Diatribes of Jay

This is a blog of essays on public policy. It shuns ideology and applies facts, logic and math to economic, social and political problems. It has a subject-matter index, a list of recent posts, and permalinks at the ends of posts. Comments are moderated and may take time to appear. Note: Profile updated 4/7/12

13 January 2017

Grading Trump’s Presidency: Benchmarks

[For a recent, very popular post opposing Jeff Sessions’ confirmation as AG, click here. For a popular recent post about Russia and our policy toward it, click here.]

Energy/global warming
Health insurance
Good jobs onshore
Economic stability
Geopolitical stability


In seven days, Donald J. Trump will become our president. You’ve got to hand it to him: he’s a master showman. By delaying his press conference for six months, he made it a standing-room-only event and a spectacle much like his campaign rallies. By holding it at the same time as the Senate Judiciary Committee hearings on his most controversial nominees, he made sure it didn’t get too much media scrutiny. By accusing CNN of making “fake news” and refusing to let its hapless reporter even ask a question, he created a frothy controversy and established a punitive and vindictive approach to the media that he may keep throughout his presidency.

Make no mistake about it. If you’re a reporter at an event Trump controls, he will be the boss. If you cross him, he will make your job harder. He will deny you a voice and will deny you access. And he will try to make a fool of you in public to boot.

So you’ll have to do reporting the old-fashioned way. You’ll have to work the phones constantly (and avoid e-mail) to find people who know something and are willing to talk. Then you’ll have to expend shoe leather visiting and interviewing them, mostly in secret. You’ll also have to spend lonely hours pawing through musty documents and secret databases to find out just what’s really going on.

Much of this work will be solitary and boring, but it will be essential. For when Trump’s nominees take over their various departments and agencies, the first casualty will be transparency. The flow of copious, complete and honest statistics that we Yanks have come to expect from our government will fall to a trickle, just like Trump’s nominees’ late and incomplete ethics disclosures. What’s left will sour with slant and “spin.” Our federal government will begin to resemble the old Soviet system of forged and fictional stats, and the GOP will use that very fact as a pretext for cutting more government.

So reporters are going to have to dig deep to find information they now get for free. They are going to have to find ways to verify its accuracy. And these (mostly) liberal-arts majors are going to have to learn to do arithmetic and maybe even some higher math.

Therefore the job of being a reporter is going to get a lot harder. But the rewards for doing it well are going to get a lot sweeter. There will be scoops galore, scandals to probe, facts and real statistics to unearth. And if you don’t think there will be Watergates by the dozen, you can’t yet see the slow-motion train wreck that is Trump’s approach to conflicts of interest and to governing in general.

Reporters willing to ditch the Trump-arranged lie fests and do their jobs the old-fashioned way will have Pulitzer Prizes for the picking. For Donald J. Trump, as president, threatens to become Warren G. Harding on steroids.

Not only is Trump a master showman. He’s also a master lier. Barack Obama was born in the American state of Hawaii, as his birth certificate says. Global warming is a reality (not a Chinese hoax), and an accelerating menace. And no, there were not “millions” of fraudulent votes cast for Hillary in our recent election. But Trump has gotten millions of gullible Americans to believe his three Big Lies to the contrary.

His doing so has two key effects. The obvious one is to strengthen the reality-free, partisan echo chamber of his and GOP supporters. Yet the most important effect is more subtle. By repeatedly propagating outrageous lies, Trump controls the national conversation and takes our collective eyes off the ball. We all report, talk and think about what he just said and whether it’s true. We forget all about whether the problems he promised to solve are getting solved, and whether people’s lives are improving.

If these trends continue, Trump will not just have been the least experienced president in our history. He will be the first president to escape serious, honest evaluation entirely.

If we are to have the ghost of a chance of giving Trump and his presidency that evaluation, we, the people, are going to have to work as hard as reporters to keep our eyes on the ball. We’re going to have to have clear and specific benchmarks for success. And we’re going to have to hold Trump and all his appointees accountable for meeting them. It’s in that spirit that I offer this essay, with benchmarks in rough order of importance.

Energy/global warming

The science of global warming and climate change is complex. It takes five years of hard study, on the average, plus above-average intelligence and drive, just to get a Ph.D. in any serious field of climate science. And a Ph.D. is only the entry ticket to real work. A few more years of postdoctoral research, at salaries that most business people would find laughable, are essential to build a reputation as a good scientist among peers.

There are tens of thousands of climate scientists who have done just this, in dozens of fields as abstruse as tropospheric mixing, paleoclimatology, geochemistry, oceanography, and marine biology. The biggest lie in the whole global-warming controversy is that the average voter can assimilate all their work better than the experts and come to rational conclusions by reading a few anecdotes about weather and some news articles about climate history.

The average voter can’t. Nor can the average pol. They might just as well try to build nuclear reactors in their backyards out of leaves and twigs, or to fly A-380s untutored. They have to rely on the experts; they have no other rational choice. And this conclusion most definitely includes Trump himself.

Climate science is hard. But benchmarking progress in energy and in fighting global warming is far, far easier than understanding the science. Only four numbers are key: the annual global output of carbon dioxide, the average annual global temperature, our globe’s remaining and recoverable fossil-fuel reserves, and the numbers of jobs in various fields of various energy industries.

That’s all. The whole field of energy and global warming revolves around four very real phenomena. First, burning fossil fuels puts carbon dioxide (directly) and other greenhouse gases (indirectly) into our atmosphere, heating our planet beyond our evolutionary comfort zone.

Second, fossil fuels are running out. It took Nature millions of years to make them—hence the adjective “fossil.” Nature ain’t makin’ any more, at least not in the short term. If we humans don’t have good ways to replace the energy from fossil fuels when they run out, our entire advanced civilization will slow or halt. It’s that simple.

Third, energy is the driver of our real economy and our modern, human standard of living. Nations that make the transition to sustainable energy quicker and better will simply have stronger economies. Their citizens will have better lives. Finally, since jobs in extracting, refining and using fossil fuels will disappear when they run out, nations that convert to sustainable energy sooner will have more good jobs to give their people, and those jobs will have better futures.

As was his wont, Ronald Reagan loved to oversimplify things. He asked voters, “Are you better off now than you were four years ago?” Yet with just a bit more nuance, we can identify good benchmarks for energy and global warming.

Have global CO2 emissions gone down, or has their rate of increase at least slowed? Despite natural variability, are there signs that global average annual temperatures are decreasing, or that their rate of increase is slowing? Has the rate of replacing fossil fuels with sustainable energy increased? Have we Yanks created the most jobs in sustainable energy, or has China or Germany? If reporters can focus on these essential questions and stop wasting so much effort, time and energy chasing Trump’s lies and Tweets, we might actually know something useful about his effectiveness as president.

Health insurance

How did Barack Obama manage to get health insurance to twenty million people who’d never had it before? It wasn’t magic, and it wasn’t “socialism.” He used subsidies, and he used “mandates”: forcing healthy people who don’t think they need health insurance to buy it.

Way back in 2007, I wrote essays about how potent a political issue the mandates would be. (See 1, 2). I was right; the mandates nearly killed “Obamacare.” But the GOP didn’t just demagogue the mandates; it’s now demagoguing the subsidies, too. It wants to cut or eliminate them so that it can cut taxes on the rich.

But here’s the rub. There’s no such thing as a free lunch. People who can’t afford health insurance can’t have it unless someone else pays (or helps pay) for it, with subsidies or mandates. And health insurance won’t get much cheaper unless the size of the insurance pool increases dramatically, thereby lowering the per-person risk. That won’t happen without single payer—something the GOP has fought tooth and nail for nearly a century.

The simple fact is that everybody wants a cut. Doctors want more pay (and many deserve it, because they work hard). Private insurers want more profit to pay their executives and stockholders. Lots of people can’t afford health insurance even today, let alone if we hike doctor’s pay and insurers’ profits.

So something has to give. We can’t insure everybody without subsidies, mandates, and/or single payer—national insurance funded by mandatory taxes. And if we don’t insure everybody, emergency rooms will fill up with people seeking grossly inefficient routine care, and people will suffer and die needlessly. And if a pandemic ever comes, low-paid workers who can’t afford to see a doctor will bring it right into your office or home. Ebola for breakfast, anyone?

Those are the choices. There’s no easy fix. And even if there were, Trump’s putting a doctor (Tom Price) who’s only interested in doctors’ pay in charge wouldn’t find it.

So here are the numerical benchmarks for health insurance. Are more or fewer people insured? Are fewer or more people seeking care in emergency rooms? Is the cost of insurance going up or going down? Are insured people, on average, happier with their insurance plans or frustrated with them? And are there more or fewer places where the uninsured can get care, such as free or subsidized urgent-care clinics?

Good jobs onshore

Long before Donald Trump was on anyone’s presidential radar, I wrote and published a “(never-given) matriculation speech” for college students. In it, I outlined the subject of mathematical intuition. If you have it and someone rails about a million-dollar waste in a $16 trillion dollar economy, you will know he’s an idiot or blowing smoke.

That’s pretty much the case with Donald Trump’s jawboning 800 jobs here and 1,000 there in an economy with 145 million employees. Nobel-Prize-winning economist and pundit Paul Krugman recently made this point in his NYT column.

The sixty thousand factories that our Yankee plutocrats sent or built abroad during the last generation aren’t coming back. Nor would we want them to. Do we want to build our national future on lawn furniture, cheap textiles, hand tools and the other stuff that Wal Mart sells? Or do we want to have better jobs making the next generation of products and performing the next generation of services?

Imposing 35% tariffs on goods already being imported from China and Mexico (among other places) would cause economic disaster. It would immediately raise the prices of all those goods by 35%, when there’s nobody at home who makes them. Plus it would cause the same kind of trade wars that help start World War II.

But The Donald has also had a more subtle idea. Suppose we apply a 35% tariff, propsectively, to products made in China or Mexico in new factories set up with American capital or technology, when those products come into our big domestic market. Might such a tariff cause our plutocrats to think twice about sending more of our factories and jobs overseas?

That sort of limited, prospective tariff might work to change our own plutocrats’ short-term and selfish perspectives. But the best strategy of all would be to make sure that our innovators and entrepreneurs have every advantage in creating the next generation of products and services right here at home.

Sure, we can’t always predict what those products or services might be. They might be reusable spacecraft and private space travel. Or not. They might be nanodevices for health care and manufacturing. Or not.

But a few things we can be pretty certain will be in the mix. Fossil fuels are running out, most probably within the lifetimes of children born this year. We will have to have some means of replacing their energy and usefulness. We will have to have solar arrays, windmills, and more hydroelectric plants. If our species ever develops safe nuclear power plants—ones that can’t ever melt down and don’t produce spent fuel that remains radioactive for hundreds of thousands of years—we will have to have plenty of those, too. Ditto for electric cars and efficient high-speed trains.

The benchmarks for the jobs that design, install and maintain these things will be simple. Are most of them going to us Yanks, or are more going to China, Germany and Mexico? Do we Yanks produce and service our own devices for generating sustainable energy, or do we just install and service imports?

Then there’s gene editing, using the CRISPR-Cas9 technology. It makes possible personalized medicine and designer farm animals, if not designer people. When a doctor takes your blood to offer personalized medicine, is it sent abroad for genome analysis, or is the analysis done here? And are the machines and biological reagents used in that analysis made in China, Mexico, or here? These are the benchmarks that will determine whether good jobs stay in this country as our species advances, and whether our nation remains a leader in scientific, technological and industrial innovation.

While addressing jobs, we should not forget Trump’s promise to create good ones rebuilding and improving our national infrastructure. That may still be possible, but the window is starting to close. Interest rates for borrowing are rising. Darth McConnell and Paul Ryan appear disinclined to forego tax cuts for corporations and the rich, making government funding less likely. Already some GOP pols are talking about privatizing infrastructure, which would make the jobs depend on private investment.

Certainly infrastructure jobs ought to count in benchmarking Trump’s presidency. But if he wants to have them to count, and if he wants the government to play a part in their creation, he’d better get moving soon.


Often immigration seems to be an unsolvable problem. There are two reasons why. First, we Yanks are divided between two strong views. One sees immigrants as weakening our nation, taking jobs from American citizens, and threatening us with crime and terrorism. The other sees immigrants as strengthening our nation by bringing us people with drive and tenacity. This second view values the kids who hike a thousand miles to escape Central American gangs, or the Syrians who ride across the Aegean in a dangerous and rickety boat and wait years in squalid refugee camps just to get here. In this view, these immigrants are highly motivated, and motivation is good for business.

There is no reconciling those two views, and each has more than a germ of truth. But the second reason why immigration seems unsolvable is much more important. Jeff Sessions thinks the flow of immigrants, at least from our south, would stop if we never offered “amnesty.” But there’s a much easier, more effective damper to the flow. If we fined every employer who ever hired an undocumented immigrant, and if the fine amounted to more than five years’ savings in wages, there would be no further jobs for undocumented immigrants.

The flow north would dry up in a month or two. The word wouldn’t even have to get to Mexico and Honduras; it would just have to get to all relevant employers here. Immigrants who came to find jobs and found none would go right back. They wouldn’t even have to be deported.

Why don’t we do this? For three reasons. First, it would deprive our plutocrats of an enormous source of cheap labor. Not only is that labor cheap; it’s also docile. People who live in constant fear of being deported don’t rock the boat. They don’t organize. They don’t complain. They don’t go on strike. They just do the work that our plutocrats don’t believe citizens would do—or would do for as low wages—quietly, efficiently and without complaining. If our business owners thought citizens would do the same work, for the same low wages, and without complaining—do you think they would risk the raids, adverse publicity, employee turnover, and general turmoil that comes from hiring “illegals”?

No wall ever built or imagined can keep out people who want to improve the miserable condition of themselves and their families. Of course our media should report on extending our existing 700 miles of southern wall, the extension’s cost, and its effectiveness, if any, in slowing the local flow of immigrants. But doing only that would miss the big picture.

The big picture is how much our government is doing, if anything, to weaken the job magnet that draws undocumented immigrants here despite the hardship of getting here and working here without papers. Is the pre-hiring system for checking workers’ papers more or less effective? How many people annually are turned down for hiring, locally and nationally, because they have no or forged documents? Is the composition of workers who are often undocumented (in construction, slaughterhouses, seasonal farming, etc.) changing, and, if so, how? Are wages for their jobs rising? Are prices for the resulting products and services? Are more American citizens taking these jobs, and are they protesting their working conditions or wages by going on strike? How much disruption does deporting non-criminal workers cause to families, communities, workplaces, specific industries, and schools?

These are the real questions to benchmark immigration policy. The length or height of the Wall is just a bit of showmanship. By itself, it means nothing. In this, as in every other, area, we shouldn’t let the Great Showman’s sleight of hand take our eyes off the ball.

Economic stability

The “good old days” of which Trump reminds his supporters actually lasted only two or three decades. In the aftermath of the most horrible war in history, the factories of Europe and Japan were devastated and inactive. For a few short years, we Yanks not only had the only Bomb. We had the world’s only working, productive economy. Freed from the pressures of war, and with a virtually untouched Homeland, our troops came home, turned our industrial swords into plowshares, and built the factories, goods, homes and prosperity of the immediate postwar era.

That short period was unique in human history. It’s not going to come again. If we try to reproduce it by war, the war might become nuclear and extinguish our species.

So we Yanks have to accept the idea that economic stability will not again derive from being the only combatant left standing. We are going to have to struggle with globalization, international competition, automation and disruptive innovation just like every other nation on our small planet. We are going to have to forge our own “stability” from good planning in the face of Schumpeter’s “perpetual gale of creative destruction.” In short, we are going to have to learn to bring stability out of chaos, again and again.

That is a tall order. So it requires tough benchmarks. We must focus on staying ahead in innovation and innovative production. We must focus on how much of our economy represents new goods and services derived from new technology and new science. We are going to have to recognize that a car like the Tesla (or Chevy’s upcoming Bolt) is worth ten or a hundred models of conventional gasoline cars, while every try by SpaceX to stick a spent-rocket landing on a tiny barge is worth five or six landings of our now-obsolete space shuttle.

In other words, we are going to have to develop new benchmarks for innovation and modernization. Of course those benchmarks must recognize the risk of failure inherent in all new ventures. But they must also internalize the understanding that the next generation of jobs will only come from of those risky new ventures that don’t fail.

Here the benchmarks may have to be more qualitative than quantitative, at least in part. Are we doing all we can to promote and encourage new ventures like Tesla, Musk’s “Gigafactory” for modern batteries, SpaceX and its two competitors, and, yes, stodgy old GM’s possibly market-beating Bolt? With antitrust enforcement and sound policy, are we giving these future-market leaders a fair shake? Are we keeping incumbents from ganging up on them, or using political power derived from economic might to strangle them in their cradles?

Are we making sure, in everything we do in our government and our economy, that the CRISPR-Cas9 technology that we Yanks invented results in factories and test labs on our shores, and not others’, and in good jobs that will create the next generation of innovation? Are we using our intellectual property, our tax laws, our economic incentives, and the strengths of our democracy to make sure that what we Yanks invent enriches us first?

And finally, is the gross economic inequality that is strangling our society and our politics better or worse? No society can run smoothly when the ratio between the pay of the top executives and that of the average worker is over 400 and getting larger by the year. And no economy can remain stable when one source of that inequality is over-financialization and the obscene pay that comes from “inventing” financial instruments that enrich their purveyors but put the entire economic system at risk. Effective financial reform and restraints on speculation (and swindling) are sine qua nons of economic stability.

Geopolitical stability

In no other field than geopolitics does Donald Trump’s advent more accurately encapsulate the Chinese definition of “crisis”: “danger” combined with “opportunity.” Nothing could be more dangerous, for Taiwan and for peace in the South China Sea, than flirting with abandoning the “One China” policy. Nothing could be more unfair than blaming China for decisions that our own government and our own plutocrats made in selling our own factories, our own jobs, and our own technology overseas. Yet nothing would create greater opportunity for human progress than the world’s first and second economy working together to ensure global economic stability and steady, sustainable growth.

Similar conclusions hold with respect to Russia. It’s a good goal to end the Cold War and not wage another. But how and at what price? Must we entirely sell out Eastern Ukraine and Syria, not to mention the EU? Can we exercise some healing influence, if not control, over Russia, if only to curtail the deaths and devastation and work together to maintain a stable, sustainable international order?

These, too, are tall orders. But the proper benchmarks are not hard to see. Has the killing in Syria and the exodus of refugees slowed or stopped? Has the civil war ended or wound down, with a political settlement in sight? Has Russia stopped meddling in Eastern Ukraine and withdrawn its troops from both Ukraine and its border areas? Has China stopped building militarized islands in the South China Sea and threatening the freedom of shipping and resource recovery in the region? Has China accepted, at least in principle, some sort of international rule of law, or some basic principles in this region, besides its old imperial whim? Has China been able to demonstrate some degree of political control over Kim, to avoid the necessity of us Yanks making plans for a possible nuclear first strike?

Trump’s general propensity to make deals, not war, is comforting. But he does have a dark, combative side. He has a macho personality and likes to win. In fact, he often defines himself as a “winner” and others as “losers,” as he sees those terms.

Unfortunately, Putin has similar traits. He likes to win at judo and skiing and appears to define his foreign policy in terms of clear gains and losses for Russia, sometimes without regard to long-term consequences. There is an irreducible risk that, if Trump and Putin come into direct conflict, personality could take over and motivate escalation on autopilot. It would be a shame if excess testosterone led to nuclear war.

For these reasons, we should be thankful that a more even-tempered and circumspect personality, Rex Tillerson, will have the most direct and frequent contact with Putin and Russia as Secretary of State. Maybe one negative benchmark for Trump ought to be the number of snarky comments and Tweets he produces about Putin, and the number of times, if any, he countermands Tillerson’s sensible, moderate or diplomatic acts or statements vis-á-vis Putin.

Most of all, we need to benchmark our global progress against the nightmare scenario that George Orwell predicted in his prophetic novel 1984. Orwell imagined a global society that achieved “stability” through perpetual enmity and low-level wars among three great empires—China, Russia and the United States. He imagined three empires each of which oppresses its own people in a totalitarian state, using the perpetual mutual enmity and war as pretexts. How close will we have come to Orwell’s dark vision, in which the threat of terrorism provides yet another, more current pretext for enmity, war and oppression?

Here our political scientists and diplomats might profitably spend some time devising new, specific benchmarks to measure our nearness to Orwell’s depressing vision. The choice between paradise and perpetual conflict—between Heaven and Hell—is ours to make, with our species’ native free will. We can have paradise on Earth if we cooperate species-wide. Or we can make a Hell out of our planet by letting global warming run away, and by persisting in the type of conflict between nations that Trump’s and Putin’s macho personalities put at risk. We need some nuanced, definite benchmarks for our people and our media to assess, day by day, in which direction we are going.


Donald J. Trump is a great showman. He’s also a successful liar and con artist. He has reached the highest level of American government by attracting everyone’s attention, including the media’s, most or all of the time. In many ways he has made us forget what we care about as we watch his spectacle.

But now the campaign is over. As Trump assumes the office of President of the United States, the time for persuasion will pass away, and the time for governing will begin.

We cannot let Trump set his own benchmarks for success, any more than we could let the bankers who caused the Crash of 2008 do so, or any more than we let VW set its own emission standards.

Our media should take upon themselves, in cooperation and competition with experts, the tasks of setting and maintaining proper benchmarks for the success of President Trump’s presidency. Then our media should apply those benchmarks relentlessly but fairly to what he does (not what he says!). If they do otherwise—if they allow the Great Showman to continue keeping our eyes off the ball—there is no limit to how quickly our nation can decline and how squalid and dangerous our global human condition can become.

Footnote 1: Dubya’s administration did something similar. At one time, the official website for the Four Corners coal-fired power plant—one of our nation’s biggest and most polluting—had announced that the plant burned 28,000 tons of coal per day. It didn’t try to hide the fact that the plant was one of the world’s most horrific burners of humanity’s dirtiest fuel. Yet some time during Dubya’s first term, that useful bit of information disappeared from the website, and the name of the mammoth power plant changed to “Four Corners Steam Electric Station,” making it sound like something cute and innocuous on a single street corner.

Republicans do love their “public relations.” In business, where many of them come from, it helps sell us rubes defective or deceptive products and services. In politics, it’s better known as “propaganda.” They rarely want us, the rubes, to know exactly what they are doing.

The whole premise of GOP authoritarianism is that the rich and powerful do and know best: they are more equal than others. This philosophy runs deep in ordinary Republicans, let alone Donald Trump, who hides his tax returns more carefully than we Yanks once did our secret ways to trigger an A-Bomb. For them, transparency is not a priority.

Footnote 2: A very recent Economist/You.gov poll shows exactly how successful these three lies were. Here is how many Americans believed them as of December 17-20, when the poll was taken:

Americans Who Believe Trump’s Three Big Lies
Lie (and Page of Report)RepublicansDemocratsAll Surveyed
“Birther” lie that Obama
was born in Kenya
is “definitely” or “probably” true (58)
Climate is not changing or change
is not due to human activity (53)
Millions of illegal votes
were “definitely” or “probably”
cast in recent election (61)

Oddly enough, this poll was taken over three months after Trump himself had recanted his “Birther” lie publicly. It shows just how persistent and damaging to public consciousness big lies can be, and how much of an existential threat to democracy they are.



  • At Sat Jan 14, 01:58:00 AM EST, Anonymous Anonymous said…

    "Ebola for breakfast, anyone?" (LOL!)

    "If would be a shame if excess testosterone led to nuclear war. (Typo, first "If")

    Jay, I did search for "robot" and it seems you may have never considered the ways sophisticated robots will soon be affecting many, many aspects of human lives.

    Lifelong computer scientist Ray Kurzweil predicts in 2028, a $4,000 computer with computational power to match a human brain. In 2029 computers/AI/robots will develop a self-identity/self-consciousness and a single $1000 computer will have 1000 times more computation power than a human brain.

    This will have extraordinary effects on what you discuss here, especially your current blog. I predict that within 2-3 years self-driving trucks will replace 2 million US truck drivers.....and after that virtually any labor and then simple service job. McDonald's may just have a manager or two overseeing robots?

    Ray's Kurzweil's recent thoughts: https://www.youtube.com/watch?v=DIIUNtUVDPI

    Interesting thought I just ran across:
    "The part that many AI developers and robotics engineers seem to forget is that by the time 2029 gets here, the AI-powered robots won’t care what we think. They’ll be in charge, and worst of all, they’ll know it."

  • At Sun Feb 05, 12:42:00 AM EST, Blogger Jay Dratler, Jr., Ph.D., J.D. said…

    Dear Anonymous,

    Thanks for catching the typo, which is fixed now.

    As for Kurzweil, he may be a visionary, but his predictions, in my view and others’, are way off.

    The Economist just published one of its famous “special reports” on artificial intelligence: The Economist, January 7-13, 2017, “Finding a Voice.” As the title suggests, the report was mostly about a small part of AI: machines translating, recognizing and responding to human speech.

    But even that small aspect of AI has encountered huge hurdles, each of which took decades to overcome. As I learned from my decade or so of programming as a scientist, computers don’t “think” at all the same way people do. They are much more “linear” and much more rigid than human brains, which are sloppier but infinitely more flexible. Even modern so-called “neural” networks are not really neural networks at all: they are attempts to simulate neural networks with silicon chips and software, which require vastly more space and energy than the equivalent in the human brain.

    Consider the following calculation. Multipolar neurons have more than one dendritic connection. One source suggests that each can have as many as 200,000. Another suggests that multipolar neurons are “the most common” of the three three types of neurons, so that at least one-third of the 100 billion neurons in the average human brain are multipolar, or about 3.3 x 10**10.

    So, assuming that the average multipolar neuron has 100,000 connections, there would be 100,000! ways of exciting that one neuron. This number is greater than 1 x 10**99,990, because each multiplicand in the factorial, except the last nine, is equal to or greater than ten. In comparison, the number of protons,neutrons and electrons in the entire Universe is estimated at 10**80, a relatively minuscule number.

    A 64-bit digital computer with a terabyte Random-access memory (64 x 10**12) would have only roughly that number squared in connections, because only two cells can “connect” at the same time, namely, by loading their respective contents sequentially into an arithmetic unit. That's roughly 4 x 10**27--- absolutely minimal in comparison.

    Another way to reach the same result is qualitatively. As The Economist's article reports, the main problem with computers doing things in the “real world,” including interacting with people through language, is that human thinking and language include an enormous amount of factual data about the “real world,” stored not as bits and bytes, but as analyzed abstractions. For example, it takes maybe ten megabytes to store a single high-resolution image and maybe 100 times that much to make “sense” of it, i.e., analyze it into useful abstractions. Yet your brains stores, in somewhat abstracted form, millions of such images, from your Mom feeding you as a toddler, through the touchdown pass you caught in high school, to your graduation and wedding days.

    So, no, I think Kurzweil is smoking something. Even if I live to be ninety—nineteen years from now—I don't expect to see cars or trucks autonomously negotiating the complex twists and turns of streets and pedestrian traffic in New York City. With active, RF sensors in the asphalt, drivers might be able to go on autopilot on uncrowded inter-city highways for limited distances, but that's all. Trucking firms would still need drivers, if only to avoid loss of the cargo when things go wrong.

    So I think the probability that AI will replace 2 million truck drivers in two or three years is minuscule. Want to bet?




Post a Comment

Links to this post:

Create a Link

<< Home