The Thread To Remind Me We're Living In The Future

Seen something interesting in the news or on the intertubes? Discuss it here.

Moderators: Zamfir, Hawknc, Moderators General, Prelates

KnightExemplar
Posts: 4989
Joined: Sun Dec 26, 2010 1:58 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby KnightExemplar » Mon May 16, 2016 5:01 pm UTC

http://venturebeat.com/2016/05/15/equit ... ng-funded/

Equity Crowdfunding is finally available in America. This means that sites like "Kickstarter" can actually offer shares in the company as a reward. I'd expect companies to be set up more like "Lending Club" or similar decentralized loan platforms however.

Kickstarter probably still has its place. The caps on crowdfunding are still relatively strict (for non-accredited investors). But true equity ought to change things.

Don't get burned people! And these companies are even smaller than "microcaps", with probably less than $1 Million in valuation. I guarantee you that some major company will be equity-crowdfunded at the start, and each of the share-holders will feel like they won the lottery. (Imagine if you invested into Google when it was worth less than $1 mil? When it finally IPOed, you'd be a millionaire easily, and you'd have hundreds-of-millions of dollars if you held onto the stock till today).

But the vast majority of these companies will likely be lemons. Still, connecting investors with companies is a good thing, as long as everyone is smart and knows how to protect themselves. Don't bet the farm on a stranger, but maybe you can bet $100 or so. Its more productive than a lotto-ticket and may define the future of the American Economy.

------------

Beforehand, only millionaires (or people who made $200k / year) were allowed to participate in early round investing. The restrictions have been partially lifted, so that "normal" people can participate in the potentially lucrative world of early investing. But those restrictions existed for the general public's protection! Just because its legal now doesn't mean its safe to do so. All the risks are still there, the politics have just changed such that the government is now willing to let us normal people participate in this activity which was once delegated only to rich folks.
First Strike +1/+1 and Indestructible.

elasto
Posts: 2915
Joined: Mon May 10, 2010 1:53 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby elasto » Mon May 23, 2016 1:00 am UTC

D-Wave claim to have built a thousand qubit computer!

Because quantum computers grow exponentially in power rather than linearly, that's an insane jump from previous attempts...

In a neat, spacious lab in Burnaby, a satellite of Vancouver, I’m looking inside what appears to be a large black fridge about 10 feet high. Within it is an elaborate structure of circuit boards, not unlike the sort of thing a physics class might construct out of Meccano, except with beautifully colourful niobium wafers as the centrepiece. It all looks fairly unremarkable, yet somewhere in here a multiplicity of different universes are thought to exist.

The lab belongs to a small company called D-Wave, a highly skilled collection of just 140 employees that prides itself on building the world’s first functioning quantum computer, which is what is contained within the large fridge-like casing. Actually it is a fridge, the coldest fridge ever assembled. The cooling apparatus enables the niobium computer chip at its core to function at a temperature of just under –273C, or as close to absolute zero as the known universe gets.

The supercooled environment is necessary to maintain coherent quantum activity of superposition and entanglement, the state in which particles begin to interact – again rather mysteriously – co-dependently, and the qubits are linked by quantum mechanics regardless of their position in space. Any intrusion of heat or light would corrupt the process and thus the effectiveness of the computer.

Exactly how and why quantum physics adheres to these science-fiction like rules remains an issue of great speculation, but perhaps the most common theory is that the different quantum states exist in separate universes. The D-Wave quantum computer I look at has one thousand qubits.


There remain doubters:

D-Wave’s first demonstration in 2007 of its 16-qubit device, which involved solving a sudoku puzzle, hardly set the world on fire. Umesh Vazirani, co-author of a paper on quantum complexity theory, dismissed D-Wave’s claims of speedup as a misunderstanding of his work, and suggested that “even if it turns out to be a true quantum computer, and even if it could be scaled to thousands of qubits, [it] would likely not be more powerful than a cellphone”.

Thereafter the company was regularly accused of hype and exaggeration. Part of the problem was that it was very hard to measure with any agreed accuracy what was happening. D-Wave came up with a test to show that entanglement – seen as a necessary prerequisite for a working quantum computer – was taking place.

But some experts doubted the reliability of the test. When D-Wave passed a different test, developed by an independent scientist, sceptics argued that while entanglement might be happening, the only real test was performance.

In 2013, the D-Wave Two was cited in one test as performing 3,600 times faster than a classical computer. But yet again these results were rubbished by several prominent scientists in the field. In 2014 Matthias Troyer, a renowned professor of computational physics, published a report that stated that he found “no evidence of quantum speedup”.

A longtime doubter of D-Wave’s claims is Scott Aaronson, a professor at MIT, who has called himself “Chief D-Wave Sceptic”. After Troyer’s paper, he argued that although quantum effects were probably taking place in D-Wave’s devices, there was no reason to believe they played a causal role or that they were faster than a classical computer.

Brownell is dismissive of these critics, claiming that the “question has been largely settled”. He cites Google’s comparative test last year in which its D-Wave quantum computer solved certain problems 100m times faster than a classical computer.

“If it isn’t quantum computing,” asks Brownell, “then how did we build something that’s a hundred million times faster than an Intel Core? It either has to be quantum computing or some other law of nature that we haven’t discovered yet that’s even more exciting than quantum mechanics. I challenge any scientist in the world to tell us: if it’s not quantum annealing, what is it?”


link

commodorejohn
Posts: 800
Joined: Thu Dec 10, 2009 6:21 pm UTC
Location: Sacramento, CA
Contact:

Re: The Thread To Remind Me We're Living In The Future

Postby commodorejohn » Mon May 23, 2016 4:26 am UTC

elasto wrote:D-Wave claim to have built a thousand qubit computer!

Because quantum computers grow exponentially in power rather than linearly, that's an insane jump from previous attempts...

Image
"'Legacy code' often differs from its suggested alternative by actually working and scaling."
- Bjarne Stroustrup
www.commodorejohn.com - in case you were wondering, which you probably weren't.

User avatar
Zamfir
Posts: 7149
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: The Thread To Remind Me We're Living In The Future

Postby Zamfir » Mon May 23, 2016 10:25 am UTC

D-Wave claim to have built a thousand qubit computer!

Because quantum computers grow exponentially in power rather than linearly, that's an insane jump from previous attempts...

Dwave is, well, odd. It's not snakeoil fraud, but you have to take a very careful look at their achievements. They are not building a typical quantum computer. Their machine cannot run any of those algorithms that quantum computers would be exponentially better at than classical computers. They are not working towards that either.

They appear to have a genuinely unique machine, that might even use quantum entanglement effects in some ways in its workings. It's just not clear if their machine is useful for anything. It's good at solving one obscure problem, much faster than general-purpose algorithms on a PC.

But specialized algorithms can solve that problem even faster than D-Wave, on a regular PC. And they don't have good results on other problems, or even theoretical arguments that they will eventualy have more general results. AFAICT, they might not even need the quantum stuff for their good results. Their machine might just implicitly exploit weaknesses in the problems, just as those specialized classical algorithms do.

They seem very skilled people, making serious progress on a route that's most likely a dead end. Not 100% certain a dead end, but it's not the normal road to quantum computing.

User avatar
Whizbang
The Best Reporter
Posts: 2188
Joined: Fri Apr 06, 2012 7:50 pm UTC
Location: New Hampshire, USA

Re: The Thread To Remind Me We're Living In The Future

Postby Whizbang » Mon May 23, 2016 11:17 am UTC

Zamfir wrote:...not the normal road to quantum computing.


Title of the next best-selling Sci-Fi thriller.

User avatar
Dauric
Posts: 3710
Joined: Wed Aug 05, 2009 6:58 pm UTC
Location: In midair, traversing laterally over a container of sharks. No water, just sharks, with lasers.

Re: The Thread To Remind Me We're Living In The Future

Postby Dauric » Mon May 23, 2016 1:42 pm UTC

Whizbang wrote:
Zamfir wrote:...not the normal road to quantum computing.


Title of the next best-selling Sci-Fi thriller.


"Two probabilities diverged in a quantum matrix and I-
- I took the less probable in this quantum reality,
And that has made all the difference."
We're in the traffic-chopper over the XKCD boards where there's been a thread-derailment. A Liquified Godwin spill has evacuated threads in a fourty-post radius of the accident, Lolcats and TVTropes have broken free of their containers. It is believed that the Point has perished.

elasto
Posts: 2915
Joined: Mon May 10, 2010 1:53 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby elasto » Mon May 23, 2016 4:03 pm UTC

Zamfir wrote:Dwave is, well, odd. It's not snakeoil fraud, but you have to take a very careful look at their achievements. They are not building a typical quantum computer. Their machine cannot run any of those algorithms that quantum computers would be exponentially better at than classical computers. They are not working towards that either.

They appear to have a genuinely unique machine, that might even use quantum entanglement effects in some ways in its workings. It's just not clear if their machine is useful for anything. It's good at solving one obscure problem, much faster than general-purpose algorithms on a PC.

But specialized algorithms can solve that problem even faster than D-Wave, on a regular PC. And they don't have good results on other problems, or even theoretical arguments that they will eventualy have more general results. AFAICT, they might not even need the quantum stuff for their good results. Their machine might just implicitly exploit weaknesses in the problems, just as those specialized classical algorithms do.

They seem very skilled people, making serious progress on a route that's most likely a dead end. Not 100% certain a dead end, but it's not the normal road to quantum computing.

I agree with all that. However, assuming they're not outright lying (and they have no history of that) a thousand qubits is (to forgive the pun) a quantum leap forward in power.

Yes, the proof of the pudding will be in the eating: They will need to demonstrate they can actually solve a useful, real-world problem to get the world to sit up and take notice, but I wouldn't put it past them to do exactly that.

Watch this space I guess...

User avatar
Link
Posts: 1297
Joined: Sat Mar 07, 2009 11:33 am UTC
Location: ᘝᓄᘈᖉᐣ
Contact:

Re: The Thread To Remind Me We're Living In The Future

Postby Link » Mon May 23, 2016 9:39 pm UTC

The word that's really lacking is "universal". There's a plethora of really cool stuff you can do with quantum systems (quantum annealing, Majorana braiding, etcetera), but from what I gather, most people don't have very high hopes for those types. When we get to use 1000 qubits in Shor's algorithm, that's when the real fun begins. Still, the fact that we're at the point where nanoscale quantum effects (as opposed to macroscopic condensate effects like superconductivity) are actually useful for doing computation at all is definitely amazing.

I do wonder where quantum computation will end up going, though. Mainly, if it'll ever find its place outside fundamental research and be something that plays in important role in your average Joe's life. For example, will a universal quantum coprocessor ever be a thing anyone would want in their PC? (Maybe Grover's algorithm turns out to be useful if you want to run a search on a zebibyte disk?)

User avatar
Zamfir
Posts: 7149
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: The Thread To Remind Me We're Living In The Future

Postby Zamfir » Mon May 23, 2016 9:44 pm UTC

However, assuming they're not outright lying (and they have no history of that) a thousand qubits is (to forgive the pun) a quantum leap forward in power.


Well, it would be a great step, if their machine scales better than alternatives. Then bigger is better, and eventually they have to pull ahead.

That's the promise of regular, gate based quantum computers. Those have good theoretical reasons to expect superior scaling, for some useful problems. So if (big if) you can build one to handle big problem sizes, it will almost certainly be relatively fast.

But that's not the case for dwave. There's no such theory for their approach, and not much experimental evidence for it either. Even on the artificial problems that exactly fit their machine. For them, a bigger machine can just handle bigger problem instances, no scale gain expected. If a PC could beat their 512 qubit macbine (and it could), it will beat them just the same on bigger instances.

KnightExemplar
Posts: 4989
Joined: Sun Dec 26, 2010 1:58 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby KnightExemplar » Mon May 23, 2016 9:55 pm UTC

elasto wrote:
Zamfir wrote:Dwave is, well, odd. It's not snakeoil fraud, but you have to take a very careful look at their achievements. They are not building a typical quantum computer. Their machine cannot run any of those algorithms that quantum computers would be exponentially better at than classical computers. They are not working towards that either.

They appear to have a genuinely unique machine, that might even use quantum entanglement effects in some ways in its workings. It's just not clear if their machine is useful for anything. It's good at solving one obscure problem, much faster than general-purpose algorithms on a PC.

But specialized algorithms can solve that problem even faster than D-Wave, on a regular PC. And they don't have good results on other problems, or even theoretical arguments that they will eventualy have more general results. AFAICT, they might not even need the quantum stuff for their good results. Their machine might just implicitly exploit weaknesses in the problems, just as those specialized classical algorithms do.

They seem very skilled people, making serious progress on a route that's most likely a dead end. Not 100% certain a dead end, but it's not the normal road to quantum computing.

I agree with all that. However, assuming they're not outright lying (and they have no history of that) a thousand qubits is (to forgive the pun) a quantum leap forward in power.


My understanding is that DWave has thirty two 32-qbit computers and is running them in parallel (or some other kind of parallelization), which is a very different thing from a singular 1000+ qbit computer.

So they aren't quite lying about having a "1000 qbit" computer (because quantum computing doesn't really have a definition yet). But their computer is definitely slower than a "singular computer that has 1000 qbits entangled". Instead, its a set of parallel computers that each have 32qbits or so... the sum total of qbits being somewhere around 1000qbits. Which is far more practical to implement with current technology. It just won't be factoring RSA anytime soon.

------------

EDIT: http://www.dwavesys.com/press-releases/ ... ve-systems

The results of the research prove the presence of an essential element in an operating quantum computer: entanglement. This is when the quantum states of a collection of particles (or qubits) become linked to one another. The research demonstrates entanglement of a two and eight-qubit subsection of one of D-Wave’s 512 qubit processors, a record number for a solid state quantum processor, throughout the critical stages of a quantum annealing algorithm.


So a few years ago they only had eight-qubits of simultaneous entanglement at a time, but were claiming "512 qubits".

It'd be like saying a modern processor is a 16GB system. Well... what does that mean? 16GB of RAM? Well... modern systems perform 64-bit arithmetic, so maybe its a 64-bit computer. Or maybe you're talking about the 128-bit memory bus? See? People just don't know how to talk about quantum computers yet. So the language is non-standardized and confusing.

It'd be nice if they had a measurement for simultaneous number of entangled qubits, which is a major limiting factor to quantum computing. Eight qubits entangled is only 256x faster than normal. Ten qubits would be 1024x. Twenty Qubits would be a million-times faster if they were all entangled at the same time.

So... how many of the 1000+ qubits can be simultaneously entangled?
First Strike +1/+1 and Indestructible.

User avatar
Zamfir
Posts: 7149
Joined: Wed Aug 27, 2008 2:43 pm UTC
Location: Nederland

Re: The Thread To Remind Me We're Living In The Future

Postby Zamfir » Tue May 24, 2016 7:18 am UTC

I think that in theory, all of their qubits are (indirectly) connected to each other. The main building block is 4+4 qubits with 16 connections between them, and each has connections to the 4 neighbouring blocks.

They don't aim to have many reliably entangled qubits. From what I read, their coherence times are hundreds or thousands times worse than for laboratory qubits, and even those laboratory bits are hard to wire together to reliable gates without all kinds of error correction schemes (which Dwave doesn't have)

To work around that sensitivity for noise, they focus on a search algorithm that gracefully falls back towards its classical version in the presence of noise. The classical version is stochastic, so the noise just takes the place of the random flips that it would have anyway. That's different from those quantum code-breaking algorithms, where unentangled effects give you noisy answers.

Trouble is, the quantum version of the search algorithm doesn't seem to be particularly good, except on highly tailored problems. And even for those tailored problems, people can write classical algorithms that work as good or better than the dwave machine.

User avatar
notzeb
Without Warning
Posts: 625
Joined: Thu Mar 08, 2007 5:44 am UTC
Location: a series of tubes

Re: The Thread To Remind Me We're Living In The Future

Postby notzeb » Tue May 24, 2016 7:38 am UTC

Link wrote:I do wonder where quantum computation will end up going, though. Mainly, if it'll ever find its place outside fundamental research and be something that plays in important role in your average Joe's life. For example, will a universal quantum coprocessor ever be a thing anyone would want in their PC? (Maybe Grover's algorithm turns out to be useful if you want to run a search on a zebibyte disk?)
Grover's algorithm is useful for a lot more than database searching. The most obvious application is that if you have an instance of SAT of size n - which you can solve by brute force in time about 2n - then using Grover's algorithm you can solve it in time about 2n/2. What would take a thousand years at a trillion operations per second to brute force classically would now take less than a second. Since basically every puzzle can be reformulated into a case of SAT, that is nothing to sneeze at.
Zµ«V­jÕ«ZµjÖ­Zµ«VµjÕ­ZµkV­ZÕ«VµjÖ­Zµ«V­jÕ«ZµjÖ­ZÕ«VµjÕ­ZµkV­ZÕ«VµjÖ­Zµ«V­jÕ«ZµjÖ­ZÕ«VµjÕ­ZµkV­ZÕ«ZµjÖ­Zµ«V­jÕ«ZµjÖ­ZÕ«VµjÕ­Z

elasto
Posts: 2915
Joined: Mon May 10, 2010 1:53 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby elasto » Wed May 25, 2016 8:24 am UTC

Adidas, the German maker of sportswear and equipment, has announced it will start marketing its first series of shoes manufactured by robots in Germany from 2017.

More than 20 years after Adidas ceased production activities in Germany and moved them to Asia, chief executive Herbert Hainer unveiled to the press the group’s new prototype “Speedfactory” in Ansbach, southern Germany.

The 4,600-square-metre plant is still being built but Adidas opened it to the press, pledging to automate shoe production – which is currently done mostly by hand in Asia – and enable the shoes to be made more quickly and closer to its sales outlets.

The factory will deliver a first test set of around 500 pairs of shoes from the third quarter of 2016.

Large-scale production will begin in 2017 and Adidas was planning a second “Speed Factory” in the United States in the same year, said Hainer.

Hainer insisted the factories would not immediately replace the work of sub-contractors in Asia. “Our goal is not full automatisation,” said Gerd Manz, head of innovation and technology.

Adidas produced 301m pairs of shoes in 2015 and needs to produce 30m more each year to reach its growth targets by 2020.

Six subcontractors of Adidas in China declined to comment on the new factories or said they were not aware of them.

In the longer term Adidas is planning to build robot-operated factories in Britain or in France, and could even produce the shirts of Germany’s national football team in its home country, said Hainer.

The shoes made in Germany would sell at a similar price to those produced in Asia, he said.

Adidas is facing rising production costs in Asia where it employs around one million workers. Arch-rival Nike is also developing its robot-operated factory.

Not horribly futuristic I know; Automation has been with us a very long time. Just a gentle reminder that the world is going to have to have a plan B once robots and AI have crowded out all the unskilled and semi-skilled work...

And if it doesn't occur during my lifetime, it surely will during the lifetime of my children...

---

Aaand a related story:

Apple and Samsung supplier Foxconn has reportedly replaced 60,000 factory workers with robots.

One factory has "reduced employee strength from 110,000 to 50,000 thanks to the introduction of robots", a government official told the South China Morning Post. Xu Yulian, head of publicity for the Kunshan region, added: "More companies are likely to follow suit."

In a statement to the BBC, Foxconn Technology Group confirmed that it was automating "many of the manufacturing tasks associated with our operations" but denied that it meant long-term job losses:

"We are applying robotics engineering and other innovative manufacturing technologies to replace repetitive tasks previously done by employees, and through training, also enable our employees to focus on higher value-added elements in the manufacturing process, such as research and development, process control and quality control. We will continue to harness automation and manpower in our manufacturing operations, and we expect to maintain our significant workforce in China."

Since September 2014, 505 factories across Dongguan, in the Guangdong province, have invested 4.2bn yuan (£430m) in robots, aiming to replace thousands of workers.

Economists have issued dire warnings about how automation will affect the job market, with one report, from consultants Deloitte in partnership with Oxford University, suggesting that 35% of jobs were at risk over the next 20 years.

Former McDonald's chief executive Ed Rensi recently told the US's Fox Business programme a minimum-wage increase to $15 an hour would make companies consider robot workers: "It's cheaper to buy a $35,000 robotic arm than it is to hire an employee who is inefficient, making $15 an hour bagging French fries," he said.

User avatar
sardia
Posts: 5310
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby sardia » Fri Jun 17, 2016 10:31 pm UTC

https://www.washingtonpost.com/news/ene ... n-decades/
This is the future frontline against climate change.
In an immaculate control room at the Watts Bar nuclear plant, green bars flash on a large screen, signaling something that has not happened in the United States in two decades.
As control rods lift from the water in the core, and neutrons go about the business of splitting uranium atoms, life comes to a new nuclear reactor — the first in the country since its sister reactor here was licensed in 1996.
By summer’s end, authorities expect the new reactor at this complex along the Chickamauga Reservoir, a dammed section of the Tennessee River extending northward from Chattanooga, to steadily generate enough electricity to power 650,000 homes. Although the opening of a new nuclear facility used to draw protesters and angry rhetoric, the Tennessee Valley Authority’s Watts Bar reactor has been mostly welcomed by local residents — and even some advocates concerned about climate change.

While they sound scary, and failures are very long lasting, I do think nuclear power is the future until the technological singularity provides us fusion*.

*If I'm asking for 1 miracle, you might as well ask for another.

User avatar
Zohar
COMMANDER PORN
Posts: 6947
Joined: Fri Apr 27, 2007 8:45 pm UTC
Location: NY

Re: The Thread To Remind Me We're Living In The Future

Postby Zohar » Sat Jun 18, 2016 6:49 am UTC

My gut feeling is this is a good thing and nuclear is safer and more sustainable in the long term compared to other types of power plants. But truthfully, I feel I don't know enough about this.
Mighty Jalapeno: "See, Zohar agrees, and he's nice to people."
SecondTalon: "Still better looking than Jesus."

User avatar
sardia
Posts: 5310
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby sardia » Sat Jun 18, 2016 5:37 pm UTC

Zohar wrote:My gut feeling is this is a good thing and nuclear is safer and more sustainable in the long term compared to other types of power plants. But truthfully, I feel I don't know enough about this.

Based on the standards we set for coal and other fossil fuels, it's very safe. Based on the standards we expect from nuclear power, it's about average safety. We don't have the newest best designs because these plants take forever to build. I think it was designed in the 80s.

User avatar
Coyne
Posts: 879
Joined: Fri Dec 18, 2009 12:07 am UTC
Location: Orlando, Florida
Contact:

Re: The Thread To Remind Me We're Living In The Future

Postby Coyne » Sun Jun 19, 2016 7:30 am UTC

sardia wrote:Based on the standards we set for coal and other fossil fuels, it's very safe. Based on the standards we expect from nuclear power, it's about average safety. We don't have the newest best designs because these plants take forever to build. I think it was designed in the 80s.

We (USA) don't have the newer designs because of bureaucratic inertia. Specifically, the government won't adopt the new designs until proven safe, which can't happen until the designs are implemented.

Of course, there's no proof the current designs are safe...and quite a bit of evidence they aren't. But hey, the good news is they are approved, so use them.
In all fairness...

User avatar
sardia
Posts: 5310
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby sardia » Thu Jun 23, 2016 8:11 pm UTC

https://www.google.com/url?sa=t&rct=j&q ... 1236,d.cWw
Rahwan and his colleagues conducted a series of Mechanical Turk surveys to study how people feel about moral dilemmas involving self-driving vehicles. And overall, they found that people were generally pretty utilitarian in outlook, believing that autonomous vehicles should be programmed such that in a case where they have to sacrifice the driver’s life to save multiple lives (by running into a wall, say, rather than running into a large crowd), the larger number of lives is saved.

But we’re not always such good utilitarians. Indeed, the surveys found “the first hint of a social dilemma” when respondents were then asked how they felt about buying such a car, knowing that it had such programming, as opposed to buying a car whose programming instructs it to always save the driver’s life (even if that would lead to more deaths overall in an accident).

“Even though participants still agreed that utilitarian [autonomous vehicles] were the most moral, they preferred the self-protective model for themselves,” the researchers report.

Meanwhile, yet another survey conducted for the study found that people were particularly uncomfortable with the idea of the government mandating or legislating that autonomous vehicles make utilitarian “choices” in key instances — even though the prior surveys had shown that people generally approve of these utilitarian choices in the abstract.

Strikingly, in one survey question, 59 percent of respondents suggested they were likely to buy an autonomous vehicle if there was no government regulation of its moral “choices,” but just 21 percent were likely to buy the vehicle if there was such regulation.

Would you buy an autonomous car if it was programmed to kill you if deemed necessary?
Turns out if you apply the trolley problem to robot cars, people get uncomfortable real quick. For example, people would rather a car run over other people instead of running over a cliff. They don't say it like that, instead they couch it in terms of maximizing self protection.

To be clear, this would only happen in marginal cases where the car can't keep everyone safe. These are not judge Dredd cars or murder bots. Maximizing self protection or not, robot cars will save millions of lives.

Mutex
Posts: 759
Joined: Wed Jan 09, 2008 10:32 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby Mutex » Thu Jun 23, 2016 8:25 pm UTC

But you also have to consider if you'd be happy someone else driving a car programmed to run you and a bunch of other people over rather than sacrifice its driver. Swings and roundabouts.

User avatar
PeteP
What the peck?
Posts: 1451
Joined: Tue Aug 23, 2011 4:51 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby PeteP » Thu Jun 23, 2016 8:37 pm UTC

I figure I should be able to give people different weights. Me, I get a ten, killing anymore than 10 people to save me seems a bit much. Friends and family also get tens. random people get 1, people I don't get 0 and people I really don't like get -1 so that the car hunts them down. Seems to be the best solution!

Though seriously the arithmetic is more complex than with the trolley problem (though of course that is to judge peoples reactions), because collisions aren't an automatic death sentence you have weigh chance to wound against chance to kill etc. I really don't want to be the one responsible for designing that part of the system.

Tyndmyr
Posts: 10117
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby Tyndmyr » Thu Jun 23, 2016 8:56 pm UTC

Mutex wrote:But you also have to consider if you'd be happy someone else driving a car programmed to run you and a bunch of other people over rather than sacrifice its driver. Swings and roundabouts.


Isn't that the case now?

Mutex
Posts: 759
Joined: Wed Jan 09, 2008 10:32 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby Mutex » Thu Jun 23, 2016 9:06 pm UTC

Tyndmyr wrote:
Mutex wrote:But you also have to consider if you'd be happy someone else driving a car programmed to run you and a bunch of other people over rather than sacrifice its driver. Swings and roundabouts.


Isn't that the case now?


Yes, but I meant as an alternative to cars programmed to sacrifice their driver if it saves lives.

Tyndmyr
Posts: 10117
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby Tyndmyr » Thu Jun 23, 2016 9:19 pm UTC

*shrug*

People will accept the status quo continuing.

Trying to sell folks on "will sacrifice you for the good of others" will probably not make sales, but will increase the difficulty of introducing a system that will save plenty of lives without that.

Mutex
Posts: 759
Joined: Wed Jan 09, 2008 10:32 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby Mutex » Thu Jun 23, 2016 9:28 pm UTC

Without regulation, cars will be programmed to save their driver at all costs. But I could see government regulation possibly being introduced to require them to save as many lives as possible.

Tyndmyr
Posts: 10117
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby Tyndmyr » Thu Jun 23, 2016 9:31 pm UTC

"Government requires your car to kill you" headlines appear everywhere, and folks flip out over it as much as nuclear power.

No doubt costing far more lives in the long run by delaying adoption.

Yeah, seems about right.

morriswalters
Posts: 6427
Joined: Thu Jun 03, 2010 12:21 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby morriswalters » Thu Jun 23, 2016 9:45 pm UTC

I always see this scenario brought up, but I don't know if I have ever seen it to be the case. To begin with in a modern cars the driver is protected more than anybody else, other than someone in another car. And any driver will already try to steer out of it if he thinks he will kill a pedestrian, even if he will hit something harder. It seems to be reflexive. I suspect this is a true for avoiding animals as well. Closing rates are such that in high speed across the median events that the driver currently just dies without being able to do much about it. And most cases of cars crashing in to crowds aren't the type of accidents that happen intentionally because a driver chose to save his life over the crowd. And it would most likely be criminal if they did. Can anybody help me by showing me where this dilemma has occurred in real life?

Tyndmyr
Posts: 10117
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby Tyndmyr » Thu Jun 23, 2016 10:09 pm UTC

It's almost certainly a super niche scenario, yeah.

It *could* arise, I suppose, but it does seem super unlikely. I see it more as a thing people talk about because it's a fun thought experiment than because it's in any way practical.

Tyndmyr
Posts: 10117
Joined: Wed Jul 25, 2012 8:38 pm UTC

Re: The Thread To Remind Me We're Living In The Future

Postby Tyndmyr » Thu Jun 23, 2016 10:11 pm UTC

It's almost certainly a super niche scenario, yeah.

It *could* arise, I suppose, but it does seem super unlikely. I see it more as a thing people talk about because it's a fun thought experiment than because it's in any way practical.

Closest I have is I once ran over an opossum when I had the alternative of steering into parked cars instead. Sucker ran right out in front of me, narrow road. I had enough time to swerve, but I opted to not. Not stupidly fast, no significant risk to myself. More money to car vs life of wild animal. Felt a little bad about it, but...yeah, that's what I picked. Not exactly the same, but maybe vaguely similar, in that it's a decision making thing with similar value weighting.

User avatar
sardia
Posts: 5310
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby sardia » Thu Jun 23, 2016 10:45 pm UTC

Tyndmyr wrote:It's almost certainly a super niche scenario, yeah.

It *could* arise, I suppose, but it does seem super unlikely. I see it more as a thing people talk about because it's a fun thought experiment than because it's in any way practical.

Closest I have is I once ran over an opossum when I had the alternative of steering into parked cars instead. Sucker ran right out in front of me, narrow road. I had enough time to swerve, but I opted to not. Not stupidly fast, no significant risk to myself. More money to car vs life of wild animal. Felt a little bad about it, but...yeah, that's what I picked. Not exactly the same, but maybe vaguely similar, in that it's a decision making thing with similar value weighting.

A robot car would stop for the animal. But assuming the sensors didn't detect it until it was too late, would you appreciate it if the car ran you into property instead?
Any funny outrage headlines would be dependent on how minor the programming was. It could be like the stupid terms of services we sign where nobody cares as long as the product works.

User avatar
EdgarJPublius
Official Propagandi.... Nifty Poster Guy
Posts: 3505
Joined: Tue Oct 09, 2007 4:56 am UTC
Location: where the wind takes me

Re: The Thread To Remind Me We're Living In The Future

Postby EdgarJPublius » Thu Jun 23, 2016 11:13 pm UTC

I'm not sure the discussion even makes sense. I don't think any of the necessary logic even exists in self-driving vehicles and it isn't necessary at all for them to operate effectively.

It's not like self-driving vehicles can tell how many passengers the vehicles around them are carrying, which is pretty fundamental knowledge to have to even begin making decisions about whether to value those lives over its own driver and passengers. They're just characterizing objects in their vicinity by size and velocity.

To the extent the vehicle makes any decisions about what to crash into, the governing principles are probably along the lines of:

1. Try not to hit anything.
2. Minimize relative velocity of unavoidable collisions
3. Especially try not to hit anything that moves and is much bigger or smaller than itself.

In roughly that order of importance. And that's just about all it needs.

Making a car that will sacrifice it's driver to save N other people isn't just about twerking some variables that would already exist in the software. You'd have to add in new sensors it wouldn't need otherwise, process and weight the information generated by those sensors, then add whole nested layers to the collision-avoidance algorithm that are otherwise completely superfluous. That's a bunch of overhead to add to an already pretty complicated system and will likely only result in marginal improvements to a small number of outcomes.
Roosevelt wrote:
I wrote:Does Space Teddy Roosevelt wrestle Space Bears and fight the Space Spanish-American War with his band of Space-volunteers the Space Rough Riders?

Yes.

-still unaware of the origin and meaning of his own user-title

elasto
Posts: 2915
Joined: Mon May 10, 2010 1:53 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby elasto » Sat Jun 25, 2016 9:51 am UTC

EdgarJPublius wrote:1. Try not to hit anything.
2. Minimize relative velocity of unavoidable collisions
3. Especially try not to hit anything that moves and is much bigger or smaller than itself.

It's the third rule where it gets tricky though: Hit stuff much smaller than you at speed and you'll probably live; Hit stuff much bigger than you at speed and you'll probably die.

Would you personally buy a car that had your rule 3 or one that prioritised hitting small things over big ones? (Or, perhaps more telling, which one would you buy for your son or daughter to drive..?)

No extra sensors are needed. This is info that's already available to it.

Mambrino
Posts: 388
Joined: Tue Aug 02, 2011 9:45 pm UTC
Location: No we don't have polar bears. Except in zoos.

Re: The Thread To Remind Me We're Living In The Future

Postby Mambrino » Sat Jun 25, 2016 12:13 pm UTC

elasto wrote:
EdgarJPublius wrote:1. Try not to hit anything.
2. Minimize relative velocity of unavoidable collisions
3. Especially try not to hit anything that moves and is much bigger or smaller than itself.

It's the third rule where it gets tricky though: Hit stuff much smaller than you at speed and you'll probably live; Hit stuff much bigger than you at speed and you'll probably die.

Would you personally buy a car that had your rule 3 or one that prioritised hitting small things over big ones? (Or, perhaps more telling, which one would you buy for your son or daughter to drive..?)


I wonder if choosing will become an option. Sounds like policies like that would be something governments (or insurance companies) would dictate. No choice other than choosing not to buy a self-driving car. Tinkering the official legally mandated AI preference would probably carry heavy penalties. (And I'd bet the government would adopt the "don't hit the smaller thing". Which situation if going to cause more loss of life, roads full of 'selfish' cars preferring to hit others to save their passengers, or cars minimizing the harm to everyone even at the sake of their passengers?)

It will be interesting to see how the legislation evolves. Which side will be found more often the culpable party in accidents between AI car and human-piloted car? And if self-driving cars get in much fewer accidents than human-piloted ones, the vehicle insurance prices will probably reflect that.

User avatar
sardia
Posts: 5310
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby sardia » Sat Jun 25, 2016 12:19 pm UTC

Edgar is taking the easy way out, program it to not hit anything, and leave the details to God and the insurance company.

User avatar
Mauthe Dhoo
Posts: 42
Joined: Fri Oct 22, 2010 4:53 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby Mauthe Dhoo » Sat Jun 25, 2016 2:07 pm UTC

sardia wrote:leave the details to God and the insurance company.


What a strangely repetitive way to phrase that.

morriswalters
Posts: 6427
Joined: Thu Jun 03, 2010 12:21 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby morriswalters » Sat Jun 25, 2016 3:00 pm UTC

sardia wrote:Edgar is taking the easy way out, program it to not hit anything, and leave the details to God and the insurance company.
He seems to be taking the human way out.

User avatar
sardia
Posts: 5310
Joined: Sat Apr 03, 2010 3:39 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby sardia » Mon Jun 27, 2016 7:03 pm UTC

http://www.npr.org/sections/health-shot ... strictions
Machine generated poetry, can you guess if a man or machine did it? I couldn't tell and got 50%right, same as guessing.

User avatar
Mauthe Dhoo
Posts: 42
Joined: Fri Oct 22, 2010 4:53 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby Mauthe Dhoo » Mon Jun 27, 2016 8:07 pm UTC

I can't tell it apart either - to me it all just looks like a news report about Whole Woman’s Health v. Hellerstedt. What a time to be alive!

I have a fond dream that there is a thread somewhere about abortion rights or SCOTUS litigation where everyone is now trying to understand the deep philosophical implications of robot poetry...

User avatar
Felstaff
Occam's Taser
Posts: 5127
Joined: Fri Feb 01, 2008 7:10 pm UTC
Location: ¢ ₪ ¿ ¶ § ∴ ® © ™ ؟ ¡ ‽ æ Þ ° ₰ ₤ ಡಢ

Re: The Thread To Remind Me We're Living In The Future

Postby Felstaff » Mon Jun 27, 2016 11:58 pm UTC

sardia wrote:
http://www.npr.org/sections/health-shot ... strictions
Machine generated poetry, can you guess if a man or machine did it? I couldn't tell and got 50% right, same as guessing.


Fallout From Supreme Court Ruling
Right-wing nut jobs found it gruelling
Wasting hate-filled use of diction
On increasing harsh restriction
Way beyond normal proportion
Just to try to end abortion
Away, you scullion! you rampallion! You fustilarian! I'll tickle your catastrophe.

morriswalters
Posts: 6427
Joined: Thu Jun 03, 2010 12:21 am UTC

Re: The Thread To Remind Me We're Living In The Future

Postby morriswalters » Tue Jun 28, 2016 12:22 am UTC

Here you go I suppose. Machine generated poetry.
Sonnet #2

The dirty rusty wooden dresser drawer.
A couple million people wearing drawers,
Or looking through a lonely oven door,
Flowers covered under marble floors.

And lying sleeping on an open bed.
And I remember having started tripping,
Or any angel hanging overhead,
Without another cup of coffee dripping.

Surrounded by a pretty little sergeant,
Another morning at an early crawl.
And from the other side of my apartment,
An empty room behind the inner wall.

A thousand pictures on the kitchen floor,
Talked about a hundred years or more.

commodorejohn
Posts: 800
Joined: Thu Dec 10, 2009 6:21 pm UTC
Location: Sacramento, CA
Contact:

Re: The Thread To Remind Me We're Living In The Future

Postby commodorejohn » Tue Jun 28, 2016 12:41 am UTC

Okay, someone needs to pull a Pierre Brassau with this thing and see what literary critics make of it.
"'Legacy code' often differs from its suggested alternative by actually working and scaling."
- Bjarne Stroustrup
www.commodorejohn.com - in case you were wondering, which you probably weren't.


Return to “News & Articles”

Who is online

Users browsing this forum: JudeMorrigan, Lazar, Mutex, ViKing and 8 guests