent than either of them.

More:

He recognizes North Korea’s nuclear strategy as a Cold War throwback in part, a concept known as decoupling. If North Korea can threaten mainland America, then it can raise the stakes for any American intervention on behalf of South Korea. As Bennett puts it: “Are we prepared to trade San Francisco for Seoul?”

I don’t think that’s the real question. The more relevant question is whether the residents of Honolulu, Anchorage, or San Francisco are willing to trade their cities for Pyongyang? Here’s another question: how will the present regime react to a threat to its existence from within?

My view continues to be that while we should be prepared to respond to attacks, preferably with a preemptive attack of overwhelming force, we shouldn’t respond to risks as though they were issues. Diplomacy has failed and is less likely to be successful with a North Korea led by Kim Jong Un than it was under Kim Jong Il. Strategic patience has been a complete flop. As long as we’re not willing to take steps to motivate China to stop supporting the Kim regime, we can’t ourselves mitigate the risks that North Korea presents and we shouldn’t pretend that we can.

7 comments

The Story of Estonia

At Country Studies there’s an interesting article on how Estonia recreated itself as an ethnic state in the 1990s:

In Estonia’s fight to regain independence, the overall strategy of asserting the country’s legal continuity as a state clearly had paid off. Yet, in terms of offering a path for the future, this strategy had many complications. One of these was the question of what to do with the 500,000 mostly Russian, Soviet-era immigrants living in Estonia. In 1990 the Congress of Estonia had been the first representative body to lay down the principle that because these people had settled in Estonia under Soviet rule, they were not automatically citizens of the legally restored Estonian state. Rather, under independence they would have to be “naturalized” on the basis of specific language and residency criteria. This position was also argued as a means of better integrating the mostly Russian noncitizen population, the majority of whom did not speak Estonian. In mid-1991, as the independence struggle seemed to languish, the Estonian government, led by Prime Minister Edgar Savisaar, showed signs of readiness to compromise on the citizenship issue in order to gain more local Russian support. However, after the failed August coup and the immediate onset of full independence, the Congress and other radical groups were emboldened to insist on the principle of restricted citizenship. Thus, the Supreme Council decided on November 11, 1991, to require the naturalization of all Soviet-era immigrants to Estonia while automatically renewing the citizenship of all prewar citizens and their descendants. In February 1992, the parliament set naturalization terms, which included a two-year residency requirement, the ability to speak conversational Estonian, and a one-year waiting period after applying.

That’s a bit of an oversimplification. There have been Russians living in Eastern Estonia for most of the last millennium and continuous settlement by Russian “Old Believers” for nearly half that. It was only after the conquest and occupation of the Baltic by the Russian Empire in the 18th century that large numbers of Russians began moving into the Baltic. When Estonia became an independent republic in the chaos of the aftermath of the First World War and Russian revolutions, Russians were about 8% of the population of Estonia. After 1940 when the Soviet Union annexed the Baltic countries, Russians began moving into Estonia again.

The language requirement was no small hurdle. Although Latvian and Lithuanian are not Slavic languages and are about as hard to learn for a Russian as they are for an English speaker, they are at least Indo-European languages. Lithuanian has been said to be more closely related to Sanskrit than it is to Russian. Estonian is an agglutinative Finno-Ugric language and very difficult for native speakers of Russian or English to learn. The Foreign Services Institute rates the difficulty of acquiring it as “significant”, meaning it requires at least a year of intensive full-time study. The only languages that are more difficult are languages like Arabic, Chinese, or Japanese. In the cases of Arabic and Chinese their difficulty is due to their orthographies. Mandarin Chinese is actually fairly easy to learn to speak but in order to become truly literate in it you’ve got to learn not only a modern foreign language (Mandarin Chinese) but also Classical Chinese. Similarly with Arabic. It’s not surprising that many native speakers of those languages are only actually partially literate.

Anyway, read the whole thing. I found it interesting. If you like law, sausage, or, apparently, nation-states, you shouldn’t watch them being made.

3 comments

The Empirical Evidence for Natural Selection

At Science Friday Jonathan Losos recounts the story of how two biologists studied the Galápagos finches for more than thirty years and recorded empirical evidence of natural selection:

Starting in 1973, the Grants spent several months each year on the small, crater-shaped Galápageian island of Daphne Major. Their goal was to study the population of the medium ground finch (so named because there are both larger and smaller ground finch species) to see whether and how the population changed from one generation to the next and to attempt to measure natural selection driving such change.

To do so, the Grants had to capture and measure all of the finches on the island every year. Only in that way could they see if the characteristics of the population—body mass, beak size, wing length, and so on—were changing from one generation to the next.

Read the whole thing. It’s an interesting story of the painstaking, detailed way in which real science is actually done.

5 comments

Ignoring the Obvious

Although I found the article at RealClearScience, “Five Elements and Substances That Turned Out Not to Exist”, interesting and informative, after reading it I was surprised to find that the author failed to include the best known example of a substance that didn’t exist: the ether.

3 comments

Proof By Cherry-Picking

I thought that Cathleen O’Grady’s Ars Technica post on the health benefits of relying increasingly on wind and solar power generation was an interesting tack:

A paper in Nature Energy this week dives into the weeds by trying to estimate the economic benefits of wind and solar power across the whole of the US. Berkeley environmental engineer Dev Millstein and his colleagues estimate that between 3,000 and 12,700 premature deaths have been averted because of air quality benefits over the last decade or so, creating a total economic benefit between $30 billion and $113 billion. The benefits from wind work out to be more than 7¢ per kilowatt-hour, which is more than unsubsidized wind energy generally costs.

I can’t read the actual journal article but I hope that it takes a less, well, selective approach to realizing its findings than appears to be the case.

At least 90% of all solar panels are made in China and the percentage of solar panels utilizing materials produced in China approaches 100%. That means that you cannot evaluate the health implications of using solar energy without taking the health implications of their production in China into account. Further the authors need to extend their investigations to a total lifecycle study. What are the health implications of the rising mountain of faulty or end-of-life solar panels? The health implications of solar energy begin before the panels are installed and continue after they’re taken out of service.

Additionally, very few people believe that we’ll be able to convert to 100% wind and solar in the foreseeable future. What are the health implications of fractional implementation? They’re unlikely to be linear.

I’m not hostile to the possibility that wind and solar energy may be healthier. I just think that we’ll never know unless we approach the question more critically.

4 comments