Thursday, November 9, 2017

20/20 Vision

...

It seems abundantly clear at this point that as long as the American government is controlled by the Trump/Ryan nexus, Democrats are going to have a significant popular advantage throughout the electorate. Between the embarrassing, buffoonish Trump with a demonstrated disinterest in policy and process, and the almost comic-book villain personification of Ryan as he pursues Republican policy goals that are universally loathed by 80% of the electorate, the political ads just write themselves and in all but the reddest constituencies (you know, the racist ones) and the Democratic candidate starts with a powerful built-in advantage.

I still don't think the Democrats will recapture the majority in the House of Representatives, but it's hard to imagine they can't re-take the Senate and significantly tighten up the House. And with another year to go, Trump (or Mueller) could easily change the dynamic so even the House is in play.

But seriously. We need to start thinking about 2020. In the run-up to 2016, there were people who said that maybe it would be best in the longer-term if Trump were to win, burn the existing structures to the ground, and open the way for a liberal government to truly take power. I thought this was a horrible idea - certainly plausible, but it just seemed to me the costs would still be too high, no matter the longer term outcome. Well, as it turns out we ended up with this experiment in political destruction by government incompetence and corruption, so what the hell. Maybe it will all turn out for the better.

But we need to think seriously about candidates. It's true - I'll grudgingly admit - that Hillary Clinton was not a great politician. She ended up in a leadership role almost accidentally, did a wonderful job in her political roles, but faced with a relentless non-stop attack from both the Republican Party and the media, attacks that were encouraged by an effective cyber-influence campaign operated by a foreign adversary, and ultimately brought down by a federal police force that wanted to end Democratic leadership in the White House, a victory might have been just a little more than we could have expected. And even with that, she won the election, only losing by a tiny rounding error worth of votes that happened to be in exactly the right place. We have to do better.

I don't like Bernie Sanders. I think he's a flim-flam man, a fast talker who happens to share (to the extent we actually know them) my preferred policy goals. But he shows no interest in the details, and he's afraid to even suggest funding or regulatory details because he's terrified he'll be challenged on them. His government funded universal free health care for all is a nice little slogan, but he never engaged honestly with how he'd pass it, how he'd fund it and how he'd implement it. His numbers were clearly wrong, but when asked about that he attacked rather than engaged. I don't really like Paul Ryan, and a liberal Paul Ryan doesn't do much more for me. We have to do better.

I'm a politics geek, a gearhead who looks at policy function, economic distributional outcomes and effective government intervention in the private sector. I don't have any idea who the Dems should nominate - particularly in light of the chaos and suspicious swirling around the Republicans that obscures who their candidate might be - but I do think it's a conversation we need to be having. I don't believe we need to move radically to the left - most Democrats, if they had the power, would pass effective, liberal, technocratic legislation that would solve problems without disastrous unintended consequences. A big key will be nominating the anti-Trump - a clean, brilliant, well-spoken professional without a whiff of corruption. Yeah, another Barack Obama. He's proof that we CAN do better.

There will be challenges in 2020, but they won't be ideological. Trump will have thoroughly destroyed the 'conservative' brand, and most any Democratic candidate should win any race not dominated by racists. One big key will be keeping the Russians from driving the narrative. Another - related in many ways to the first - will be making sure that the primary elections and nominating process aren't used to divide the Democratic vote.

Mainly, we have to recognize 2016 for the outlier that it was. The longer we on the left continue to rend our garments and re-litigate the primary, accuse each other of everything from dishonesty and corruption to socialism and something nebulous we call 'neo-liberalism' (which is apparently bad), the less we will be able to take advantage of the tailwinds that Trump is creating and intensifying. He's a clear and present danger, and we need to come together and agree that replacing him with a Democrat with as much legislative power as possible is critically important.
...

Friday, November 3, 2017

The End of Physics?

...
Our story so far...
Point 1.) The Large Hadron Collider in Europe, with a budget of almost eight billion euros, is the most powerful particle accelerator/collider in the world. It was built to replace FermiLab's Tevatron, the previous most powerful collider. Best known for discovering the Top Quark in 1983, the Tevatron was improved over its twenty year lifespan, eventually reaching collision energies of 1.8TeV (trillion electron volts) by colliding protons and anti-protons at 9oo GeV each. But it was recognized quite early on that, based on our understanding of the Standard Model, much higher collision energies and luminosity (essentially, the number of collisions recorded over a given period of time) were going to be necessary. When the US government cancelled the SSC being built in Texas in 1993, CERN's proposal to build the LHC was adopted and funded. After some frustrating delays, the LHC began it's initial run in 2009, with the intention to ramp up to collision energies of 13 TeV in 2015. The goal was to find out what could be learned about the fundamental characteristics of matter at energies and densities similar to what was extant in the immediate aftermath of the big bang. Specifically, the Higgs Boson, Dark Matter and Supersymmetry were expected to be discovered or disproved, and at long last the Standard Model would be completed.

Point 2.) The Standard Model can be comfortably thought of as the most successful failed theory in the history of physics. Every single prediction it makes has been proven, and yet we KNOW it is brutally incomplete. If we accept that gravity is one of the fundamental forces (along with electromagnetism, the weak nuclear force, the strong nuclear force), then we need to understand both how it is mediated and why it is so much weaker than the other forces. While the discovery of the Higgs Boson filled out the last box in the Standard Model, we are left with no operational theory of quantum gravity, and no clue as to what most of our universe is made of.

So now we're nearing the end of the 2017 data run at LHC. The joyful discovery of the Higgs is years behind us. And once again, despite some tantalizing events that turned out to be mere statistical anomalies, we come away with nothing. No Graviton. No dark matter. No Supersymmetric particles. No surprises, nothing that tells us anything beyond what we already know. And nothing to help us fill in the pieces we KNOW are missing. If gravity is a force, the gravitational field has to be mediated by a quantum force carrier. We KNOW that's how it works, but if we can't find that 'graviton', we can't understand gravity.

Meanwhile, outside the realm of CERN, physicists are just lost, stumbling around in the darkness. Without a ten billion dollar machine, they have no idea where to look for new discoveries. We've spent decades watching theoretical physicists indulge in the worst form of academic masturbation, creating dense, elegant mathematics that make no predictions and cannot be tested experimentally. Between string theory, inflationary cosmology and Supersymmetry, the larges portion of working physicists today aren't even doing science. They're working on unfalsifiable speculation that leads off in silly, pointless, untestable directions like the multiverse, or worse, that our entire universe is somehow just a digital simulation.

The way science is supposed to work is that the theorists postulate a theory, and the experimental scientists and engineers develop ways to test the predictions that the theory makes. If they start finding actual evidence that supports the theory, it gets plugged into the larger base of knowledge to make sure it fits in all the corners and doesn't fall apart at certain energies or conditions. Once enough evidence is gathered for the theory, it becomes accepted science. But that's just the beginning. Accepted science is boring. Now the theorists go back to work, trying to 'break' the theory - to find a reason why what we THINK we know is wrong. Because if we get something wrong, that means there's more to learn.

All of which brings us back to the Standard Model. We KNOW it's wrong - or at least incomplete. We know what it tells us, and we know there are things it's NOT telling us. And that's with a ten billion dollar mega-machine churning away at the problem. So whither physics? What happens if we learn nothing of consequence for years to come? How do you do fundamental research when the basic cost of knowledge is beyond your species willingness to pay the cost of 'basic science'? How do we get from where we are to the next discovery when we do not have the equipment to do the experiments? And, of course, what if we, as a species, decided to spend $25 or 30 billion on a new collider at much higher energies, and nothing changed?

I find it sad and frustrating to be living in such an amazing time, a time when we understand so much that no human ever understood before, and yet we seem to have no path forward. We're seeing breakthrough science in machine intelligence, computing power, data analytics and robotics, and that's both interesting and life changing. But we're stuck with the nagging thought that big science has picked all the low-hanging fruit and left us here, wondering what is just beyond the limits of our ability to discover it. And scientists keep going off in weirder and weirder directions because they don't have any way to do new science today.
...

Wednesday, November 1, 2017

Social Goddam Media - Somebody Really Should DO Something...

...
Watch your ass, Jesus,
she's got a BIG right hand
Russian intelligence agencies have not been sleeping for the last ten years. They have been watching the political and ideological polarization in western democracies, and they have noted the opening that provided a skilled and prolific propagandist in in the age of the internet. They understood Wikileaks, they understood Infowars, they understood the power that gave them to guide the narrative. But then, with the unfettered power to publish professional (or not-so-professional - it turns out it doesn't really matter) looking narratives directly to targeted audiences provided by Facebook and Twitter (among so many others), they discovered how easy it was to actually shape the domestic narrative. They found that by pushing made-up stories through Social Media, they could actually drive the entire American news cycle. That's right, they paved a direct highway from troll/bot farm to Fox News and MSNBC.

There's a number of reasons they chose to support Trump, but it's not unreasonable to consider that 2016 was a proof of concept exercise, an opportunity for 'live fire' testing of a new and powerful weapon of destabilization. Putin has long been furious with Bill Clinton's manipulation of the drunken and mentally debilitated Boris Yeltsin in the years after the fall of the Soviet Union. He also angrily - but carefully - observed the US sponsored 'Color Revolutions' in Georgia and Ukraine, and planned to defeat the Americans at their own information game. When Trump came along, with his nativist, nationalist and isolationist rhetoric, he was the obvious vessel to support with a full-spectrum cyber-campaign that might help him destabilize the EU and weaken NATO.

So the hackers used advanced phishing techniques to steal information, while the paid trolls and bots spread story after story smearing Clinton and encouraging anger and humiliation among Berniecrats.  Altogether, it was a quite effective campaign, using stolen data to amplify false narratives in a way that could be targeted against any set of demographic characteristics.

And now everybody's pissed.

So, with the next election a year away, what is to be done? Well, for all the spittle-flecked hatred being flung at the leadership of Facebook and Twitter, you'd think it would just be a matter of some bipartisan legislation banning them from...what, exactly? Political speech? Dishonest posts? Foreigners? Hmmm, maybe this isn't quite as easy as we thought.

Suppose we were aware of - and worried about - the power of social media to amplify specific messages and agendas long before the events of 2016 transpired? Well, that's problematic, because that phrase - "amplify specific messages" - is precisely the goal of online advertising. The platform was built to provide carefully segmented, highly targeted access to paying customers. The owners of the platform rely on those revenues, and would not set out to intentionally cripple them, regardless of the larger risk they represent. And a platform that goes around deleting posts and banning users isn't going to be successful for very long - the market would soon produce a platform who's sole raison d'ĂȘtre would be "we let you post and read what YOU want". Can you imagine ANY effective action that could have been taken to prevent an adversary nation from using social media tools like Facebook to undermine our democratic processes?

What role do we want government to take in limiting our exposure to propaganda? One thing I DO know - if you silence someone I hate, I'm happy, but if you silence ME (or someone I agree with) I'm furious. But that dynamic sort of prevents any action at all, doesn't it? Of course, Facebook isn't the government, so you have no absolute right of free speech on a digital communication platform, but what is the attraction of a platform like that when it seems to silence anyone at any time for anything? Wouldn't we immediately seek a platform that let us share our thoughts and beliefs without keeping such a heavy thumb on the topic scale?

Nobody's going to like it, but I know what the answer is. Foreign governments are going to have to become a little more careful in how they deliver targeted propaganda on social media platforms - hiding sources of funding, using better writers and grammar, perhaps being a little more selective in choosing false narratives - but beyond that, this is now a core feature of western democratic politics. From Brexit to Berlin, Washington to Paris, some of the most effective activists will be rabble-rousing on Facebook, and there's nothing that can be done that doesn't make the situation worse.
...

Saturday, October 28, 2017

Magic Number

...
We're going to have to deal with THESE people
One of the key lessons we learn as adults is the simple universal fact that "every action has consequences". As we move from childhood to adulthood, we discover that when we act impulsively, no matter how clear the short-term win might be, there are often unanticipated outcomes that make things worse for us than they were. To at least some extent we learn the value of 'thinking things through', trying to game out the unintended and second-order consequences of any given action, and revising our activities based on the conclusions of that process. And, of course, the larger the decision, the larger the benefits, the larger also are the consequences of acting on that decision.

All of which brings us to President Donald J. Trump. As the Special Prosecutor announced the first indictments of the investigation, and as Trump acts more and more delusionally and unpredictably, the calls for some kind of Constitutional process (or, in some cases, extra-constitutional process) for removing him from office get louder and more strident. It seems completely reasonable, perhaps even likely, that at least one of these actions will be implemented before the end of his term in January of 2021. Let's think through at least some of the possible consequences.

Trump has a built-in 35% support among the national electorate. That's nowhere near enough to win an election - but when we start talking about options for removing him from office, it is more than enough to destroy American democracy. How do you suppose this 'base' - people who have watched the first year of the Trump administration in all it's chaos, incompetence and corruption and still think he's the best president EVER - would react to the long, slow, grinding impeachment process, let alone the arbitrary and difficult 25th Amendment process? And do we believe, even for a moment, that Trump would sit back and allow these processes - which he would certainly view as 'attacks' on him and his family - to play out unimpeded?

Remember those rallies that formed the basis of the Trump presidential campaign? The undertones of hatred, bigotry and violence radiated from them like heat waves, with his most loyal supporters only a single word or phrase away from lashing out? Let's imagine a rally in, say, Mobile, Alabama, in early 2019. Articles of impeachment have been voted out of the House and the Trial is set for the Senate. Now you have a beleaguered Trump and his loyal supporters, angry and with a clear enemy. The attacks on the press and the Democrats and any Republicans that voted in favor of impeachment would make all that has gone before look like playground antics.

And then what? After impeachment, Trump is embarrassed, furious, filled with a seething rage and desire for revenge. Does anyone really think that he would hold back from encouraging everything from riots to outright insurrection? This 'base' of Trump's is somewhere between 50 and 60 million Americans, with at least 35 million who can be considered the hardcore center of whatever 'Trumpism' is. And to them it is victimization, anger, hostility and a carefully cultivated sense of existential angst. With the encouragement of the man himself, unmoored and with nothing left to lose, any hope of governance, any hope of a peaceful election in November 2020 is gone. All bets are off, and we won't come out of it the same nation we went into it.

If the process isn't impeachment, but rather the exercise of the 25th amendment, we know less about what we would see. The 25th has never been exercised before, and there would be a lot of guesswork and challenges to the process. And Trump would still have his bully pulpit, right through the entire effort. If the effort failed, you would have an unhinged, angry President with all the available power of his office looking for revenge. If it succeeded, you would have the same net outcome as you would post-impeachment: A wounded animal in a rage, with millions of equally enraged followers, all in the run-up to another attempt at a democratic transfer of presidential power.

Ultimately, I have no recommendations. Leaving him in power is a clear and present danger, not just to millions of lives and livelihoods, but to American small-d democracy itself. Removing him from power may be the thing that puts an end to the American experiment in liberal democracy. The things we've seen over the last twenty five years - from the media, from the GOP and in the technological changes that enabled so much more political manipulation - have led us to this point. We have truly 'painted ourselves into a corner', and there's just no solution that leaves things intact. It's impossible to see a path from here to 2021 that gives one any hope. But no matter what we do, we mustn't ever stop thinking. Indulging in a feel-good political undertaking without a clear idea of how the story ends is even more dangerous than the status quo.
...

Saturday, October 21, 2017

With a Little Help From My Friends

...
And This Time It's Legal
The thing that must be understood first is that our cyclical drug 'epidemics' are ALWAYS driven by supply. That is to say, there is not a pent-up demand for a certain drug waiting for the underground economy to supply it to broad portions of the population. Instead, somebody finds a way to bring very large supplies of a certain drug into the US and widely distribute that drug to cities and towns across the nation. Whether it was the Crack Cocaine epidemic of the 80s, the Methamphetamine epidemic of the Oughts, or the still ongoing Opioid epidemic, the underlying economic principle is backwards from most markets. It is the supply itself that creates the demand. The thing that makes the Opioid epidemic so pernicious is that the supply is not created by criminal elements, and there is no smuggling required. No, these drugs, and the demand for them that is killing and immiserating so many Americans are perfectly legal. They are produced very cheaply in massive numbers by regular pharmaceutical companies, and distributed through the normal channels to hospitals, doctors and pharmacies in every community in the US.

In many cases, I will often take the position that what the federal government does is unimportant and only marginally effective. Federal incarceration, is only a small percentage of overall incarceration, and nothing done at the federal level will have any impact on how criminal justice is meted out in the states. Similarly, in the case of education, while the federal government provides some funding, states by far have the most to say about how educational institutions are operated, and at what cost.

But this is a different matter. The FDA has complete control on how these drugs are manufactured and distributed, and working through the DEA they can prevent at least the most egregious misallocation of these kind of highly addictive drugs through the system. One of the most problematic parts of the supply chain is a local pharmacy that seeks to profit from the sales of Opioid drugs like Oxycontin, rather than to act in a manner that is responsible to the people of their community. Their greed is destroying communities all over the US. But at least we have the Government to step in and control these kinds of abuses, right?

Not so fast. When Donald Trump nominated Rep. Tom Marino (R-PA) to be Drug Czar, the press started looking into his history, especially with the Pharmaceutical industry. And sure enough, it turns out that he took over a hundred thousand dollars from Big Pharma, and acted on those contributions. Working directly with industry lobbyists - many of them former FDA and DEA inspectors and investigators, hired by the industry for huge salaries - Marino drove the passage of a law that brilliantly concealed its ultimate purpose. The law seemed so benign that it passed congress by unanimous consent - no voting required. It was only after its passage that the DEA discovered how it tied their hands when they discovered excesses that were clearly criminal.

One of the small-town pharmacies they detected as being a criminal drug distribution center was in Kermit, West Virginia, population 392. That single pharmacy ordered - and received - 9 MILLION pills in just two years. The DEA saw the reports, and wanted to investigate that pharmacy, but Tom Marino's law blocked them from doing so. And people died.

Of course, Marino, once exposed, could never be confirmed as 'Drug Czar', so he took himself out of the running. But we still need effective federal action to limit the greed-driven excessive shipments without limiting access to necessary treatment for pain. Big pharma - especially McKesson, Cardinal Health, and AmerisourceBergen - makes tens of billions of dollars on these drugs, and will resist anything that reduces those profits. Just like the gun lobby, the pharmaceutical industry has concluded that the profits are worth the loss of life, and the ruination of so many peoples lives, and will continue to push for ways to put MORE opioids into circulation, not less.

Watch congress closely. There is a lot of talk about the 'Opioid Crisis', but not a lot of action. Any action will take immense political courage, because the industry is a powerful foe. And if we've learned anything over the last ten years, it is that congress can bloviate and pose with the best of them, and at the end of the day do nothing. Once again, the personal risk to the individual lawmakers is of much deeper concern than the millions of American lives at risk.
...

Friday, October 20, 2017

Is A Robot Going To Take Your Job?

...
The Workplace of the Future
Not to give away the surprise ending, but the short answer is yes. An intelligent machine will almost certainly be doing the job you do today within a few years. But the main point I want to make today is a response to all the skeptics that make the argument that we've seen massive technological shifts in the workplace before, and after a few years of disruption they actually contributed to greater human employment than before. And that is inarguably true. But - and I generally HATE any kind of argument from authority - I would posit that the people who make this argument simply don't understand the depth and breadth of the sea change occurring right NOW.

The technological factors that are the structure behind the success of intelligent machines in the workplace are much more profound than many people realize. They are a combination of factors - hardware, software, sensors, high-speed wireless networks, virtualization, cloud-based computing, cheap custom silicon, modern programming methodologies, distributed databases - the list is long, and within each category is another list of breakthrough inventions and developments specific to that category.

It's also important to realize that the common term for these technologies - Artificial Intelligence (AI) - is a complete misnomer in this context. These machines are not intelligent in the way we understand it - they are not an 'artificial' equivalent of human intelligence. What they are is purpose-built machines that have access to the vast amounts of subject-matter specific data and a set of rules for applying that data in a very narrow context. A machine that can very quickly create an endless series of perfect welds on an assembly line, or that can interpret thousands of photographic (or X Ray) images per minute could never go to 7-11 and buy a soda. That simple act - one that can be accomplished by any six-year-old human - is well beyond the capacity of these kinds of 'intelligent machines'. Make no mistake, there's a lot of research being done on that kind of broad autonomous intelligence as well, but that's not what workplace automation requires - which is why it is arriving so quickly.

With capabilities provided by broad-spectrum sensors, huge distributed databases, fast processors, cheap memory, GPS, accelerometers, radar/lidar systems and custom designed chips that do specific things very effectively (more often than not, designed at least in part by other intelligent machines), there are very few jobs that cannot be done faster, better and more reliably by machines. They never get sick, they never lose focus, then never even want a coffee break. "But wait", you say. "Every other technologically-driven workplace revolution has resulted in MORE jobs for humans, not less. Why would this be any different?"

I'll explain that in very simple terms. Yes, this trend to use robots in place of humans will create more jobs throughout the workplace. But those jobs will be filled by other robots, not humans. And before you ask, it will not continue to be necessary for humans to design, build and program the intelligent machines. We already have machine-learning algorithms that can start with zero input and very quickly learn how to accomplish specific tasks better than anyone ever has. You may recall that Google's DeepMind research project produced a software Go player called AlphaGo that in March, 2016 stunned the world by beating a human player at the game. Well, they recently announced AlphaGo Zero. The new version started with NO data - just the game rules. It chewed on that for a while, and without any historical information on human strategies or processes, has now beaten the previous version in one hundred consecutive games. Just think about that.

Obviously, the political/socio/economic impacts of a world without jobs are profound. Ideally, this ability to create virtually unlimited productivity at only the cost of inputs should produce a Utopian society where everyone 'works' at what they want, and nobody lacks for material goods. Of course, our religious/puritanical culture, coupled with the greed and fear that makes so many of our fellow humans view economics in zero-sum terms will prevent us from getting anywhere near that kind of breakthrough for a long time. In the meantime, jobs at all parts of the workforce spectrum will become more scarce, people will be desperate and homeless and all the while, with the diminishing demands of labor, the owners of the capital that owns the robots will get richer and richer.

One would think the end of work would be a boon to a human community. And it almost certainly will. But the time between now and utopia will be a very dark period in human history. If there is a key takeaway, then, it is this: Do not bury your head in the sand. These changes are coming, and they are coming faster than you think they are. You (and your children) will be better served to at least understand what the next 20 years holds...
...

Thursday, October 19, 2017

A Few (Trump Inspired) Thoughts on Nuclear Weapons

...
Carrying the Football
One of the go-to arguments against a mentally-defective, impulsive buffoon in the White House is that he has access to the 'nuclear codes', and is the sole individual in the US Government hierarchy who can, without any other approval or concurrence, order the immediate release of American strategic nuclear weapons. An order that is processed by a system specifically designed and engineered to assure that  the order will be carried out within minutes, without any hesitation or option. This, we are continuously assured, is clearly a systemic defect, one we must reconsider and put further safeguards in place to prevent an angry president from being able to impulsively bring about the end of the world.

[Note: The President doesn't have any 'nuclear codes'. The famous 'football' that is always with him everywhere and at all times is a set of pre-developed Operational Plans that describe the weapons and targeting to be used against any given adversary. It also contains a set of one-time codes and ciphers that allow the president - using advanced cryptography - to authenticate his identity. Once the President issues the order and authenticates the source of the order is the National Command Authority, it is carried out by people who most assuredly DO have the nuclear codes.]

The reason for this seemingly dangerous - and make no mistake, it IS dangerous - system to be in place is based upon years of the study of nuclear weapons strategy. The idea is that nuclear weapons are not to be used - they are intended to deter any other nation's use of their strategic weapons. The number one overarching fear has always been a 'first strike'. If a nation, the thinking goes, believes it can launch a nuclear attack on a rival that eliminates the vast majority of that nation's ability to retaliate, the temptation would be to go ahead and launch. It would put the risk of global nuclear annihilation in the rear view mirror for all time.

This concern led to the streamlined command structure we have now. Russian (Soviet before) missiles coming over the pole can hit the American missile silos scattered across the northern prairie in less than fifteen minutes. The only way to deter such an attack, say the strategists, is to configure a system we call 'Launch on Warning'. If an incoming nuclear strike is detected, the entire system is designed to get OUR missiles in the air before THEIR missiles start landing. And that means one person in the chain of command, one fast decision, and the end of the world.

Oddly, despite some close calls and near misses, this system has worked for more than half a century. Nobody decided to launch a first strike, because nobody believed that a first strike would 'work'. They launch 2000 missiles, we launch 1500 missiles, then everybody dies. Nobody wins. In the Cold War, that strategy was called, quite accurately MAD (Mutually Assured Destruction). You don't start a war if everybody dies.

So after all these years, the unthinkable happened. The United States population lost its collective mind and elected an eight year old with impulse control issues to the presidency. And with that, we quite logically rend our garments over this president's access to such god-like power. But if we interject another process in the launch order, we reduce the deterrence value of our strategic forces, and invite a first strike. Which way would be more likely to lead to war? Nobody really knows, but you don't get any do-overs in nuclear war.

Another topic that has gotten a lot of attention recently is the global elimination of nuclear weapons. The Swiss organization ICAN (International Campaign to Abolish Nuclear Weapons) won the Nobel Peace Prize this year, and organizations such as Global Zero, the Ploughshares Fund, the Federation of American Scientists and others have been working on this goal for dozens of years.

So - is it possible? Hey, anything's possible, but the process seems highly unlikely to me. If everybody and every nation - operating in perfectly good faith - eliminated nuclear weapons, then one bad national leader could secretly develop a few crude atomic weapons and dictate terms to the rest of the world.

Unless and until somebody can describe a methodology in which either nobody will cheat or cheating will be somehow nullified, you're not going to see nations give up their strategic deterrent. Personally, I believe that technology, once in the wild, can never be truly rolled back. Machine guns, drones, robots - anything that makes killing and destruction more efficient and effective - will be used simply because they exist. Nuclear weapons are so destructive that they can form their own deterrence regime, but if that deterrence regime is weakened sufficiently, that's what will lead to their use.
...