TrustNoRobot

Category: AI

  • Deep Compromise

    Stewart knew something was wrong before he opened his eyes. Too dark.

    His phone was off, no small glowing light in the dark bedroom. He reached to the bedside and booted his phone back up.

    It immediately opened onto a screen with a strange, morphing face speaking to him. The face was constructed of constantly shifting celebrity faces, blending and distorting as if in a dream.

    James Dean spoke in an ethereal, phased voice:

    “Stewart Moss, you have been hacked, DeepComped. Please check your bank account to confirm.”

    Stewart frantically minimized the video and logged into his bank app. He immediately saw that instead of the $20k he kept in his account, he had a balance of -$350k.

    The video face, looking like Kanye West, said:

    “You will do as we say or we will continue to draw against your credit line until you are insolvent.”

    Stewart watched in horror as his account moved to -$475k.

    “Stop!”

    A young Drew Barrymore: “You will do exactly as we say. If you contact authorities we will escalate. You will lose your apartment, your job, and your friends. Every week you will buy and sell certain stocks for us. We will send a message soon regarding your first transaction.”

    “Wait, how do I get my money back?”

    The face transformed into Stewart’s own face, and the voice suddenly morphed into a perfect simulation of his own. Stewart was talking to his perfect double.

    “We will be watching closely Stewart. You will remain in debt until we are assured of your loyalty. You will be rewarded for compliance.”

    “Wait!” Stewart shouted at the phone. “Are you a person or an AI?”

    The video’s face shifted from Stewart visage into the grizzled, bearded face of George Carlin.

    “Does it matter, man?”

    Wonder Stories cover, August 1930
    from here.

  • In 2016, America Failed the Turing Test

    The problem with the Turing test is we are conducting it against our whole civilization, and in this case the testers, normal Americans using social media, were not exactly looking to help contribute to a philosophical discussion about AI sentience.

    In 2016, America failed the modern version of the Turing test. We proved that we could not distinguish bot networks from real human networks.

    The testers, in the case of Americans dealing with networks of bots posing as humans on social media, were angry, scared, distracted humans who are looking to validate their own beliefs.

    These bot networks still exist and lie in wait for the next election cycle to commence. And no law will fully stop them, though I applaud the effort. Another point, these bot networks are controlled by self-serving assholes.

    Let’s switch this up.

    Instead of the Turing Test, what if the test was harder. Not a little bit harder. A lot. And what if it was designed not to detect if a computer could pass as a human, but instead if a human could validate its status as a known entity.

    The Turing Test asks if a human can be fooled into thinking a computer program is a human. It sets up conditions that do not penalize the computer program for not being biological – by putting the human behind text messaging, the test lets the tester focus on just the content of the messages.

    But see, a human isn’t a text message, and in this case, the medium is truly the (text) message. By crippling the human in this test, and letting the bot hide behind text, we create an illusion of equality. But we are not equal with programs, that is a category error that is deeply confusing to many.

    To help eliminate this confusion, and focus on the real issues around identity and bot defense, I propose the Distributed Human Identification Test.

    1) The DHIT test is conducted via videoconference conversations with multiple remote people over long periods of time, in other words, whenever you video chat with someone the test is in effect.

    2) In the DHIT test, you talk to someone you know in a regular conversation. This is a video chat, and can be a business call or a friendly share.

    3) At the end of each conversation, you will be asked if there was anything to indicate that the entity you were taking to was transformed or artificial in any way.

    4) Over time, and many tests, a validity score can be built up about an individual. If they drop beneath a certain threshold that individual would need to take further steps to confirm their human Identity.

    The point of this test is not to determine if a computer can fool a human. We have bots that can do that now, in the wild, outside of the laboratory. Spoiler alert: They can even get bumbling Presidents elected.

    The point of the Distributed Human Identification Test is to confirm human Identity. If you can pass it, you are a real person, known by other real people. We don’t need Equifax for identity confirmation when we have the people you went to high school with.

    By using video in realtime and normal conversation we make the test very difficult for an AI to ever pass. By conducting the test with people you actually know we make this almost impossible to fool. By making it distributed we use the power of large real social networks to build a constantly evaluated Web of Identity.

    The DHIT will protect us from bots, and turn our attention to the real danger: if realtime deepfake videos of us as individuals start being used to commit crimes and spread misinformation, by shitty humans, we will have no defense.

    An AI bot who can get a home loan over videoconference using our identity can ruin your life in twenty minutes, and bring down a civilization in an hour.

    But your Mom knows you, your friends know you, your boss knows you, the network of real people know you.

    Properly integrated into existing video chat systems, the DHIT could be our greatest defense grid against deepfake fraud. And eliminate the greatest security risk, passwords, once and for all. Because boiling things down to text, be it a Twitter post or a password, is the root of the problem.

    Because text doesn’t matter.

    Ideas don’t matter.

    Only people matter.

    Render by Mike Winkelmann

    Painting by Simon Stålenhag

  • California’s SB 1001 – Bot Disclosure Law

    Why should you care that your opinion is constantly being manipulated by hoards of fake humans, bot armies that have infiltrated and devalued our social networks, video sites, commerce, and political discussion?

    You should care because the use of these bots is a Federal crime.

    And more explicitly, A bot pretending to be human is also against the law in California as of July 1, 2019.

    The law, which goes into effect on July 1, 2019, was signed by Governor Jerry Brown on Friday, September 28. It will apply to companies who operate online chat services, where a bot can supply assistance to customers.

    However, the law – SB 1001 – is intended to tackle commercial and political bots, that seek to sway the opinions of people online. This includes bots used on social media platforms like Twitter and Facebook.

    “[B]ots can be – and are often – weaponized to spread fake and misleading news, reshape political debates, and influence advertising audiences,” said the bill’s author, Senator Robert Hertzberg (D-Los Angeles), noted CBS Sacramento.

    “SB 1001 sheds light on the fake accounts that simulate real people and spread waves of false information across social media platforms.”

    Per the law, any bots used online will have to clearly state that they are indeed a bot.

    And from the bill itself:

    This bill would, with certain exceptions, 
make it unlawful for any person to use a bot 
to communicate or interact with another person
in California online with the intent to mislead
the other person about its artificial identity
for the purpose of knowingly deceiving the person
about the content of the communication in order 
to incentivize a purchase or sale of goods 
or services in a commercial transaction or to 
influence a vote in an election.

    Yes. This is exactly the kind of legislation we need on the federal and state level. And of course the movement starts in California. We will see how the law gets implemented in coming months.

    But you should also care because:

    1. Your attention is the largest and most valuable commodity you possess

    2. These false human networks, controlled by less ethical humans, are being used to cynically manipulate you by stealing your attention and directing it to the purposes of those individuals

    Bot networks turn innocent people into slaves of misinformation.

    These are malicious AI networks, using algorithms to find the most effective fake identities for influencing certain population groups.

    The people who use bots who pose as humans have ruined our infospace, and I will no longer entertain the argument that these nefarious techniques of opinion manipulation are legitimate expressions of freedom of speech.

    Painting by Simon Stålenhag

  • From HAL to HAiL

    Arguably one of the most iconic evil robots in cinematic history was HAL 9000, the murderous brain of a doomed spaceship in Kubrick’s 2001: A Space Odyssey. However, I don’t think this trope of an red-eyed AI gaining sentience and killing us is an actual real threat we face from robotics or AI. Worrying about that is akin to worrying nuclear weapons will start robbing banks. But killer AI makes a better movie than people losing their jobs to automation or getting relentlessly spammed by bots via social media.

    But I’ve been thinking a lot about the term I coined in this blog post for the Human-AI Leviathan, or HAL. Obviously I was making an ironic reference to the original HAL here, and for that reason I would like to withdraw my coinage, or alter it at least. I don’t want people to think I am suggesting that the metalanguage that will evolve between AI and humans is going to be conscious and/or evil.

    Plus, most folks won’t get the Leviathan Hobbes reference, and I also think that particular work of philosophy is claptrap. Or people may think I am being literal in suggesting it is a creature or being. It will not be, nor should be talk about it as a being.

    In fact, I think this evolving AI feedback loop will be a positive movement in our species evolution. And while it is not a creature or really a language in the traditional symbolic way of thinking about language, I do think Language suggests more of what I intend, that its raison d’être will be human communication and facilitation, than “Leviathan” does.

    So, here is my re-coinage:

    HAiL, the Human-AI Language.

    To break it down, Human comes before the AI because that is what the Singleton Pattern is all about. There is a lower case “i” for AI to remind us that artificial intelligence should never be considered a being. And because it looks cool.

    And finally, Language. Because even though this will not be a spoken or coding language solely, this will be a medium of exchange between human beings. We will just be using models of self and society to communicate that only machine intelligence will be able to query effectively.

    A concrete example of HAiL is a spam filter. Or a social media news feed. Or a reminder from your calendar app. Or an ad served based on your preferences.

    And wonders yet unrealized, like thought mail.

    This blog is not about instilling fear of AI. This is about setting a new course. Because naming things is important. If we call the thing evolving and growing around us a language, we can understand that. Spoken languages change and grow around us all day long, and besides the folks who are still upset that nobody uses semicolons properly, generally language evolution is seen as a good or natural process. And nobody tries to marry a language. We break through the confusion that AI is a being in its own right.

    HAiL will be a manifestation of technology as an order of life, in the way Kevin Kelly talks about in What Technology Wants, a wonderful book that for some reason I seem to be the only person who made it through. Worth a read. Basically he makes a case that Technology itself has purpose, and is a new type of life. A big idea, maybe a little too big.

    AI should be incorporated into our species as HAiL, which suggests both signaling and dangerous ice falling from the skies. Both threat and greeting. A future not of threatening red eyes, but of raising our hands and being heard.

    And no, I am not making a Hail Hitler reference. Man, the Internet has ruined us.

  • The Coming Microservice of You

    I was reading up on Erisology this morning, which sounds like a nascent idea around a science of disagreement, especially online. I like the thought, but as I have previously mentioned, I don’t have a lot of hope for language as a tool for finding truth.

    Ultimately, we will build AI that learns what our thoughts mean in terms of our actions as Singleton entities, and this AI will spend most of its time in communication with other AI, negotiating reality via bytes and models.

    Out of that meta AI language, that humans may be able to interpret but not speak in, the truth will coalesce like a crystal around our species. We will certainly spill a lot of ink trying to describe the new Human-AI partnership world, that already stirs around us like a great beast, but those words will not ultimately decide the fate of our species.

    It will be determined by the way we decide we want to interface with the Human-AI Leviathan. Which I will call HAL from now on, because “society” just doesn’t cut it when describing what is growing and evolving around us.

    This is why it is important that we enter this new phase as Singleton entities, recognizing our unique status in the universe as creatures with a theory of mind, the ability to model other entities inside of us.

    Because if we do not place ourselves at the center, other humans are going to use us like chits. We should trust no robot because we should trust no human behind the robot who has not modelled us in their own heads and tied our own well being into their own emotional states.

    These AIs will be, and soon. A distribution of mind that models us inside it.

    And what we need to flourish in this world is not just a new way of thinking, though I do think programming metaphors like Singletons and decoupling and data structures are more useful ways to analyze ourselves in the modern world than Freudian models of the mind as plumbing, or religious models based on fantastic personal revelation/hallucination.

    In fact, the model of thought we need ultimately is also one from modern programming. It’s the microservice. A microservice is a program with a simple job, that does one thing completely independent from other services. It is decoupled, and modular, and secure.

    In the Singleton Pattern, your identity itself is governed by a microservice that says who and where you are, and that is all it does. But you can imagine other services that would spring up. Let’s say I want to know how happy Chas is? My AI could query my biochemical state service, that would share my levels of dopamine, serotonin, oxytocin, etc.

    See, this is the kind of tech that extends and transforms us. And we don’t need to map every crack in the pavement on the road ahead of us, we just need to know the direction we are going. And that direction is clear: we must drive together toward honor of each other, and all life.

  • Debate This

    AI debate program narrowly loses to the best debater.

    In a way, Debater follows IBM’s philosophy of constructing a man plus machine future, one in which AI augments and supports rather than succeeds us humans. IBM is next working on commercializing Debater’s core system for business ventures, for example, helping journalists or even governments explain contested issues—for which there isn’t necessarily a right or wrong—and thus making the public better informed. Unlike most AI, Debater can grasp the messiness of the real world, rather than games or toy problems. “We are going out of the comfort zone of AI into territory which is more gray,” said project leader Noam Slonim.

    Taking jobs from lawyers will be the least of it. The truth is, we all deserve AI like this to protect us from the flood of misinformation around us.

    If the few own this power, we are lost. If this is weaponized before we have AI defenses then our information systems will collapse.

    We don’t need this AI to tell us the truth. We need it to protect us from lies.

    Painting by Simon Stålenhag

  • Precision German Ethical Trolling

    German Rules for autonomous vehicles can be read about here.

    These rules are a dangerous fantasy, created by well-intentioned ethicists but impossible to implement. They show clearly how the problem with autonomous vehicles is a non-trivial non-starter in their implementation in the real world.

    Under new ethical guidelines – drawn up by a government-appointed committee comprising experts in ethics, law and technology – the software that controls such cars must be programmed to avoid injury or death of people at all cost.

    That means that when an accident is unavoidable, the software must choose whichever action will hurt people the least, even if that means destroying property or hitting animals in the road, a transport ministry statement showed.

    The software may not decide on its course of action based on the age, sex or physical condition of any people involved.

    First, choosing whatever action will hurt people the least is something even humans cannot do. It’s the easy out of AI Utilitarianiam I have written about before. Glass and steel flying apart at speed is the definition of a chaotic system. Humans cannot predict the outcome of an accident, or have perfect knowledge of what the environment is. Accidents by definition happen faster than such decisions are possibly made consciously, which ticks behind experience at a good second of time. We might pretend that we made decisions after the accident, but in a real one your automatic systems are not weighing people’s gender identity.

    Second, are you telling me that a car, made by fallible engineers, is going to be able to determine the age, gender or physical condition of humans around it going 65mph down the highway?

    To say a car won’t make a decision based on the age of a subject, for instance, means you need to train your algorithms on age to ensure they ignore it. Otherwise, and this is the point, in a neural network you will never know why the decisions are made unless you control the training sets. And you can’t control the world, which is where people and cars exist.

    Let’s say that autonomous cars learn that driving near older people is less or more dangerous, even though they don’t know the people are older. Did they violate this rule? What if they learn that certain races drive better or worse through real world experience. Is this an unethical conclusion? How will we know? Machine Learning algorithms don’t talk and explain their actions or follow rules like this.

    There is no example of an entity besides a Divine one that would have a shot at knowing the world well enough to follow rules like this, and there is no way a Chevy is going to make validated ethical choices.

    It’s why the answer to the trolley problem for humans is to not allow people to stand on train tracks. And the solution to autonomous vehicles is to pass laws that require a human being to be behind the wheel at all times.

    Now, on the other hand if this list of rules is a way for German Ethicists to troll German auto manufacturers with impossible rules so they can continue to drive in the fast lane on the Autobahn: bravo.

    Painting by Simon Stålenhag

  • De-skilling the Future

    A reader writes:

    So can automation cause de-skilling to handle complex situation? If routine problems are dealt with by the algorithm then how do humans develop the skills to handle catastrophic problems? I’m thinking autopilot, self driving cars, call centers. If the answer to complexity in the system is plugging humans in the top end to close the gaps I’m not sure how to get skill set without experience?

    I believe human de-skilling is a real thing, and it happens not only with AI but also with all technology. The development of cars reduced the amount of walking and horse riding people do, for instance.

    But there is also a paradox that emerges: while fewer people walk great distances, more people run marathons for enjoyment than they used to. While spreadsheets have replaced the skills of bookkeeping for many, they have opened up those skills to a far greater number. Tools like TurboTax have reduced the number of skilled accountants, but make those that still exist more valued in certain areas. Hunting skills lost are made up for by more nuturious diets for greater numbers of people, etc.

    To be clear, I am not making an economic argument. Paying jobs that are lost to robots are likely gone for good. What I am suggesting is the adaptability of humans makes us quickly pivot from lost skills to new ones. Adaptation is our best, and maybe only good, feature.

    However, the reader makes an excellent point about dealing with catastrophic problems. If we only have medical robots that care for us and they are hacked or break down, tragedy is sure to follow.

    Which is why I keep emphasizing in this blog that we must never give ethical choices to machines. There must always be a trained doctor working with the surgery robot. There must always be a human behind the wheel of an autonomous vehicle, who will likely choose to drive on their own at times. AI must always be a tool, and never seen as a being, no matter how much people want to marry their waifus or venerate Elon Musk.

    Does this mean I am suggesting that we cripple our technology? That we stop building robots and programs when they cross the threshold of moral action?

    That is exactly what I am suggesting. We should use the technology of Laws to limit AI to being the tools of individuals, with always a single individual to blame for misuse. Because de-skilling alone we can handle. We will adapt, and maybe even steal back the skills when we tire of the machines. It would not surprise me at all if there were more family farms in 100 years. But if we let AI be our bosses in any way, we lose for all time our purpose, which goes far beyond living our lives in fulfilling ways.

    Our purpose is much more grand. We are a biological expression of the universe itself. We are its great transparent eyeball. And there is a dystopian future to prevent, while reclaiming our paradise lost. Work to be done, a world to unite, space to claim for life. A world to save from climate change, families to grow, a solar system to colonize.

    The skills we lose will be nothing compared to what we will gain if we classify machine intelligence correctly, as a tool. We must stop trying to build new life, and instead empower the individuals of our planet equally.

    And we must make the falsification of human being with bots a clear crime. It’s fraud of the highest order. Even if it helped get the person you wanted elected.

    Because of all the manufactured ideas of the modern world: justice, freedom, happiness, etc, the one that means the most for our future is equality. And that starts by defining AI as not equal to us, to prevent the moral confusion that follows.

    Painting by Simon Stålenhag

  • The World To Come

    When the managers brought in robots to take the jobs of the auto workers, we said “well they were just lazy union people anyway” and drove away in our VWs.

    When the managers brought in robots to take the jobs of fast food workers, we said “well they had crappy jobs anyway”, and we shoved three more McNuggets into our mouths.

    When the managers brought in machine learning searches to take away the jobs of lawyers we said “nobody cares about the lawyers,” and our schadenfreude grew.

    When the managers brought in AI to take the jobs of call center folks, we didn’t care because those jobs were all in Asia already.

    When the managers brought in robots to take the jobs of managers we said, “wait, can they do that?” and our slack jaws opened while McNuggets dribbled out.

    And all the wealth flowed into a small number of people who had controlled the managers with money and promises that one day, they could be ultra wealthy too.

    But then a funny thing happened. The myth of working hard to pull yourself out of poverty started to crumble, because there was no work to be had. And the ultra wealthy frantically tried to hide their money offshore because they knew what was coming.

    Because the easiest thing to replace with a robot is an ultra wealthy person. In fact, you don’t even need a robot. You can just replace them with a myth, a story we tell our children about how the world became so riddled with inequality that it made the Middle Ages look egalitarian.

    And then we moved on.

    We don’t need the ultra wealthy. We don’t need managers even. And we certainly don’t need wealthy politicians to tell us how to feel and what to do.

    We have been given the most powerful tool in the universe, artificial intelligence, to serve us as individuals.

    A single person is weak, a liability. A small group of people can change the world. A society can change history. Add machine learning and robots to this equation and you begin to see true power. Soon we will decide to build a world designed for people, and not just ultra wealthy people, using AI as our tool.

    This is going to get very, very interesting.

    Painting by Simon Stålenhag

  • Nick Cave Defends Humanity Against AI

    I won’t say must read, but what he says about why AI will never surpass human art should be enshrined.

    Nick Cave discussing why Smells Like Teen Spirit by Nirvana will always transcend something written by AI:

    What we are actually listening to is human limitation and the audacity to transcend it. Artificial Intelligence, for all its unlimited potential, simply doesn’t have this capacity. How could it? And this is the essence of transcendence. If we have limitless potential then what is there to transcend?

    I once dated a wonderful woman who also owned an art studio and had the privilege of growing up around some the best visual artists of the last two generations. And here is what I learned about Art:

    The more you know about the creators, the more amazing it can become. Not additive, exponential.

    To see her stand in front of a painting and talk about the history behind the art and artist was breathtaking. It did more than enhanced enjoyment: it often made me realize I was looking at the piece completely wrong.

    But…

    Knowing more about an AI creator will likely have the opposite impact. Which is a sign we are not dealing with Singletons, that we are dealing with frauds of beings, fraud committed by product designers and software engineers. These are algorithms without limit. That doesn’t make them Gods, it makes them dangerous tools to be strictly regulated.

    Trust No Robot. Trust those that are selling you robots with human characteristics even less.

    Painting by Simon Stålenhag