Monday, July 17, 2017

Silicon Valley Wants Inside Your Head—Literally

A recent article in the engineering professional's magazine IEEE Spectrum reveals that several powerful Silicon Valley entrepreneurs are sponsoring initiatives to breach the barrier separating our brains from the rest of the world.  They all fall into the category of "brain-computer interfaces" or BCIs. 

For example, Facebook wants to develop a noninvasive (meaning you don't need surgery to wear it) system that would let you type five times faster on your smart phone than you do now.  A former Facebook executive named Mary Lou Jepsen is trying to develop an MRI-type device that will "interpret the patterns of neural activity associated with thoughts"—mind-reading, in other words.  Elon Musk, true to form, has thrown caution to the winds with his program to implant a sensor in your brain, bypassing the old-fashioned eyes, ears, and fingers and mainlining the Internet straight to your hippocampus, or wherever the thing will be attached. 

There are two things I'd like to say about these projects.  One is technical, and the other is moral.

The technical aspects of BCI projects are daunting, to say the least.  While some research has been done already into ways of communicating with the brains of people with "locked-in" syndrome (e. g. sufferers from Lou Gehrig's disease who can no longer move any voluntary muscles), progress has been slow and the systems have been customized to each individual.  The brain is the final frontier of biology, in that it is the most complex organ known and probably the one we know the least about in comparison to what there is to know—which, in a sense, is all human knowledge, since all human knowledge is, materially speaking, contained in brains.  The self-reflexive nature of brain research makes me wonder if there isn't something analogous to Gödel's incompleteness theorems at work in the brain's attempt to understand how the brain works. 

Mathematician Kurt Gödel showed in 1931 that every mathematical system of a certain complexity is bound to have statements in it that cannot be either proved or disproved without going outside the system.  The brain analogy of this is that the brain may not be able to understand exhaustively everything about itself. 

Whether or not that is the case is a purely speculative question at this point—just the kind of issue that the Silicon Valley types are not interested in.  They want to do something with the brain, not understand it, and their research is way toward the development end of R&D, with explicit timelines and the whole apparatus of high-tech development programs favored by those with essentially infinite amounts of cash.

What a contrast it is to the way some wealthy corporations used to behave.  Physicist Mark P. Mills points out in a recent article in the journal New Atlantis that U. S. corporations spend only about 7% of their total R&D money on basic research, which the government's Office of Management and Budget defines as "study directed toward fuller knowledge or understanding of the fundamental aspects of phenomena and of observable facts without specific applications toward processes or products in mind."  Mills makes the telling point that while the basic-research labs of the old pre-breakup Bell System and IBM can count thirteen Nobel Prizes to their credit, the free-spirited pursuit of knowledge wherever it leads is no longer in favor in the U. S. corporate world.  Though the lag between a discovery and the awarding of a Nobel Prize can often be decades, Mills looks in vain for any comparable scientific achievements from the tightly-application-focused "moon-shot" projects currently favored by Silicon Valley.

The technical point here is that those pursuing BCIs may have bitten off more than they can chew, and the nature of the problem might require a longer-term, less focused perspective.  Even if the goal of brain-computer interfaces is worthy of pursuit, we may be in for a long marathon instead of a sprint.

Now for the moral issue.  Is it right to read another person's mind?  Especially if they are not fully aware of what is involved in the process?  Ah, the corporations say, we would never do such a thing without your consent.  Yes, I reply, the same kind of consent I give whenever I load a new piece of software on my computer and lie that I have read and understood eight pages of legal gobbledegook when I click the button that will let me load the software. 

We have already been trained to allow snooping at a scale that twenty years ago would have been regarded as outrageous.  Everyone who gets online has probably had the experience of doing a web search for a consumer item in one place, only to find ads for it popping up later during a completely unrelated activity.  A combination of cookies and data-sharing among Internet companies on a grand scale means that privacy, at least when it comes to things you search for online, is mostly a thing of the past. 

Should we let the greedy hands of the Internet reach into the last remaining sanctuary of privacy, the human mind itself?  I am reminded in this connection of a passage in one of C. S. Lewis's Chronicles of Narnia series, The Voyage of the Dawn Treader.  In it one of the English children transported to Narnia is named Lucy, and at one point she is alone in a magician's house, perusing a great book of magic.  She comes upon a spell "which would let you know what your friends thought about you."  She says the magic words, and a kind of television process shows her two friends of hers in a train.  She hears them talking about her, and not in a nice way, either. 

A bit later, Aslan the Lion appears, and says to her, "Child. . . I think you have been eavesdropping."  When she replies that she didn't think it counted as eavesdropping if it was magic, he replies "Spying on people by magic is the same as spying on them in any other way."  I don't know how popular the Chronicles are in Silicon Valley, but it's just possible that a moral lesson a child could understand needs to be taught to some of our most powerful technical leaders.

Sources:  The IEEE Spectrum article "Silicon Valley's Latest Craze:  Brain Tech" by Eliza Strickland appeared on pp. 8-9 of the July 2017 print issue.  The Spring 2017 edition of The New Atlantis carried Mark P. Mills' article "Making Technological Miracles" on pp. 37-55.  I also referred to the Wikipedia articles on Gödel's incompleteness theorems and the IBM Zurich Research Laboratory.  Lucy's exploit with magic is found on pp. 131-135 of the Macmillan paperback edition of The Voyage of the Dawn Treader, originally copyrighted 1952.

Monday, July 10, 2017

Does the U. S. Need a New Star Wars Program?

On the Fourth of July last week, the world saw one rocket's red glare that wasn't fired in celebration:  North Korea launched the latest in a series of intercontinental ballistic missile (ICBM) tests.  The timing was intentional, and the North Korean news agency quoted its leader Kim Jong-un as saying, "The American bastards must be quite unhappy after watching our strategic decision."  Not exactly diplomatic language.  Although the test missile went mostly straight up and down and landed harmlessly in the Sea of Japan, if directed toward the east, experts say it could have reached as far as parts of Alaska.  According to the New York Times report, it is unlikely that Pyongyang has a small enough nuclear weapon to fit on their ICBMs, but they seem to be devoting a great part of their pitifully small GNP to reach their ultimate goal of being able to threaten the continental U. S. with a nuclear warhead.

The North Korean government is one of the few remaining bastions of old-fashioned, dictatorial despotism, and rational behavior is not to be expected from them.  But missiles are. 

There are some parallels between this situation and the way the final years of the old Soviet Union played out.  When President Reagan announced his Strategic Defense Initiative ("Star Wars") in 1983, it arguably contributed to the eventual downfall of the USSR as that nation's confidence waned that they could counter the U. S.'s initiative with anything as effective.  As it turned out, Star Wars as originally planned never reached the deployment stage, but by 1991 the USSR had cracked apart, and it wasn't needed.

North Korea is different, in that they will probably never have more than a few viable nuclear ICBMs.  But even one nuclear bomb can spoil your whole day, so since about 2000 the U. S. has been developing a kind of mini-Star Wars system called the Ground-based Missile Defense (GMD). 

Shooting down even one ICBM on the fly is a very delicate undertaking that has been likened to shooting a bullet with another bullet.  Nevertheless, in 18 tests the system has successfully destroyed 10 targets.  These are not encouraging odds, but it's not bad for a system whose funding and support has fluctuated wildly over the years with the political climate in Washington. 

Austin Bay is a retired colonel in the U. S. Army Reserve whose service record goes back to the 1970s, and for some years has written regular columns on national affairs from a military perspective.  I have always found his viewpoints to be solidly grounded in factual information, and before the latest North Korean missile launch, Bay noted in a May 31 column that the GMD program was doing as well as you could expect in view of  the "sparse and fitful" testing it has had. 

Compare the record of an average launch of one per year for the GMD to the series of manned as well as unmanned spaceflights that took place in the 1960s, leading up to the moon launch:  14 launches (3 of which failed) for Project Mercury, 14 launches (two partial failures) for Project Gemini, and 10 successful flights that led up to the triumphal landing on the moon in 1969.  All this happened in only ten years, 1959 to 1969, which saw an average of nearly four launches a year.

The world and the U. S. were very different then, and the 1960s space program ate up a much larger proportion of the federal budget than Washington is likely to tolerate today.  But in North Korea's missile launches, we face a threat that is much less predictable than the old Soviet Union was, and one that could quite possibly lead to hundreds or thousands of American deaths in a nuclear attack.  This is serious business.

In contrast to what worked with the USSR, merely announcing a greatly expanded GMD is not going to make much of an impression on Kim Jong-un.  As Bay points out, the alternative to missile defense is diplomacy, and when the Clinton Administration made an agreement with North Korea in 1994 to quit making plutonium, evidence shows that the regime ignored us and went right ahead with their nefarious plans. 

It looks like North Korea won't quit rattling their nuclear saber until we grab every one they flaunt and crack it over our GMD-equipped knees, to stretch a metaphor.  But we can't afford to attempt a shoot-down of one of their missiles and miss—that would be worse than sitting on our hands.  Bay thinks, and I agree, that the time has come to get serious about ICBM defense, and that means a focused, well-publicized, and well-funded effort, as independent as possible of politics, to come up with a system that can be relied on to shoot down North Korea-style missiles, say at least 90% of the time. 

In the current fractious political atmosphere in Washington, such a plan is way down toward the bottom of most politicians' priority lists.  It may take a genuinely frightening incident such as an apparent attack by North Korea to motivate enough voters to call for protection.  But nobody (on our side, anyway) wants to go that far.

There are other things we can do about North Korea, but unfortunately most of them are not unilateral:  asking China to squeeze them a little, solidifying alliances with Japan, South Korea, and other eastern nations against the crackpot North Korea regime, and so on.  While China doesn't want its little neighbor incinerating the planet by mistake, it is much more tolerant of North Korea's human-rights abuses and other misbehavior than we can accept in the U. S., and there is little hope that the North Korean regime will change in response to anything that China does.

In the meantime, the U. S. needs to defend itself against attacks by foreign powers.  Everybody—Democrats, Republicans, Libertarians, you name it—agrees that defense is one of the bottom-line functions of the federal government.  However misdirected the defense budget has been in the past, the problem of North Korea won't go away.  We need to finish the job that the current GMD program has started, and develop and test it to the point that people in Alaska and the rest of the western United States can go to sleep without worrying that a rotund guy in Pyongyang is going to wake up one morning and decide to drop a nuclear bomb on their heads—and nobody can stop him.

Sources:  I referred to a report on the latest North Korean missile test in the New York Times carried on July 4 at  Austin Bay's commentary on previous GMD tests appeared on May 31, 2017 at  I also referred to Wikipedia articles on the Strategic Defense Initiative and NASA's 1960s space program. 

Monday, July 03, 2017

The Legacy of Hanford

One era's triumph can turn into another era's disaster, and perhaps no better example of that in the field of nuclear energy and weapons is the Hanford Site in south-central Washington State, about 200 miles from Seattle.  During the height of World War II, physicist Enrico Fermi designed a nuclear reactor for the Dupont Corporation to produce plutonium that was needed for nuclear weapons, as part of the ultra-secret Manhattan Project.  The small farming community of Hanford, Washington was selected for the site of the reactor and associated chemical processing plants, and more than 40,000 construction workers swarmed to the bank of the Columbia River in 1943 to build what became known after the war as the Hanford Nuclear Reservation. 

Because plutonium is one of the most deadly radioactive substances known, plant designers had to come up with novel ways of transporting large volumes of liquid and solid plutonium-containing material while keeping workers either far away from the load or behind several feet of radiation shielding.  Accordingly, one of the first industrial applications of closed-circuit TV was to view remote-controlled plutonium-handling equipment.  In view of the hazards of spills during transportation from the producing reactors to the processing plant, a railway tunnel was constructed of timbers and steel, buried in a foot or more of earth on top.  Plutonium that went into the "Fat Man" nuclear bomb used on Nagasaki, Japan probably passed through this tunnel, as did dozens of tons of plutonium used to make nuclear weapons during the Cold War.

Beginning in the 1960s, plutonium production ceased at Hanford, as it was realized that the site was heavily contaminated with long-lasting radioactive material and was no longer usable by then-current safety standards.  When the U. S. populace felt its back was to the wall during the war, not many people raised issues about long-term health hazards of working with nuclear weapons.  But as the threat of nuclear war declined after the Partial Test Ban Treaty between the USSR and the US in 1963, and especially after the collapse of the Soviet Union in 1991, most production activity ceased at Hanford and instead, a massive cleanup became the top priority.  The U. S. Department of Energy now spends billions of dollars a year on the Hanford cleanup, employing 8,000 people at the site and taking reasonable precautions about keeping workers safe.  But since President Trump's appointment of former Texas governor Rick Perry to head the Department of Energy, the media has paid more attention to the Department and any problems it may have, the most recent of which is the collapse of part of the roof of the old railroad tunnel used to transport plutonium.

The hole in the tunnel, more than ten feet across, was discovered on May 9, and as a precaution, many employees at the site were told to shelter in place until measurements could be taken to tell if substantial amounts of radioactive material had been released.  Investigation showed that no such release occurred, and the hole has since been covered in plastic and plans made to fill the old tunnel with grout.  Several railroad cars used to transport plutonium remain in the tunnel, which is altogether too radioactive to be inspected by humans, although robotic inspections are possible.  A second larger tunnel built in the 1950s has also shown signs of structural instability, and Hanford managers are planning to do something about preventing its collapse by August.

It would be nice if engineering ethics consisted of a set of unchanging rules, and doing engineering ethically simply meant understanding and following the rules.  But a phrase I recently came across expresses nicely the difference between the discipline of ethics and the disciplines of the hard sciences. 

Ethics is a "humane science"—meaning not that it's kind to animals, but that its "laws" are really just generalizations that depend on the nature of humanity, and so cannot show the ironclad reliability and constancy of physical laws.  This is not to argue for relativism—the notion that all ethical principles are relative to particular times, places, and cultures.  Rather, it is to confess both ignorance—no finite human being can possibly know all the relevant considerations in a particular ethical situation—and the fact that as human cultures and societies change, what is regarded as ethical behavior in a given circumstance can also change. 

In the case of Hanford, what has changed the most is our sense of priorities.  In 1939, the U. S. suspected Hitler of building a nuclear weapon, and Japanese troops were showing signs of fighting to the last man on the last domestic island of that nation.  For good or ill (plenty of both, actually), Roosevelt gave the green light to the Manhattan Project, which led to the first production and use of nuclear weapons six years later.  Both leaders and ordinary citizens felt seriously that the U. S. was fighting for its life, and in such a situation, concerns about exposures to levels of radiation that might possibly lead to cancer in twenty or thirty years, or might pollute the environment for hundreds of years, simply faded into the background.

Having enjoyed relative peace in the North American continent ever since the end of World War II, the U. S. can now afford to deal with the messes it created during the war, Hanford being the leading example.  Many opponents of nuclear power take the acres of lethal radioactivity at Hanford to be proof sufficient to lead us to swear off all use of nuclear power forever, amen.  And it must be admitted that disasters such as the 1986 Chernobyl nuclear-reactor fire in Ukraine are uniquely horrible.  Shutting down all nuclear plants would presumably avoid such incidents in the future. 

But nuclear energy is also uniquely suited to address the increasingly prominent issue of global warming.  While it is an open question whether renewable energy can compete economically with nuclear energy for the world's short-term energy needs, it would be shortsighted to rule nuclear out altogether because of an emotional reaction against it not based on an objective view of the facts.  Unfortunately, there are lots of facts to view, and so nuclear power remains controversial, as it probably always will simply because its first public use was to bring us the horrors of nuclear war. 

Sources:  I referred to news reports on the Hanford tunnel-roof collapse carried by the Washington Post on May 9 at, and the Seattle Times on June 30 at  I also referred to the Wikipedia article on the Hanford Site.