The explanation afterwards typifies my skepticism of machine learning and the kind of magical thinking that makes people think that limitless tasks can be automated beyond set domains.

Of course, algorithms with enough data are going to be effective at determining scary, tender or action segments from movies. But just like how they admit, a human touch is required to then piece it together in a way that resonates on an emotional level.

Trailers ultimately are pretty formulaic so they may be automatable but there are bound to be a whole host of areas where either a deterministic result is not practical or the noise of the algorithm response will be high enough to render the prediction meaningless.

Also too bad the movie's getting panned by reviews, I was kind of excited about watching this.

And the reason why a large pizza isn't twice as expensive as a medium is because the primary driver isn't materials, it's labour.

making a large takes slightly longer than a medium, but making two mediums takes almost twice as long as a large.

If you could automate the pizza making process, you would see the price more accurately reflect the size, but who wants automated pizza?

That's what scares me the most about these automated systems..."It worked 1,000 times...just not this time". And I know that there are tons of shitty drivers out there, and when self-driving cars become the majority, maybe accidents will go down...I fear too much complacency.

The inherently chaotic event that exists in the otherwise predictable / trainable environment of driving a car is the unplanned / unmeasured disturbance. In control systems that are adaptive or self learning, the unplanned disturbance is the killer - a short duration, unpredictable event for which the system is unable to respond to within the control limits that have been defined through training, programming and/or adaptation. The response to an unplanned disturbance is often to default to an instruction that is very much human derived (ie. stop, exit gracefully, terminate instruction, wait until conditions return to controllable boundary conditions or freeze in place) which, depending on the disturbance, can be catastrophic. In our world, with humans behind the wheel, let's call the unplanned disturbance the "mistake". A tire blows, a load comes undone, an object falls out of or off of another vehicle (human, dog, watermelon, gas cylinder) etc.

The concern from my perspective (and I work directly with adaptive / learning control systems every day - fundamental models, adaptive neural type predictors, genetic algorithms etc. ) is the response to these short duration / short response time unplanned disturbances. The videos I've seen and the examples that I have reviewed don't deal with these very short timescale events and how to manage the response, which in many cases is an event dependent response. I would guess that the 1st dead person that results from the actions or inaction of self driving vehicles will put a major dent if not halt to the program. Humans may be fallible, but we are remarkably (infinitely?) more adaptive in combined conscious / subconscious responses than any computer is or will be in the near future in both appropriateness of response and the time scale of generating that response.

In the partially controlled environment (ie. there is no such thing as 100%) of a automated warehouse and distribution center, self driving works. In the partially controlled environment where ONLY self driving vehicles are present on the roadways, then again, this technology will likely succeed. The mixed environment with self driving co-mingled with humans (see "fallible" above) is not presently viable, and I don't think will be in the next decade or two, partially due to safety risk and partially due to management of these short timescale unplanned disturbances that can call for vastly different responses depending upon the specific situation at hand. In the flow of traffic we encounter the majority of the time, would agree that this may not be an issue to some (in 44 years of driving, I've been in 2 accidents, so I'll leave the risk assessment to the actuaries). But one death, and we'll see how high the knees jerk. And it will happen.

My 2 cents.
TB

Perhaps it would have been better to say that fuel isn't the only reason. The Airbus A320 in this example has roughly 55% better fuel efficiency than a pre oil crisis Boeing 707, although as Jimbo's big bag'o'trivia points out, that's barely better than the 1950s era prop planes like the Douglas DC-7.

Better automation has also allowed the A320 to reduce the staffing requirements, the 707 required 3 or 4 crew to operate the aircraft, but the A320 only requires 2. The DC-7 also requires 3 crew, but only seats half the passengers (doubling the flight crew costs per passenger).

Greater competition is probably a larger factor. Talking about airline profitability and competition, Warren Buffett joked that had a farsighted capitalist had been present at Kitty Hawk for the Wright Brothers' first flight, he would have done his successors a huge favor by shooting Orville down.

This is controlled by software, which can be tweaked... the whole point of having test pilots fly the thing. These jets are fly-by-wire as they are bricks-in-the-sky compared to a standard plane so they need automated assistance to fly the plane like 'arcade' mode in a video game. And if you've ever beta-tested a flight-sim you'll know they start off shit and eventually get better as they are refined.

We all have our biases.

Conservatives are likely to believe conspiracy nonsense that appeals to their biases (Obama is a muslim/kenyan/communist, etc).

And progressives are just as likely to believe conspiracy nonsense that appeals to their biases (see: anything to do with monsanto).

The trick is acknowledging your biases and fact checking anyway.

I wasn't kidding about automating that. If someone could write a browser plug-in that detected when you were submitting a URL to twitter or facebook and ran it against a snopes-style DB, that would be a boon for humanity.

Stock market trading is almost completely automated now. Just computer algorithms trading with algorithms.

As far as the "will there be a crash this year?" question, my guess is probably not.

@RedSky

First, if it were up to me, you could take over as Minister of Finance in this country tomorrow. Our differences seem miniscule compared to what horrendous policies our last three MoF have pushed. The one prior, ironically, was dubbed the most dangerous man in Europe by The Sun.

We're in agreement on almost everything you mentioned in your last comment, so I'll focus on what I perceive differently.

First, I'd differentiate between fiscal stimulus and fiscal spending, the former being a situational application of the latter. As you said, fiscal stimulus during an economic crisis tends to be inadequate with regards to our macroeconomic objectives. You can neither whip out plans for major investments at a whim nor can you mobilize the neccessary resources quickly enough to make a difference and still be reasonable efficient. Not to mention that it only affects certain parts of the economy (construction, mostly), leaving others completely in the wind. So I'm with you on that one, it's a terribly inefficient and ineffective approach.

Automatic stabilizers work magnificently in this regard, but they barely take any pressure from the lower wage groups, especially if unemployment benefits come with a metric ton of strings attached, as is the case in Germany. A basic income guarantee might work, but that's an entirely different discussion.

The problem I see with merely relying on reasonable automatic stabilizers in the form of payments is that they do put a floor into demand, but do very little to tackle the problem of persistent unemployment due to a lack of jobs. As useful as training and education are, the mere number of highly educated people forced to work mundane jobs tells me that, at best, it doesn't work, and at worst pushes a systemic problem onto the individual, leading to immense pressure. Not to mention the psychological effects of being unemployed when employment is tauted as a defining attribute of a proper person -- aka the demonization of the unemployed.

It's still somewhat decent in Australia, but in Europe... it's quite a horrible experience.

Anyway, my point is that I'd rather see a lot more fiscal spending (permanent!) in the shape of public sector jobs. A lot of work cannot be valued properly by the market; should be done without the expectation of a return of investment (hospitals, anyone?); occurs in sectors of natural monopolies -- all of that should be publicly run. A job guarantee, like your fellow countryman Bill Mitchell advocates quite clearly, might be an approach worth trying out. Economy in the shit? More people on the public payroll, at rather low (but living wage!) wages. Do it at the county/city level and you can create almost any kind of job. If the private sector wants those people instead, they'd have to offer better working conditions. No more blackmail through the fear of unemployment -- you can always take a public job, even if it is at a meagre pay.

I should probably have mentioned that I don't buy into the notion of a stable market. From where I am standing, it's inherently unstable, be it through monopolies/oligopolies, dodging of laws and regulations (Uber), impossibility to price-in externalities (environmental damage most of all) or plain, old cost-cutting leading to a system-wide depression of demand. I'm fine with interfering in the market wherever it fails to deliver on our macroeconomic objectives -- which at this point in time is almost everywhere, basically.

Healthcare is all the rage these days, thanks to the primaries. I'd take the publicly-run NHS over the privately-run abomination in the US any day of the week. And that's after all the cuts and privatizations of the last two decades that did a horrible number on the NHS. Fuck ATOS, while we're at it.

Same for the railroad: the pre-privatization Bundesbahn in Germany was something to be proud of and an immeasurable boost of both the economy and the general standard of living.

In the mid/long run, the effects of automation and climate change-induced migration will put an end to the idea of full employment, but for the time being, there's still plenty of work to be done, plenty of idle resources to be employed, and just nobody to finance it. So why not finance it through the printing press until capacity is reached?

As for the Venezuela comparison: I don't think it fits in this case. Neither does Weimar Germany, which is paraded around quite regularly. Both Venezuela and Weimar Germany had massive supply-side problems. They didn't have the production capacity nor the resources to meet the demand they created by spending money into circulation. If an economy runs at or above its capacity, any additional spending, wherever it comes from, will cause inflation. But both Europe and the US are operating faaar below capacity in any measurable metric. You mentioned LRAS yourself. I think most estimates of it, as well as most estimates of NAIRU, are off quite significantly so as to not take the pressure off the wage slaves in the lowest income sector. You need mass unemployment to keep them in line.

As you said, the participation rate is woefully low, so there's ample space. And I'd rather overshoot and cause a short spike in inflation than remain below potential and leave millions to unneccessary misery.

Given the high level of private debt, there will be no increase in spending on that front. Corporations don't feel the need to invest, since demand is down and their own vaults are filled to the brim with cash. So if the private sector intends to net save, you either have to run a current account surplus (aka leech demand from other countries) or a fiscal deficit. Doesn't work any other way, sectoral balances always sum up to zero, by definition. If we want to reduce the dangerous levels of private debt, the government needs to run a deficit. If we don't want to further increase the federal debt, the central bank has to hand the cash over directly, without the issuance of debt through the treasury.

As for the independant central bank: you can only be independant from either the government or the private sector, not both. Actually, you can't even be truly independant from either, given that people are still involved, and people have ideologies and financial ties.

Still, if an "independant" central bank is what you prefer, Adair Turner's new book "Between Debt and the Devil" might be worth a read. He's a proponent of 100% reserve banking, and argues for the occasional use of the printing press -- though controlled by an inflation-targeting central bank. According to him, QE is pointless and in order to bring nominal demand up to the level we want, we should have a fiscal stimulus financed by central bank money. The central bank controls the amount, the government decides on what to spend it on.

Not how I would do it, but given his expertise as head of the Financial Services Authority, it's quite refreshing to hear these things from someone like him.

In the short term, absolutely. But on a longer term, automation will mean there are simply NO jobs. It'll happen in my lifetime, and the only way it doesn't is if we run out of energy to power said automation before that.

When it happens, we will need a serious rethink about our entire economic system.

Huh, never thought of that. So true...and I will add to this. In the past jobs in service have been able to absorb people into it to take up the loss of jobs (Say in agriculture and such.) So when Henry Ford made his assembly line, jobs were created pretty much everywhere. Restaurants are but one great example of this.

But with the techno revolution, the service sector was already pretty full. Now it is saturated. If I see one more new gas station down here in Florida, or another restaurant open, it will be too soon. I remember TWO, TWO Starbucks in the same mall. Such a false economy...

Now add automation and boom...

Automation is not the only problem, true. But here is the funniest part. Everything in a free market is commodified. That is for sale. This includes life saving services even. I.e., a free market. And yet people somehow complain when corporations invest in politics. They claim that investments should not be made in that sector, which I laugh at...

But anyways, to suggest that corporations will lower prices (I.e., profits) just because transportation costs go down is ignoring traditional responses. Lets say grocery stores--which have enormous competition and have been failing let and right--jacked up prices because of rising fuel costs...then fuel went to 1.70 a gallon. Whew, those (lack) of savings are rolling on in... When the labor becomes 0 dollars, we shall see exactly that back...

Now, with that said, hell yeah automate driving. I never implied we shouldn't. Just the 39,000 lives saved, insurance costs, etc are worth it. But I am pointing out a cumalative stacking of bad effects coming up (which have already been slowly hitting.)

I guess the question is then are we going to be like the grasshopper or the ant? Will we prepare for the eventuality that automation and political corruptness (based on the demands of cheap employment pools and the money they receive from corporations desperate to keep that status quo) will merge together for the perfect storm? My problem is the attrition has been slow, just compounding the problem...

I would argue that automation still isn't the job killer #1. Plain old political decisions, such as sound finance, deficit hawkery, and austerity lead by a mile in this category. Neither is being addressed properly, but I find it hard to focus on the employment effects of automation when the Eurozone, for instance, runs at >10% unemployment strictly due to policies enacted by (non-)elected officials. We don't need technology to cause mass unemployment, humans can do that all on their own.

Additionally, even the amount of work available is a matter of perspective. Within the current system, the number of jobs with a decent salary is already dwarfed by the number of people looking for one. The amount of work to be done, on the other hand, is not.

Case in point: our (read: German) national railroad company is short-staffed by about 80.000-100.000 people, last I checked; our healthcare system is short-staffed by at least 200.000 people, probably a lot more; law enforcement is short by about 50.000; education is short by at least 20.000. Let's not even talk about infrastructure or ecological maintenance/regeneration. These are not open positions though, because nobody is willing/able to pay the bill.

So while I agree that we should be discussing how to deal with technological change, a more pressing matter is either to alter the system or to at least take back control over the vast sums of dead currency floating around in the financial nirvana or on Stephen Schwarzman's bank accounts. First stop: full employment. Then, gradually, guaranteed basic income when automation does, in fact, cause mass unemployment.

Finally, I don't think automation will do as quick as sweep as some presume. The quality of software in commercial machines is quite absymal in many cases, since it was written in the normal fashion: do it now, do it quickly, here's five bucks. Efficiency improvements generally come at the price of QA, and it shows. Europe's most modern railway control center is nearby, and it never went online -- Bombardier cut corners and never had the proper railway expertise to begin with. Meanwhile, the center build in '53 is working just fine, and so are the switches put in place when Wilhelm II was running the show.

Edit: That said, I'm thrilled to see mind-numbing labour being replaced by machines. Can't happen quickly enough.

It's different this time though. Every technological advance moves jobs from humans to automations once the automation is good/cheap enough.

Right now, automations aren't good/cheap enough to do most of the jobs humans do (if they were, they'd already be doing it).

But that's going to change. Even for "creative" jobs (music, writing, art, etc), computers are getting better at it. Remember, they don't have to be perfect or even as good as the best humans, just better and cheaper than most.

Eventually the number of jobs that actually require human input will be vanishingly small.

This is going to happen.

http://videosift.com/video/Humans-Need-Not-Apply