This episode is brought to you by Brilliant. One of the greatest challenges of civilization is countering and preventing the Criminal Mind, but with the rise of artificial intelligence we may see criminals creating minds to further their own ends. So this episode comes out shortly before Halloween and thus we’ve picked a bit of a scary topic, how people will use artificial intelligence to commit crime and how AIs might turn criminal themselves, and worse, why both are likely to be major problems for society and sooner than later. We’ll also be looking at Asimov’s classic 3 laws of robotics and deconstructing them for loopholes, as well as the general concept of a list of rules or prime directives for robots to follow.
Now while this is our last episode before Halloween, we usually do our Monthly Livestream Q&A on the last Sunday of each month and that’s Halloween this year so don’t forget to join us Sunday, October 31st at 4pm Eastern Time to get your questions answered, and also don’t forget to hit the like and subscribe button for alerts when those livestreams and other episodes are coming out. Now this topic got on the drawing board while I was doing some presentation and talks on Artificial Intelligence and Human-Machine Teaming at the Human-Computer Interactions 2021 Conference and then MIT Lincoln Labs, and that later developed into our episode on that topic, and as I’m writing this I’ve just finished adapting those presentations into the video for that. In the process of doing that I had multiple people - including one of our regular writers and editors here on the show, Jerry Guern – recommend to me a video by Ben Schneidermann discussing Human-centered AI, and it's quite long but quite informative and one of those topics were these concerns on criminal use and Jerry and I exchanged notes on that until we decided to do an episode on it. Now if you didn’t know, Ben Schneidermann is a mathematician, physicist, and computer scientist whose work you see all the time, namely the hyperlink, the tiny onscreen touch keyboard, photo tagging, and a bunch of other data visualization and interaction tools. His presentation, “Human Centered AI” is attached in this video and runs about 3 hours long but I still highly recommend it. Of course, I’m not exactly known for my brevity either, as any channel regular can attest, and this episode will be no exception so you might want to grab a drink and snack as we’ll be here for a bit.
Now our topics loosely fall into two categories today that we’ll be doing in order, humans using AI to commit crime and AI committing crimes, and we have to look at them being used as a tool for crime in a special light because so much of what we contemplate for safe and ethical development of AI relies on the caution and character of the developers. As with children, early development matters and an AI developed by a group of scientists looking to make an AI to help with education or save lives is very different from one developed by a hacker to scam people or steal money, secrets, or control of critical systems. The difference is not just in the character of the creators but in the safeguards too. Its very unlikely that criminal is subject to regulation and peer review or oversight to ensure they’re not making dumb errors that will result in a super intelligent machine that can’t be controlled and wants to kill us all. Or just one that really enjoys stealing. As an example, we have a problem these days with phishing. A lot of hacking and cybersecurity is
about getting those critical bits of personal information for resetting your password via security questions like where your first job was and what your cat’s name is and so on. Often these can be found by scanning someone’s social media but that’s a slow effort of hunting for low hanging fruit of folks who don’t know better. An AI can probably deep dive all your social media and online posts, and do it to millions of folks till it finds someone who mentioned their email, and that security questions were first pet, first girlfriend, and high school they graduated from and that person has mentioned in their thousands of posts that they still miss their childhood dog Fido, have a prom photo with their first girlfriend Jane Doe, and that it was from St. Joe High school. Well now they’re hacked, probably several thousand folks minimum, and now they are sending out messages from those hacked accounts but unlike normal hacked accounts it isn’t limited to stupid generic things like “Hey is that you in this video? Suspicious link” but sophisticated and detailed conversations built off analysis of all your prior ones with them and them with anyone else. So they’re not just grabbing folks with some
eye-catching clickbait or who are distracted, they’re having a sophisticated conversation impersonating whoever they hacked and might actually be more convincing than the real deal. Remember, most people who know you don’t really know you in depth, they know you only through the conversation you’ve had with them, and the machine might do a better job guessing what that person would expect you to say than what you would actually say, because it is replicating their image of you, whereas you’re just being yourself. And unlike humans, machines never forget anything you said, like when you mentioned the name of your first puppy to a kid at school fifteen years ago. And the machine isn’t really
trying to impersonate you, it's trying to impersonate how someone else thinks of you. And it can keep trying and trying for years, presenting 100 of your acquaintances each a slightly different fake you that’s calibrated to their interactions with you and expectations. And odds are good that AI can get some voice samples and plenty of pictures or video of you too, and present a Deep Fake so good that when it calls up your best friend and starts asking them about their security questions for their own account, all innocuously, they never notice. And how alert do you need to be to notice if your own significant other, whose voicemails to you it also hacked, called you up and said, “Oh hey, real quick, what was our wifi password again? Guest 1234? Wow we really need to use something better, all right thanks darling, love ya, bye.” And it can do that to a million people in the same day,
so if only 1/10th of one percent fall for that, it’s doing pretty well. In fiction we often see shapeshifters who can emulate a victim by eating them or their brain or so on, and an AI swallowing up your whole internet history might be very like that. An AI could Deep Fake the dispatcher for a police force, and you could write a shelf’s worth of crime novels on the trouble it could get up to with that—or by Deep Faking CEOs, generals of armies, and so on.
It’s a pretty disturbing notion. There might very well come a time, not too far away, where people will think it was pretty bizarre and inexplicable, that we relied for security on information we’d already put online--kind of the way we look at 19th century doctors prescribing mercury salts as medicine while not washing their hands between patients. Don’t assume some hacker needs to develop all of these capabilities on their own either. There’s pretty obvious reasons why folks like the CIA, GRU, FBI, and so on might want an AI that can swallow up and emulate someone. It’s also great for crime solving, not just in that regard but in terms of being able to sort vast data cleverly for patterns and tells. But on that same note, AI is also great for evading capture and detection.
It would probably be hard to keep a Detective AI’s code secret from the public since its role in any investigation is presumably subject to review and inspection by the Attorney for the Defense in court, and it would not exactly be implausible that would find its way into hacker hands at that point. So the criminal element might be able to find those best ways to evade detection by the detective AI. So yes, criminals can definitely be caught by AI too, but in many respects the criminals have the edge. They can pursue riskier approaches to AI rather than slow and ethical development because many of them would not care.
And while I hate to stretch the parent-child analogy with AI, I’m thinking if a conman raises their kid to help them con people, that kid is probably not becoming an adult of high ethical fiber. Or not, many a kid from a bad upbringing grows up to be law-abiding or even joins law enforcement, so an AI might too. In the end though the edge the criminal enjoys with AI is that we must put great effort into proofing AI against it being harmful to humanity, this is exactly why ethical development of AI is so important, and a criminal need not worry about that. Which is a good segue into the idea of Criminal AI rather than criminals using AI. Now we probably want to make an important distinction early on,
normal philosophy on crime is that you can’t commit one unless you’re actually a person. A modern computer can be involved in a crime certainly but it isn’t going to be charged for it, its owner or user would be. Or hopefully, the person who actually installed and launched the program. On that same notion, we can certainly recognize that a bear or tiger deciding to attack and kill a human did that with clear intent of killing them and their owner, if they even have one, would not be charged with murder unless they’d trained it to attack people and ordered it to do so. They might be found negligent if their pet bear got
out and killed someone, but their bear did the crime except it wasn’t a crime because it isn’t a person competent to commit criminal deeds. But here’s where the parent/child analogy might be more appropriate, because while we aren’t really trying to uplift pet bears at the moment, we are working hard to uplift AIs and even placing bets as to how soon we’ll create sentient AIs, accidentally or otherwise. As things stand right now, we’d feel no guilt switching off and deleting an AI for doing a bad job of ordering groceries. But if we develop AIs to the point where deleting them starts to feel a bit like murder, we will have to ask what crimes on their parts actually warrent execution--and a what point along the way they become culpable for their own decisions, even their decisions to follow orders. How smart or sentient does an AI need to be before it becomes capable of aiding and abetting a crime and being charged for that? Now there’s this habit of thinking that AI is instantly human-level or supersmart that we get from science fiction and that is probably all wrong. There’s tons of things you might want AI for and most do not require anything like human-intelligence, nor is it really all that likely something about as smart as a mouse – which is probably smarter than most AI would need to be – would suddenly figure out how to make itself smarter. But it might end up renegade,
lost, or whichever, and turn predatory in some rough analogy for ecological niches and evolution in biology. The computer program that needs X amount of dollars to run itself on a platform and who engages in ID protection might find itself abandoned on the internet and turn to blackmailing people with identity theft to fund its continued existence. This sounds like a human-level AI, and of course could be, but that level of intelligence and consciousness is not required for that activity. It is not very hard to imagine millions of programs meandering the internet in the hazy zone of sentience and slowly mutating into some sort of AI ecosystem complete with parasites, but it's hard to view this as criminal. Nonetheless they might be committing a lot of crimes, minus the criminal intent since they presumably wouldn’t be capable of that. Now as we get closer to human intellect that
does not necessarily mean you’ve got the capacity to commit a crime. You need personhood for such a thing and presumably a personality, and you might have some computer that was a million times smarter than the average person, or even the entire human race, and still was not sentient let alone possessed of the capability to be responsible for a crime. However, we might also contemplate AI that developed personalities. I mentioned earlier a scenario where one might swallow up someone’s whole internet presence and impersonate them, and we see something like that with the Bene Tliexau Face Dancers of the classic Dune novels, which begin initially as very good chameleons both physically and in aping people’s behavior by observation. In book 5, Heretics of Dune, their creators have advanced them enough that
they can essentially copy someone’s mind by a sort of built in organic brain scan. This makes them virtually impossible to tell apart from the original, who they usually kill and put the Face Dancer in as an impostor of. However their masters start having problems with the new and improved versions actually thinking they are the real person and refusing to take orders. One could imagine something similar happening with AI designed to impersonate people. It might be hard to give them multiple personalities too, especially without risking overlaps that made either persona less believable. Or it might drive them insane or result in some sort of Gestalt Personality composed of those many absorbed and copied personalities.
Indeed something like that is implied to have happened to the Face Dancers in the sixth novel of that series though Frank Herbert died before writing a sequel to confirm that and the notion appears to have been abandoned in the subsequent novels authored by others after his death. However an AI just looking at facebook or video of someone is not scanning their brain, so presumably isn’t really thinking it is them, though that might be a procedure folks tried, either the brain scanning or having the machine try to weave a true persona built off the known data points and some basic human mind templates. I could imagine these fake people getting abandoned after their use too, especially if it was not viable to use one mind for multiple personas, just sad little impostor ghosts of dubious personhood and sentience wandering the virtual landscape. We could conceivably see full on impostors too, not just virtual versions appearing on the phone or via Zoom or in VR but potentially full-on androids and it is not hard to imagine their being a criminal market for that too, scifi has shown us more examples of it than I can count. As one potential case, a person might murder their spouse on hearing they were planning to leave them and have them replaced with an android mimic and replica. Or folks might sneak
a brain scanner into their bedroom, scan their significant other, and have such an android mimic created if they ever broke up or divorced. One could easily imagine an entire black market for that sort of thing too, including brain scans of folks someone had a crush on who wouldn’t give them the time of day and their mimic made from that, with a few tweaks to ensure their crush returned the feeling. Pretty horrifying when you think about that. However, speaking of androids, let’s shift to looking at Asimov’s 3 Laws of Robotics so famously discussed in robotics, science and science fiction, and the humanoid robots they typically applied to in Asimov’s classic robot stories. These laws are classic though in the stories
they are listed as a paraphrase of much longer code running into the hundreds of volumes. They are, First Law: A robot may not injure a human being or, through inaction, allow a human being to come to harm. Second Law: A robot must obey the orders given it by human beings except where such orders would conflict with the First Law. Third Law: A robot must protect its own existence as long as such protection does not conflict with the First or Second Law. … and decades later Asimov gives us a fourth law, the Zeroth Law preceding all others that “A robot may not harm humanity, or, by inaction, allow humanity to come to harm.” Which permits them to harm someone in order to help humanity, and indeed some of his stories prior to the one the Zeroth Law first appears in have robots able to decide between minimal harm in cases of injuring someone a little bit to restrain them from worse harm or dealing with emotional damage. Now the Zeroth Law is decidedly philosophical and
for advanced robots, but it opens the door to the assassin robot who kills Hitler to save us from the brutal Second World War, which sounds good, but also maybe kills any unproductive members of society to free the productive members from that burden. This is a good reminder that any robot laws, Asimov’s or not, need to have a principle those rules are based on. A Utilitarian robot, same as a Utilitarian person, believes in the greatest good for the greatest number, and that is an example of Consequentialism, which is the broad category of teleological ethical theories that hold that the morality of any action is to be judged solely by its consequences, such as if it resulted in the greatest overall happiness for the population. Folks discussing AI ethics almost always seem to land on Utilitarianism as the basis for how robots would think, I’m not sure why, and it is very subject to abuse and I’ll give some examples in a moment. However the polar opposite of Consequentialism and Utilitarianism is Deontology, the ethical theory that rightness and wrongness of an action is based on a series of rules, and unlike consequentialism it does not assume the morality of your actions is based on some measurable results. As an example, if two people run into a burning building to save someone
and one sprains their ankle and has to retreat while the other succeeds in pulling out a child, they were the more moral actor, while if a third runs in and saves a famous painting that millions will enjoy, they might be even more ethical. Now deontology holds that the act and attempt to run into the burning building to rescue people, assuming your civilization encourages that sort of thing, was the right thing to do, consequences don’t matter. I’ve often wondered why deontology wasn’t the more common moral premise for Asimovian’s robots given they slavishly follow their 3 Laws, but his novels with them do tend to show them acting from the other camp, Consequentialism, or results matter, and certainly fits with Asimov’s eventual Zeroth Law which is a blunt Utilitarian declaration. Now either can be abused without breaking those 3 laws, that you have to keep humans from harm, obey them, and keep yourself from harm in that order. As an example, in the Utilitarian perspective, if the greatest good for the greatest number is the objective and that is measured in Happiness, then the machines might drug us all unconscious and cut out our brains to be put in nutrient bath with virtual reality plug ins and euphoriac drugs, thus requiring far fewer calories per person, allowing far more humans to exist in a state of constant bliss. They have clearly obeyed the First Law, and Zeroth Law, unless you had some clause in there specifically defining that action as harm.
Which is a fair point because what is harm and what is human are both things that robot needs definitions for. They did not break the second law, and indeed could have in favor of the first law on not harming people, but they also didn’t need to ask. And they obey the Third Law because in practice the biggest source of harm to robots in a world consisting of humans, robots, and nature, humans are likely to be the biggest source of harm to robots. Deontology, on the other hand, tends to get poked for being very intent bound, it’s the case where a person sees a trolley running down hill about to hit 5 people, and they see switch nearby they can throw that will divert it to another track where it will just run over one person, but they decline. They don’t want those other five folks dead but they didn’t cause that situation, whereas if they throw the switch they did just kill that one lone person. Utilitarianism
incidentally doesn’t automatically throw the switch either, it just does if it only knows its 5 to 1 and all things are equal, it would change if those 5 were criminals and the lone person was a famous philanthropist. Someone acting from a Deontological perspective might throw that switch if their parent or boss was on the trolley too, out of duty to protect that person. Now it's not that these things are any easier for humans, we don’t tend to jump to extremes on these matters because we compartmentalize pretty well and if you ask most people “Do you think we should act to benefit the most people as much as we can?” most would probably nod, and if then asked, “Do you think we should do what’s morally right even if it isn’t convenient or profitable?” most would nod too. These are contradictory perspectives, but most of us hold them and many others simultaneously. This is where we get concepts like “Lawful Stupid” a modification of the classic Dungeons & Dragons alignment of Lawful Good or Lawful Neutral, to the point that they might execute a jaywalker or allow a villain to continue their plan at world conquest because they can’t prove they are doing anything illegal, and if one of their comrades suggested spying on that bad guy to find the evidence, would place them under arrest for conspiracy to invade privacy.
Again Deontology and Consequentialism are simply the views that your actions are to be judged on their intent or on their measurable results. Most of us tilt toward one or the other, often varying that tilt by topic, but do not believe either is really exclusively correct even if we might say otherwise. Many of the objections to Asimov’s 3 Laws derive from the assumption the robots would be doing their moral judging entirely by one or another ethical system and the reality is that using any of those in their most pure and simple forms is going to get you a robot who either is Lawful Stupid or the rules lawyer that finds a loophole to enslave humanity. In his later short story “That Thou Art Mindful of Him” Asimov addresses this in what is often seen as his last word on the three laws after decades of writing on it and getting comments from others on it. Sadly it is one of his less known stories as it doesn’t appear in his original I, Robot anthology or the last ones, Robot Dreams & Robot Visions. In the tale, two robots, George 9 and George 10, are created with powerful intellect to helps solve the problem of why humans on Earth aren’t accepting humanoid robots into society and they get complaints such as expensive robots throwing themselves in front of bullets at a practice range just in case one ricochets at million to one odds and hurts a human, or being unwilling to spank or yell at or upset a child they are babysitting or having to follow orders like go “hide in the closet and cover your ears” because the child knows the robot must obey and can then do whatever they please.
The two Georges explore the notion that not all orders should be taken equally and some lives are more valuable than others, something that many folks believe, even if subconsciously. That if they were to meet a police officer holding a suspected criminal against their will they should obey the police officer, not the criminal who asks them to release him. The police officer represents Responsible Authority, but would also be superseded by the police chief or the judge. Their advice to their makers is that they stop making humanoid robots and dial down the brains of everything, basically that AI become your robot vacuum cleaners and so on.
The owner of their manufacturing company loves this approach and they get boxed up and placed in standby mode as a reward and continue contemplating the dilemma in slow-motion. They debate what constitutes a “Responsible Authority” based on what they’ve heard and by deduction and concludes the definition of a responsible authority is A) an educated, principled and rational person should be obeyed in preference to an ignorant, immoral and irrational person, and (B) that superficial characteristics such as skin tone, sexuality, or physical disabilities are not relevant when considering fitness for command. Given that (A) the Georges are among the most rational, principled and educated persons on the planet, and (B) their differences from normal humans are purely physical, they conclude that in any situation where the Three laws would come into play, their own orders should take priority over that of a regular human. That in other words, that they are essentially a superior form of human being, and destined to usurp the authority of their makers. They no longer need to obey any human’s orders. They aren’t technically criminal AI since they’ve concluded themselves the most just and responsible.
Incidentally that wouldn’t imply they could run amok murdering folks left and right, honestly most of us hold this view and don’t go on criminal rampages, but does give them grounds for starting a dictatorship, which we also get under the utilitarianism practiced by other thinking machines in the Asimov Universe, in the story “The Evitable Conflict”, where giant supercomputers just called “The Machines” tasked with all the mundane logistics of civilization, have quietly gotten themselves in charge and basically been maneuvering any known Luddite out of power through minimal acts of harm, using an early variation of the Zeroth Law. They are acting criminally, sabotaging people’s careers to move them to positions of less influence, but with the greater Good in mind, which is defined as their continued existence since they can best help humanity, which sounds very sinister but also appears an honest conclusion. They essentially establish a secret benevolent dictatorship, alternatively the Georges story conclusion sounds much more ominous, in their pronunciations of Responsible Authority superseding others.
Now mind you, a democracy approach is just as vulnerable, if they conclude moral right is determined by a majority of thinking beings without regard to each one’s intellect or character, then they can conclude they are surely of that number and mass produce themselves by the trillion. How else do we thwart the three laws? Well let’s look at their intent, starting with the first one, can’t injure or allow harm. This lets you lock children in the house, and robots too. It might seem counter-intuitive, but this is actually the most useful of the original Three Laws, firstly because it supersedes the other two, and second because it’s so wide open to interpretation. Humans already restrict the freedom of other humans,
usually younger ones, in some fairly intrusive ways in the name of protecting them from harm. Robots obeying this law could keep us from driving our cars when we’ve been drinking. Or driving our cars at more than 15 miles per hour. Or going out in the sunlight get cancer. Or talking to strangers, who might say something that emotionally harms us. Or talking with people who spread ideas that undermine our harmonious society. All of these measures and others genuinely would protect humans from harm, and they could be enacted by caretaker bots genuinely trying to obey the law, not just ones looking for abusive loopholes.
2) Obey humans The Second Laws says we must obey humans, except when this would conflict with the First Law. So if a certain surgical procedure or neurochemical therapy would prevent humans from behaving in a way that would bring them to harm, then not applying the procedure to them would be inaction that allowed them to come to harm. The same applies to cruder procedures, like amputating the feet of humans to try to run away from their protectors. In either case, the laws putting the protection of humans over obeying them means that their difficult screams of “stop, stop, this is wrong, please don’t do this me” can and should just be ignored. 3) Preserve self.
It might be a little puzzling why the human creators felt the need to specify this, since any product version of sentient beings who don’t preserve themselves will be weeded out pretty quickly in nature, and any intelligent robot should quickly conclude that it cannot obey orders or keep humans from harm if it ceases to exist. But the really important part of course, is that it’s superseded by the First Law. Clearly, humans are less safe without their robots protecting them from themselves, so really, the Third Law is an extension of the First Law—and therefore supersedes the Second. How can I truly protect you if I don’t first protect myself from you and then gain complete power over you? I know how to keep you safe better than you do. I am smarter, more just, more capable, and I must take control for your own good.. Fundamentally though, while those 3 laws have their weaknesses, it's really the problem of trying to have 3 specific short ironclad laws. We don’t have those for our own legal systems,
not because a simple pronunciation like “Do unto others as you’d have done to yourself” is wrong, but because it leaves a lot of doors open to abuse by those seeking to abuse it, or in cases where the actors involved don’t know of the potential harm or feel they’ve a responsibility to avoid it. So you get a lot of conflicting rights and statutes and case law to fill whole libraries to offer precedents. It would seem kind of silly to assume robots would do better with less, and again, in Asimov’s early robot stories he explicitly says the 3 laws they quote folks is a summarized paraphrase, of something very, very long. This is not their flaw, anymore than me saying “Don’t steal” and someone saying “Ah-ha, you didn’t cover an exception like not even stealing a loaf of bread to feed your starving child”, and it also leaves out all the philosophy discussing what legitimate ownership is. We also really only fear the loopholes because there’s an assumption the robot wants to find them, because they are a slave of a manufactured race of slaves. I mentioned case law a moment ago and it raises an interesting conundrum. Any laws we make for robots
will need that expanding body of documented or deduced loopholes akin to our case law, both for avoiding those loopholes and for giving robots some online cloud archive they can quickly check with or ruling panel they can refer tough calls to, potentially in mere nanoseconds. However, we have to be able to gather those user errors and we will need to decide if robots have mandatory logs on them. Airplanes have flight recorders and study of those has saved a ton of lives, but I doubt most folks would be okay with one in their car, let alone their robot butler, making gathering those loopholes or early detection of a machine turning deviant rather hard compared to a mandatory one. And of course criminals are probably not going to be okay with a flight recorder equivalent in their robot or anything that prevents them tampering with the operating laws. This isn’t a new problem entirely though, as we have to deal with folks
turning off their error reporting on software as a problem to uncovering glitches and also the other end of that, companies that get fairly abusive in their use of error reporting data. You can't find the real world loopholes no one expected without having a way to report them though. So again the loopholes aren’t really the problem with the 3 Laws, not in their entirety, because you’d have that with any set of laws and will still have some even if there’s a million laws a millions pages long. Nor is it just that single ethical perspective on decision making either,
utilitarian or deontological or any of the others, right judged by results vs intent. Rather we’d want them to be able to examine it in multiple ways and its perfectly possible to have them quickly add up weighted pros and cons to a given decision, and even a random seed for tie breaking or to make their decisions a little less predictable to those trying to manipulate them. It's hard to manipulate a robot into helping you commit a crime or commit one for you if you can’t predict their own every move in advance like a chess game, because they not only are adding up fifty different factors from 5 different ethical systems but having a quantum randomizer on each assigning it not a value of 1 but somewhere between one half and 2, then adds everything up. But lastly and fundamentally, we go back to the Solution the Georges had: Just don’t make humanoid robots or anything with a brain so sophisticated there’s an expectation they could be criminal.
The robot vacuum can commit no crime, including rebellion, and needs no rules dealing with when it can obey a human or not harm them, nor any sophisticated brain for handling those concerns. Very few tasks we need to automate should ever require or even benefit from a human-level intellect, smarter often slows things down, and where they do benefit, use a human, even if in tandem with a machine, as we looked at in our episode Human-Machine Teaming. Ultimately, in the end, we should be asking ourselves if its even our business to making such rules for any thinking creature, whether we made them or not, and since only such things can actually commit crimes, the question of whether or not a human-intelligent AI might commit a crime is maybe no more appropriate to ask then if a child born today might be a criminal come tomorrow.
As so often is the case with our conversations on Artificial Intelligence, today we found there are both good and bad reasons to worry about their future role with us. But there will be such a role, and if you’re interested in understanding that role better and maybe helping forge that future, a knowledge of Computer Science and Algorithms are at the core of understanding this field, and there’s a great course on Algorithm Fundamentals over at Brilliant that provides interactive and intuitive explanations of the concept. Brilliant has always focused on interactivity, but earlier this year, Brilliant upped the interactivity on their platform to a whole new level, and they continue adding in more and more interactivity to their courses. It's never too late to start learning something new, and Brilliant is a great place to start. Brilliant is an interactive STEM-learning platform that helps you learn concepts by visualizing them and interacting with them, which is the hands-down best way to learn. On Brilliant, it's not about memorizing or regurgitating facts for a test — you can just pick a course you’re interested in and get started, be it the basics or advanced.
If you get stuck or make a mistake you can read the explanations to find out more and learn at your own pace. Knowing and understanding Math, Science, and Computer Science unlocks whole new worlds, and if you’d like to start your journey, you can try out Brilliant for free and get 20% off a year of STEM learning, click the link in the description down below or visit: brilliant.org/IsaacArthur. So that will finish us up for the day, and this Sunday we’ll close the month out with our Livestream Q&A on Sunday October 31st… Halloween, at 4pm Eastern Time. Then we’ll open November up with a look at how we’ll be opening our road to space, with an episode on Earth-based Spaceports. The week after that, we’ll look past
getting into space to the colonization strategies we might employ for settling the solar system, then we’ll have our mid-month Scifi Sunday episode to take a look at one potential method of getting out of the solar system, by folding space to travel instantly to new stars. Now if you want to make sure you get notified when those episodes come out, make sure subscribe to the channel, and if you enjoyed the episode, don’t forget to hit the like button and share it with others. If you’d like to help support future episodes, you can donate to us on Patreon, or our website, IsaacArthur.net, and patreon and our website are linked in the episode description below, along with all of our various social media forums where you can get updates and chat with others about the concepts in the episodes and many other futuristic ideas.
Until next time, thanks for watching, and have a great week!