AI learns to write crappy text, and I fail to be alarmed (yet)

One wonders how much of this is media-friendly hype versus actual scary breakthroughs in artificial intelligence:

The creators of a revolutionary AI system that can write news stories and works of fiction – dubbed “deepfakes for text” – have taken the unusual step of not releasing their research publicly, for fear of potential misuse.

OpenAI, an nonprofit research company backed by Elon Musk, Reid Hoffman, Sam Altman, and others, says its new AI model, called GPT2 is so good and the risk of malicious use so high that it is breaking from its normal practice of releasing the full research to the public in order to allow more time to discuss the ramifications of the technological breakthrough.

At its core, GPT2 is a text generator. The AI system is fed text, anything from a few words to a whole page, and asked to write the next few sentences based on its predictions of what should come next. The system is pushing the boundaries of what was thought possible, both in terms of the quality of the output, and the wide variety of potential uses.

Here’s what the program wrote after being fed the first line of 1984:

 “I was in my car on my way to a new job in Seattle. I put the gas in, put the key in, and then I let it run. I just imagined what the day would be like. A hundred years from now. In 2045, I was a teacher in some school in a poor part of rural China. I started with Chinese history and history of science.”

Ugh.

Reminds me of the alarmist and mostly fake stories about how Facebook panicked and “shut down” an experiment in which two bots started to talk to each other in an incomprehensible language. It wasn’t exactly Skynet becoming self-aware, but the headlines tended to make you think that.

Granted, I’m probably not doing the OpenAI research justice here. I’m just very skeptical of AI-related stories that contain words like “fear” and “dangerous” in the headline and lead.

Paralyzed people in Japan controlling robot waiters with their eyes

A brilliant and inventive use of robots gives disabled people a new form of gainful employment in Japan:

On 26 November, a ribbon cutting ceremony was held in the Nippon Foundation Building in Akasaka, Tokyo for a very special kind of cafe.

Called Dawn ver.β, it’s staffed entirely with robot waiters. While these days that’s hardly something new, these aren’t mere robots.

Developed by Ory, a startup that specializes in robotics for disabled people, the OriHime-D is a 120 cm (4-foot) tall robot that can be operated remotely from a paralyzed person’s home. Even if the operator only has control of their eyes, they can command OriHime-D to move, look around, speak with people, and handle objects.

Very cool. All we need now are remote-controlled robots that stand around all day drinking coffee and reminding their employees that they’re gonna need those TPS reports, mmmkay?

The Internet of Incredibly Scary Things

If you are thinking about getting one of those wifi-enabled home security cameras, you might want to think again. At the very least, if you do get one, please create a password for it that you have not used elsewhere.

Otherwise, your fancy “Internet of Things” device might get hijacked by a hacker and start barking nuclear missile warnings at you or talking to your children in a deep, scary voice – which are actual things that happened to families in California. The hackers might also mess with your thermostat while they’re at it.

I take a dim view of the “Internet of Things.” It seems like a total scam to me. I do not see how our lives are materially improved by connecting our devices to the Internet, especially when you factor in the time-wasting complexities of managing these devices and the opportunities for hacking and massive privacy violations that they create.

If you have to get one of these cheap security camera systems that require a stable internet connection to function, for Pete’s sake choose a unique login password:

Nest’s parent company, Google, said in a statement that Nest’s system was not breached. Google said the recent incidents stem from customers “using compromised passwords … exposed through breaches on other websites.”

I recommend using the open-source password manager KeePass – you can download it here. It will store all your passwords in a file that is controlled by a single master password. KeePass will even generate unique random passwords for you every time you create a new entry. Whenever you need to log in to an online account, just copy and paste the password from KeePass.

The ultimate road/rail trip

Trans-Siberian road map

Source: CNN

It would have be the fanciful, proposed railway and superhighway connecting New York to London via the Bering Strait. With high-speed rail, I reckon that you could hurtle from one city to the other in about three days (read to the end for how I calculate that).

From the article:

At a Russian Academy of Science meeting in March, Vladimir Yakunin, the 66-year-old head of Russian Railways, unveiled detailed plans for what may seem like an impossible infrastructure project. Yakunin proposed engineers could build a high-speed railway through the entirety of Siberia, dubbed the Trans-Eurasian Belt Development (TEPR)—the final destination of which would be the mouth of an underwater tunnel crossing the Bering Strait. Highway, too, could be constructed adjacent to the tracks, effectively making ground transportation possible from Anchorage to Moscow—or for that matter, New York to Paris. Or, if we’re going to go there, Miami to Johannesburg.

“This is an inter-state, inter-civilization, project,” The Siberian Times reported Yakunin saying at the meeting. “It should be an alternative to the current [neo-liberal] model, which has caused a systemic crisis,” by which he means an economy based on investing in derivatives and stock buybacks and, in consulting engineer and infrastructure expert Dr. Hal Cooper’s words “things that are easy to do on your computer, but which don’t benefit the real world.” The idea is to instead focus on reviving economic forces that revolve around building something—and in this case a very big, maybe impossibly ambitious something—in the physical world.

Impossible? Of course not. But very, very expensive:

Of course, in order to do this, approximately 12,500 miles of road and new railway would have to be built starting at Russia’s eastern border—which would include the 520 miles between the frigid shores of Nome, Alaska, and Fairbanks, the northernmost point of the Alaskan Highway. And then there’s that 55-mile Bering Strait tunnel itself, which has been priced at somewhere between $25 billion and $50 billion. And what Dr. Hal Cooper calls the “Worldwide Railroad Network” in a 2007 report could range from between one and $1.5 trillion which, Cooper notes, “will be the equivalent of what the United States will spend in total on the Iraq War, for which there will be no measureable benefit to anyone.”

Estimated cost of the US “war on terror” through FY 2019: $5.9 trillion, or about four Worldwide Railroad Networks.

Interestingly, the idea has a long history:

Dreams of bridging the East and West across the Bering Strait have been percolating since the 19th century. Cooper told me that as early as 1846, then-Colorado territorial governor William Gilpin invested in a study to build railway up to northern Alaskan shores. And, it turns out, even decades after being ousted from office, Gilpin was still publishing plans for the “Cosmopolitan Railway” which would fuse together all continents chiefly via the Bering Strait.

The real problem is organizing the resources and manpower to get it done, which requires both political will and a population that is able and motivated to work together on difficult projects. This is becoming more rare. The last man walked on the Moon in 1972, and the Concorde supersonic flights that zipped passengers across the Atlantic in three and a half hours were discontinued in 2003.

The next question, then—which has been on the table now for 150 years—is whether anyone would be willing to invest in a project that could collectively cost trillions of dollars and whose anticipated economic yields would be a generation away.

According to back-of-the-envelope calculations by CNN, the proposed superhighway from New York to London would stretch about 12,910 miles. Assuming high-speed rail could be built along this entire length, with trains running at the average speed of the Shanghai to Beijing bullet train service (181.4 mph), then we’re talking about a rail trip of 71.2 hours, or just under three days.

If the trains could run at the top speed of the Shanghai Maglev Train, or 268 mph, then you could expect to blast from New York to London in just over 48 hours. Now, would that not be the world’s most awesome trip?

UPDATE: China is also mulling a high-speed rail project across the Bering Strait. Of course!

The greatness of film

David Hmmings in Blow-Up

I mean film photography, not movies. (Although don’t get me wrong, movies are also great.) Photographer Graham Carruthers has a blog post concerning the merits of film photography versus its more-efficient digital replacement. That post is a response to this article in PetaPixal about the sudden trendiness of film cameras among the “nostalgic hipster” set.

While I’m hardly a photographer, my limited experience with film has given me an appreciation for the joys of analog shooting. Here’s my comment on the blog:

Speaking as a hobbyist, I always enjoyed using my Nikon film SLR. Film has a tangible existence, which makes it seem more “real” than a digital file, and there is something cool about handling film – loading it into the camera, snapping the back shut and hearing the whir of the take-up spool, dropping off the roll at the local pharmacy and getting back an envelope full of prints and negatives that can be displayed, passed around, stored in an album… It’s definitely less convenient than digital but there is something special about the process that can’t be replicated on a computer screen. By the way, a professional photographer once told me that the best way to learn photography skills is to start with a film camera, because digital cameras make everything too easy.

And, of course, you can make art even with a chintzy disposable camera.

Year of the drone

Drone attack Venezuela

This year there were two major and ominous drone-related incidents. One was the attack on Venezuelan president Nicolas Maduro by a pair of exploding DJI drones in August:

However, in this attack, it is not the drones as such that should be getting the attention. Instead, the interesting part is that it was flown by a non-state group (reportedly the ‘T-Shirt Soldiers’) and that this attack happened in a civilian context, rather than in a conflict zone. It is this aspect that makes the Caracas event a departure from what we have seen before. But it had been a long time coming.

For decades, drones have been used by militaries around the world, with approximately 90 countries using military drones of some kind today. Civilian drones, by contrast, are much younger. But even though the commercial drone market is still in its infancy, millions of hobbyist drones have been sold around the world. Of course, this did not go unnoticed by non-state actors such as terror groups.

Note that the drones used in the Caracas attack retail for about $5,000.

The article reports this amazing fact:

In response, anti-drone technology has boomed: a 2016 Goldman Sachs Investment report estimated that almost 10 percent of US defense research and development funds goes into financing counter-drone systems.

The second was the shutdown of England’s Gatwick Airport for nearly two days in the week before Christmas, grounding 1,000 flights and affecting some 140,000 travelers, due to an alleged drone sighting:

But just how many drones caused this massive disruption, who was operating them and, most importantly, why? Over a week and a half later, there are still no answers, no culprits, and no drones recovered. The best British authorities can offer is that they are “absolutely certain” there was at least one drone.

The attack, if it even was that, succeeded spectacularly in sowing the chaos, confusion and systems disruption that are the hallmarks of fourth-generation warfare:

On Saturday (Dec. 29), in an interview with BBC Radio 4’s Today program, Sussex police chief constable Giles York revealed the latest on what the area’s police force had learned about the incident at Britain’s second-largest airport. He said there had been 115 reports of drone sightings to police, including 93 confirmed as coming from credible sources, such as law enforcement and air traffic employees.

But beyond those eyewitness accounts, things get rather muddy. A couple that had been arrested by police and held for 36 hours was released without charge, and said that they felt “violated” by their experience and the release of their identities. York also had to concede that it was possible that police drones launched to catch the perpetrator(s) during the ordeal caused “some level of confusion”—suggesting that some of the reported sightings could have been of drones operated by police. Additionally, York said that two drones found nearby were ruled out of being involved in the incident, and searches of 26 sites in the immediate area were not fruitful. Further complicating matters, last week a senior Sussex police officer was quoted saying there was a possibility there hadn’t ”been any genuine drone activity in the first place.” This was later called a misstatement, and blamed on poor communication.

Economic and psychological damage: enormous. Cost of attack: virtually nil. The math works!

This is not going to stop. To the contrary, the combination of ultra-low cost and the potential for massive, even cataclysmic disruption means that drone attacks could become increasingly common. One way or another, modern society needs to be comprehensively hardened against malicious drones or there won’t be a modern society anymore.

Prevent

The home of the Magna Carta continues its transition into Airstrip One, as the University of Reading warns students reading a Marxist essay on political violence that the authorities might be watching:

Part of a larger anti-terrorism strategy, Prevent was designed to prevent radicalization and seeks to monitor supposedly vulnerable people for evidence of extremism in the materials they peruse and the ideology they express. The idea is that, once identified, these individuals can be steered by authorities away from negative outcomes. […]

Primarily targeted at potential recruits to Islamist terrorist groups, but also at Northern Ireland-style sectarian violence and extreme right-wing terrorism, Prevent suffered mission-creep pretty much right out of the gate. In 2015, a politics student at the University of East Anglia was interrogated by police after reading assigned material in an ISIS-related publication.

The kid clicked a problematic link, which was thereafter removed from the course materials.

Younger students are being scooped up for alleged radicalization, too. In 2016-17, 272 children under 15 years of age and 328 youngsters between ages 15 and 20 were flagged under the Prevent program “over suspected right-wing terrorist beliefs.” The proportion of individuals referred to government officials “as a result of far-right concerns has risen from a quarter in 2015 to 2016 to over a third in 2016 to 2017,” according to Britain’s Home Office, so that likely represents only a fraction of young people questioned and “mentored” for their suspected ideological deviance.

Under 15 years of age? Guess you have to nip these things in the bud.

Where do these referrals come from? Well, anybody can contact the authorities, but the situation is complicated by the duty the law imposes on both public and private institutions to report people seen as being at risk of radicalization, with very little guidance as to what that means beyond cover-your-ass. The imposition of the duty resulted in a surge in referrals by schools to the authorities.

Informing on your fellow citizens for potential thoughtcrimes is just part and parcel of living in a country full of extremists. Comrade Pavlik would have approved.

“Laws such as this restrict the core democratic right to freedom of expression,” a legal analysis published last year in the Utrecht Journal of International and European Law charges. It “indicates a concerning trend of liberal States embracing opportunities to impose severe restrictions on ‘extreme’ speech.” […]

Parliament is currently considering a Counter Terrorism and Border Security Bill that would go beyond monitoring people for extremist ideology and hauling them in for questioning. The proposed legislation would criminalize voicing support for banned organizations, and even make it illegal to view or otherwise access information “likely to be useful to a person committing or preparing acts of terrorism.”

I would say this defies belief, but sadly, it all fits a familiar pattern. Outlawing speech in defense of an organization is the sort of thing one would normally associate with, say, Cuba or North Korea, but it seems the British have met Big Brother and he is them. Seriously, what is happening in Britain is almost as bad as the garden-variety repression seen in certain dictatorships. Not quite as bad, but moving in that direction fast.

If hauling students in for questioning because they clicked a link to “extremist” material sounds like something out of Orwell, Facebook’s AI monitoring system could have been ripped out of a Philip K Dick story:

A year ago, Facebook started using artificial intelligence to scan people’s accounts for danger signs of imminent self-harm.

Facebook Global Head of Safety Antigone Davis is pleased with the results so far.

“In the very first month when we started it, we had about 100 imminent-response cases,” which resulted in Facebook contacting local emergency responders to check on someone. But that rate quickly increased.

“To just give you a sense of how well the technology is working and rapidly improving … in the last year we’ve had 3,500 reports,” she says. That means AI monitoring is causing Facebook to contact emergency responders an average of about 10 times a day to check on someone — and that doesn’t include Europe, where the system hasn’t been deployed. (That number also doesn’t include wellness checks that originate from people who report suspected suicidal behavior online.) […]

In the U.S., Facebook’s call usually goes to a local 911 center, as illustrated in its promotional video.

I don’t see how the quantity of emergency calls proves that the system is working well. It could just as easily indicate rampant false positives.

More importantly, is this a technology that we really want to work “well”? As the article points out, “There may soon be a temptation to use this kind of AI to analyze social media chatter for signs of imminent crimes — especially retaliatory violence.”

There is a well-known story and movie that explores the concept of pre-crime. Do we really want to go there? And just as AIs patrol Facebook for signs of suicidal tendencies and Community Standards-violating speech, will AIs also be used to augment the growing efforts by governments in Britain and elsewhere to flag, investigate and prosecute people who read the wrong materials and think the wrong thoughts?

Facebook Zuckerberg VR dystopia

Networked Confucianism

Masonic Eye of Providence

John Robb describes China’s now-infamous social credit system as the world’s first “networked tyranny.” Allow me to coin the term “networked Confucianism” to describe the same system.

Combine Confucian ethics with modern surveillance technology and social networking, and this is what you get:

China’s plan to judge each of its 1.3 billion people based on their social behavior is moving a step closer to reality, with Beijing set to adopt a lifelong points program by 2021 that assigns personalized ratings for each resident.

The capital city will pool data from several departments to reward and punish some 22 million citizens based on their actions and reputations by the end of 2020, according to a plan posted on the Beijing municipal government’s website on Monday. Those with better so-called social credit will get “green channel” benefits while those who violate laws will find life more difficult.

The Beijing project will improve blacklist systems so that those deemed untrustworthy will be “unable to move even a single step,” according to the government’s plan. Xinhua reported on the proposal Tuesday, while the report posted on the municipal government’s website is dated July 18.

According to the Party, the overall social credit system will “allow the trustworthy to roam freely under heaven while making it hard for the discredited to take a single step.”

China’s greatest philosopher might have approved:

When Confucius was asked 2,500 years ago what a ruler needed to govern a country, he said 信credit, faith, or sincerity; food 食; and an army 兵. But if he could only have one, it would be the first 信. The Chinese character we translate as “credit” has thus long been a core concept of Chinese governance. […]

At first glance, the official goal of the SCS appears to have little to do with financial credit. It is “construction of sincerity in government affairs, commercial sincerity, social sincerity, and judicial credibility” (State Council 2014), which is more a call to embrace traditional Confucian moral virtues than a vision for high-tech governance. The plan document cites a laundry list of social ills that stem from the lack of trust and trustworthiness at all levels of a fragmented Chinese society. These include tax evasion, factory accidents, food and drug safety scares, fraud, academic dishonesty, and rampant counterfeiting of goods.

How to lose a war without firing a shot

I’m a little rusty on my Sun Tzu and Clausewitz, so I don’t recall what those great military theorists had to say about the bold strategy of allowing your most sensitive defense technology to be sold to your chief geopolitical rival:

China has obtained the big screen software used by Nato and the United States for war room mapping, putting its forces on an equal organisational footing with some of the West’s elite military operations.

Luciad, a defence contractor based in Leuven, Belgium, is selling the Chinese government high performance software used for situational awareness by the military commands of the North Atlantic Treaty Organisation, according to information from Chinese government contractors verified by the South China Morning Post.

The package includes LuciadLightspeed, a program that can process real-time data, including that from fast-moving objects, with speed and accuracy. […]

The same software is used by the United States Special Operations Command at MacDill Air Force Base in Tampa, Florida, where covert missions for the US government – including the raid that assassinated al-Qaeda leader Osama bin Laden in Pakistan in 2011, originated. […]

“Luciad is the Ferrari of GIS software. It comes to the right place at the right time,” said a geospatial information engineer from an aerospace company in Beijing.

Truly, this is a level of cunning strategery that makes Alexander the Great look like Sergeant Klinger! But seriously, what’s the point of having NATO if a Belgian company is blithely selling off crucial military technology to the People’s Liberation Army? Why bother even having a military at all? Wouldn’t it be easier and more profitable to disband NATO, dismantle all the Western armed forces and auction off our technology and weaponry to the highest bidder?

Has the US weighed in on this reported sale?

Revelation 13:16-17

Swedish commuters chip implants

Because cell phones weren’t convenient enough

Swedes have learned to stop worrying and love the chip:

In Sweden, a country rich with technological advancement, thousands have had microchips inserted into their hands.

The chips are designed to speed up users’ daily routines and make their lives more convenient — accessing their homes, offices and gyms is as easy as swiping their hands against digital readers.

They also can be used to store emergency contact details, social media profiles or e-tickets for events and rail journeys within Sweden.

What’s remarkable to me is how marginal the benefit is here. O, the hardship of having to carry around a key card! The agony of having to swipe your phone to get on a train! We are approaching the reductio ad absurdum of modern convenience, where you will be able to go anywhere without having to walk –

Robotic exoskeleton

Get in

– talk to anyone without having to flap your gums, and do anything without having to move a muscle. At that point, the only remaining challenge will be how think without needing to use your brain.

Proponents of the tiny chips say they’re safe and largely protected from hacking, but one scientist is raising privacy concerns around the kind of personal health data that might be stored on the devices.

Around the size of a grain of rice, the chips typically are inserted into the skin just above each user’s thumb, using a syringe similar to that used for giving vaccinations. The procedure costs about $180.

So many Swedes are lining up to get the microchips that the country’s main chipping company says it can’t keep up with the number of requests.

Dystopia now. Seriously, why would anyone want this? It’s physically invasive and has all the privacy and security risks of a Yahoo email account – only it’s under your freakin’ skin!

“Having different cards and tokens verifying your identity to a bunch of different systems just doesn’t make sense,” he says. “Using a chip means that the hyper-connected surroundings that you live in every day can be streamlined.”

You know what? No.