In our modern world computers are everywhere. They’re in our phones, airplanes, bureaucracies, communication systems, and even in our microwave ovens; yet politicians seem to have very little understanding of how these machines work. Hopefully this (admittedly lengthy, though hopefully interesting) blog post will be able to help.
But first, what is a computer?
Let me tell you a story.
Once, a very long time ago, there was a war. On the one side there was a dictator who wanted nothing short of the complete annihilation of his enemies. He was a man whose very name would someday become synonymous with genocide.
In his quest for world domination, the engineers who worked for him invented many new weapons. They had invented explosives that could be propelled to their destination on their own. These “rockets” as they were called were not particularly good at hitting their targets though. The rockets couldn’t think. They couldn’t understand where they were going. They were, after all, just machines.
On the other side of the war there were the allied forces. Those who knew exactly what this dictator wanted, and knew that he had to be stopped.
The communication lines used by the dictator used a special machine called “the enigma”. It was a powerful cryptography device. The only way to crack the code would be to crack it before the end of the day. If you managed to crack the encryption code after the end of the day then it wouldn’t matter because they changed the encryption key every day.
The world’s best minds simply couldn’t keep up with the enigma machine.
Machines can often do tasks faster than humans can, but this particular task required a lot of thought. Which naturally raises the question: can a machine be made to think?
A mathematician named Alan Turing believed that, yes, machines can indeed think. He believed that he invented a machine that could do exactly that. He had invented the Turing machine.
To create a Turing machine you take a tape that is infinitely long, and divide it into segments called “cells”. Each of these cells contains information. That information could be a number, or a letter, the location of another cell, or, importantly, an instruction. Next you would modify many of these cells to contain a series of instructions. These instructions would be read by the read/write head which would follow the instructions and modify the information on the tape based off of what those instructions tell it to do.
The modern computer, to put it plainly, is a machine that follows instructions. It’s a turing machine. The only difference between the turing machine and the modern computer is that modern computers have a finite amount of memory circuits instead of infinite tape, those circuits are divided up into bytes which are made up of ones and zeros, and the read/write head is replaced with a processor circuit that is made of logic gates.
The modern computer only understands certain commands. These commands include things like moving information around memory, doing basic math operations like addition, subtraction, multiplication, and division, comparing two piece of information on the tape, and jumping from one instruction to another based on whether or not those two pieces of information are the same (that can mean deciding whether or not to skip over many instructions, or jump back to an earlier instruction).
These Turing machines, as it turns out, can actually do any kind of thinking that’s possible. They can, in theory, be programmed to think exactly like a human (although we don’t yet know how to do that), or they can be programmed
to think in some other way. The only kind of thinking they can’t do is the kind of thinking that would solve a paradox since paradoxes, by definition, are unsolvable.
Most people don’t know how to troubleshoot computers because computers don’t SEEM like machines that blindly follow commands.
The commands that modern computers follow go all the way down to the level of controlling the electrical signals that go to the screen of the device you’re reading this with to tell it how to display this message.
So now let’s dive into computers and see what needs to be known about them to deal with modern computer-related issues.
1. Operating Systems.
In the early days of home computing the programs that computers ran would typically assume full control over the device. If you hooked up a printer to the computer then the program that’s currently running would talk directly to the printer. Since we live in a capitalist society, there’s more than one printer company out there.
This means that different printers would communicate with the program in different ways. This meant that one computer program might only be compatible with a certain number of printers. This, of course, is a terrible way for things to work.
Thus one day kernels were invented. A kernel is a special computer program that talks to any hardware hooked up to the computer on behalf of the computer programs running on the machine. This way if a new printer comes out the kernel can be updated to work with that new printer, and now all the programs you can run on your computer will be compatible with that new printer. The kernel also enables the computer to run multiple programs at the same time.
The kernel, and all the programs that come with it, all come together to make what’s known as an “operating system”. Modern personal computers need operating systems
in order to work. Jake Roper from Vsauce3 made a great video on the subject here.
Unfortunately most computers come with the Microsoft Windows operating system installed by default. It can be hard to find a computer that doesn’t have Windows already on it. At the current time of writing this isn’t too huge of a deal since Microsoft is currently giving away copies of Windows for free but, as I’ve talked about in a previous blog post, software that’s given away for free can be, and often is, a scam.
Microsoft’s entire business strategy is vendor lock-in. That’s why they have so much money, and why they had so much money in the past. There are alternatives to Windows such as Linux, but most people don’t know how to replace their operating system. This problem has gotten better in recent years since Google started selling their own laptops (chrome books) with their own operating system (chrome OS), but still it would be better
if it were easier for all of us to move away from using Windows.
2. The Internet.
In order to understand the problems the Internet faces we need to know how it works, and what it is. The Internet is a giant computer network that connects almost every computer on Earth (and possibly some not on Earth) together.
Like any computer network it’s goal is to allow a computer program on one computer to talk to another computer program on another computer. Computer programs can actually talk to each other and, in fact, that’s what they statistically spend most of their time doing. There are many languages that computer programs use to communicate with each other. A language that a program uses to talk to another program is called a protocol.
One such protocol that computers use is called the internet protocol or IP protocol for short. When one program wants to connect to another computer over the Internet it needs to know what program it wants to connect to, and what computer it wants to connect to.
Every computer in the world has an associated IP address which is what computer programs use to specify what computer they want to connect to. The IP address is just a series
of four numbers.
The Internet is running out of IP addresses. There can only be up to 4.2 billion of them, and we actually ran out a really long time ago. Thankfully there’s a replacement called IPv6 that will give us more addresses than we could ever need, but to implement it we’d need to change a lot of infrastructure, and nobody wants to do that. There is a short-term solution that the Internet has been using called NAT (network address translation), but that can’t hold forever.
3. The client/server model.
When a program wants to connect to another program, one of them needs to be the client and one of them needs to be the server. On the Internet the way that works is one program tells it’s kernel that it wants to be a server program, and then the kernel allows it to accept connections from clients over the Internet.
Then the client will specify to it’s kernel what IP address it wants to connect to, and what program it wants to connect to and then the kernel will send connection requests (which are in the IP protocol language) over the Internet and then the server program on the other end of the Internet will decide whether or not to accept the connection or reject it. If it accepts then they can send messages to each other. If it rejects then the client will get an error message and have to decide what to do.
There isn’t really a problem with this except for the fact that the internet service provider company can detect what computers your machine is connecting to by simply looking at the IP address in each internet protocol connection that gets made through it. One thing that companies can do is artificially slow down the connection to certain sites and offer to speed the connection back up to normal speeds in exchange for more money. This is where net neutrality comes in.
The idea of net neutrality is that internet service providers may not artificially slow down anything.
4. Domain Names.
In the very earliest days of the Internet if you wanted to connect to another computer you needed it’s IP address. When the Internet went mainstream people wanted
a better system for connecting to computers so domain names were invented. Here’s how it works: when a program wants to connect to another computer, it doesn’t just use the IP address. Instead it asks one of a few special computers called the “domain name servers” what the IP address for the domain name is.
So if I want to connect to google.com, I would tell the program to connect to google.com, and it would ask the domain name servers what the IP address is for it, and then it would have the IP address and would go about connecting like normal.
In order to get a domain name you need to pay for it. It’s kind of like a phone book except that it’s a phone book that you need to pay to put your name in, your phone number can change automatically, and you have to call a certain known phone number in order to ask it what the phone number is for a given person.
Originally the idea was that there would be more than just eight of them. The idea was that every country could set up their own DNS servers.
The more DNS servers there are, the better because if those eight servers went down all at the same time then the Internet would stop working almost entirely.
5. Search Engines.
When the Internet was invented it originally was much more difficult for the average person to use. Originally they had to write computer programs to communicate with each other over the Internet.
There were several protocols people experimented with and each had their own associated client and server programs, but one protocol would one day become synonymous with the Internet in the minds of ordinary people:
the web. The web is not the same thing as the Internet. The web is something built on top of the Internet. It’s a protocol (called “hyper text transfer protocol” or “http”) that’s used by the client (called a “web browser”)
to access documents from the server (the documents are called “web pages” and the servers are called “websites”).
These web pages often contained references to other web pages (called “hyperlinks” or “links” for short). This way all the web pages on the Internet were connected in a giant web of connections. To navigate this web special websites called “search engines” came about which enabled people to simply type in a few words into a text input and then press a button and then a web page would be automatically generated (Yes, GENERATED. The web page wasn’t made by a human, a computer program on the website would create it on the spot), and then you’d get a list of links to web pages that contained the words you typed in.
This way you could quickly find information about whatever topic you wanted to even from sites that you had never heard of and might not have been able to stumble onto yourself. There was no more “stumbling onto” things. It could all be pulled from a search engine. Just ask it a question and you’d get the answer instantly.
In order to make money off of having your own search engine you would generally use advertisements. The user would search for something and get a relevant advertisement for that thing as well as their search results. This isn’t a bad thing on it’s own, but what many find creepy is that while you’re learning things from the search engine, the search engine is learning things about you.
Many search engines, such as google, are constantly gathering data on everyone, feeding that into a supercomputer, and figuring out what kinds of advertisements might be relevant to you. Companies like google gather vast amounts of information on everyone. Having privacy on the Internet isn’t easy because of the sort of power for spying on people that the likes of google have. Speaking of which…
One of the things the Internet was supposed to be good at was anonymity. It works like this:
Once you’ve been banned off of a website for spreading the absolute TRUTH about the rampant corruption that the administrators of that site engage in (and I’m only half joking, some of these people really do accept bribes to keep
certain messages hidden), you can create another account and do it again.
Or can you? This led to many sites keeping track of the IP addresses that a banned user came from. So to get around this proxies were invented. The idea is simple: instead of directly connecting to the website,
your computer instead connects to another computer called a “proxy” which will then connect to the site for you and then when you get banned again you can just use a different proxy.
This also helped prevent the websites from collecting huge amounts of data on the users.
This method became so popular a way of ensuring anonymity that it was expanded upon with the TOR network. The tor network works like proxies, but it’s an entire distributed network.
The way it works is you connect to a computer called an “entry node” and then that connects to another computer, and that connects to another one, and so on until one of them connects to an “exit node”
which then connects to the website. This creates layers of encryption and makes it very hard to spy on people.
Unfortunately people have found ways to fight this. All a site needs to do is use the tor network to connect to itself again and again and then it knows which ip addresses are tor exit nodes, and can refuse to accept connections from those computers.
Also any proxy that an ordinary person can know about can probably also be known about by the website, and so they know other ip addresses to ignore. This problem is still being worked on, but to make a long story short there is no perfect solution to getting anonymity online.
Anonymity is important. Without it the people can be spied on by corporations and governments alike, and censorship can only happen when they know who is saying the messages that are being censored. While it is true that hackers generally use anonymizers to remain hidden, it’s important to understand that there’s more to hacking than just using anonymous connections, which brings us to:
In computer programming a bug is a mistake that a programmer makes when writing a computer program.
You may recall that computer programs are just lists of instructions for the machine to follow. Some programs are larger than others though. Some programs have many instructions, while others have very few. As a general rule the more features a program has, the more instructions it’s made up of. Some features take up more instructions than others, so it’s hard to say how many instructions a given feature will have.
The more instructions there are, the more places there are for bugs to exist. In theory it’s possible to write a program that has no bugs; however in practice it isn’t.
In my experience there are generally two types of bugs: the first kind are the ones that you’d never know existed because you’re a normal human being who’s NOT constantly trying to break things, and the second kind are the kind that ordinary users will notice because they show up when just using the program normally.
For example: I’ve heard that in the early days of Amazon you could order 0.1 of a product and get a 90% discount. I don’t know if that’s true or not, but that’s the perfect example of the first kind of bug. Nobody who uses Amazon normally would think to do that, while there’s a very specific type of person who would immediately think to try that.
A great example of the second kind of bug is how when you play a video on the youtube app for Android it will sometimes briefly show the first frame of the wrong video.
Most proprietary software is full of bugs. While most freedom software has far less bugs but can often lack certain features that people need.
“Piracy is almost never a price issue. It’s almost always a service issue.”
-Gabe Newell, CEO of Valve Software.
One of the problems with writing software is finding a way to make money off of it.
The most obvious method is to sell the programs as if they’re physical objects; however this brings with it a problem: what if the users just make copies of the programs
and give them to each other so that only one person would ever need to buy the program? This is where DRM (digital rights management) comes in. DRM is any kind of system that would attempt to stop piracy.
Unfortunately for the software companies there’s nothing that really can be done to stop copying from happening. No matter how you write the program, so long as the program is running on my computer,
I can just modify the program to not do any of the things that would prevent it from being copied and used by everyone.
The only defense that software companies had against this was to get the government to make it illegal to modify computer programs.
Software companies have tried many different approaches to implementing DRM, but the DRM systems tend to do very little to stop piracy, and tend to just annoy the customers who payed for the product.
In response to this many programmers added end user license agreements that explicitly allowed people to modify the programs, but that’s another story from another blog post.
That DRM exists at all. It’s not real technology and it tends to just make life harder for the people who payed for the software.
There’s nothing that can be done to make it impossible to pirate a program. Except…
To prevent people from being able to copy programs software companies are now experimenting with the idea of having the program run on their own servers that they have full control over, and simply allowing their users to connect to these servers and remotely use the program. This method for prevent piracy is almost completely fool proof at stopping piracy from happening, and it only comes at the cost of customers never being able to preserve a copy of the program.
Soon enough it may be the case that you could pay hundreds of dollars for the chance to “buy” a program which could stop working at any time. Back when people could have a copy of the actual program on their own computer they could just keep using it forever (unless it was programmed to slow down over time or just quit working one day, which is something that many companies do).
With this system, the instant that the program becomes no longer profitable the company can just shut down the server computers running it and then the product would stop working entirely.
This has huge implications for the user and those implications are all bad. Keeping copies of old programs is just as much a form of archival as keeping copies of old newspapers or old books. Some people put a lot of effort into keeping copies of old information so that society as a whole can remember it, and this threatens archival at a fundamenetal level.
Now that we’ve connected to the Internet we need to start thinking about security. Unfortunately Windows is not secure. At all.
Thankfully though, Windows does give us some nice examples of how NOT to do things. Security is a complex topic, but as I’m sure you know it generally involves attackers (or what might call a “hacker”) who tries to break
past the security system of something they’re hacking, so let’s begin there. Remember before when I mentioned that there’s a certain kind of person who’s always looking to break things? That’s what a hacker is.
There are a few different kinds of hackers: the first are black hat hackers. These are the kinds of hackers who hack into systems illegally. They generally do it using two kinds of attacks. The first kind of attack is where they exploit bugs, the second kind is where they overload the target with information that’s coming in too fast for the computer to process it. There’s a reason why programmers often care about fixing all bugs, including
ones that no ordinary person would ever come across. It’s to prevent black hat hackers from getting in.
Unfortunately most programmers aren’t as good at hacking as a hacker is, so that’s where the second kind of hacker comes in. White hat hackers are penetration testers (or “pen testers” for short) who break into computer systems so as to check for ways to get in.
Companies pay white hat hackers to test their systems for them, and some of them get payed a lot of money for this service. The way they do it is they try all kinds of things to try and find bugs in the program that can be exploited and then they report those bugs to the programmers. Of course, some people would rather keep those bugs a secret in case they want to hack into systems. Those are called “gray hat hackers”. They do a bit of both kind. Finally there are red hat hackers. They’re hackers who work for the government and hack other governments.
Most companies that have servers such as twitter, youtube, facebook, etc tend to have security that’s good enough to keep amateurs out, but not good enough to keep a determined professional out.
Some companies care more about security than others, but what they all have in common is that they all only care as much as they’re forced to. Freedom software projects care a great deal about security, while proprietary
software tends to care not at all.
Encryption, to put it plainly, is the act of modifying information so that it only has meaning to it’s intended recipient. Obviously there’s a lot that can be explained about cryptography, but for now let’s focus on encrypted connections over the Internet.
When a computer forms an encrypted connection to another computer it sends it’s messages to that computer through intermediate computers. That’s just how the Internet works. Now if computer A (“Alice”) forms an encrypted connection with computer B (“Bob”) then the messages have to pass through intermediate machines. If one of those intermediate machines is trying to spy on Alice and Bob, then they could take the messages from Alice, pretend to be Bob, form an encrypted connection with Alice, and then form an encrypted connection with Bob.
This way the spying computer could take messages from Alice, pretending to be a simple messenger, decrypt those messages, read them, and then send them through it’s encrypted connection to Bob.
This is known as a man in the middle attack. To solve this problem digital signatures are used. A digital signature is a way to make sure that a message was sent by the person it was supposed to be sent by and that it couldn’t have been modified.
The way that works is your computer generates two files (called “Keys”). One is the public key, and the other is the private key.
The idea is that you give the public key away to anyone who will accept it, and keep the private key a secret sort of like how you’d keep a password a secret except that this password is too large to remember and so it has to be kept in a file.
The private key can be used to encrypt files in such a way that they can be decrypted by the public key, and ONLY the private key can encrypt files so that the public key can decrypt them. Of course the public key is meant to be given away to everyone so this isn’t good as a means of encryption, but it is good for digital signatures. If Alice sends a message to Bob, she wants to make sure that any man in the middle can’t modify it (that’s how man in the middle attacks work. They require the man in the middle to modify a message sent by Alice). So to do this she encrypts it with her private key.
The man in the middle receives this encrypted message and can decrypt it since this man has Alice’s public key just like everyone else; however he can’t modify it and reencrypt it.
To modify a message is to make a new message, and in order to encrypt this new message the man in the middle must have Alice’s private key, which he doesn’t have. The only option he has is to send it to Bob. Thus Alice and Bob are able to form an encrypted connection that the man in the middle can’t spy on.
This is exactly what websites do except with one small problem:
How does the website get it’s public key over to the web browser? The website can’t just digitally sign it and then send the digitally signed version because the only way to verify that the digital signature is legitimate is to already have the public key.
This is where trusted third parties come in. The idea is simple: any website that wants to form an encrypted connection pays a trusted third computer to digitally sign the site’s public key with
the trusted site’s public key, and the web browser happens to already have the public key of the trusted third party and can thus receive the public key for any site. This might seem complicated but it’s really not as complicated
as it sounds.
This system relies on trusted third party companies whose whole job is to sign public keys for websites and NOT make it possible to spy on people’s connections.
Naturally this is not a very good system because third parties might be bribed into compromising their clients’ security. Nobody has been able to come up with a better system than this, unfortunately.
Another problem is that all forms of traditional cryptography could be broken by a new technology called “quantum computing”. Quantum computing makes it possible to break any kind of old encryption system, and makes it necessary to have quantum encryption systems. So far nobody has been able to make a quantum computer that can deal with enough information to be able to break the current encryption systems that we have, but that could change soon enough.
Cryptocurrencies are currencies that are specifically designed to be decentralized. Decentralized means that there is no one central server that could take the entire system down if it went offline.
The way it works is that there’s a ledger that keeps track of the transactions that people make using the cryptocurrency. This ledger is distributed across the various mining systems. A mining system is just a system that keeps track of the ledgers, tries to solve a complicated cryptographic problem for that ledger, and in return it gets a samll amount of currency added to the system that it’s owner gets to keep as a reward for keeping things running.
There’s no regulations in these economies so scams are rampant. Also the exchange rate fluctuates wildly for some reason.
13. Video Games.
In 1982 Disney released the movie TRON. TRON was a fictional story about what the inside of a computer was like. It was a nice little story in which programs were played by human actors wearing silly little glowing costumes, and they played video games against each other to the death like Roman gladiators. Of course what the digital world is actually like (as discussed at the beginning) is a bit different. What is the inside of a computer world like?
It’s whatever we tell it to be like. By teaching the computer how to display triangles in three dimensions it’s possible to create what look like actual 3D objects since all 3D shapes can be approximated as a bunch of triangles.
In these virtual worlds it’s possible to create things (called “objects”) that can do stuff. This has been used extensively to make video games. Each individual object can be given it’s own programming and players can interact with these objects using some kind of input to the computer.
That input can be a keyboard and mouse, or joysticks, or a motion controller, or whatever else we can use to communicate with a computer device. In the early days of computer games people would generally play a game by themselves. In first person shooter games (“first person” meaning it’s seen from a first person perspective, “shooter” meaning there’s a gun in the corner of the screen) the players would often fight some kind of enemies by shooting at them. The enemies were objects just like everything else in the game.
The objects would have instructions associated with them telling the computer where the enemy should be, what the enemy should be thinking, and, importantly, how to display the enemy on the screen.
Eventually people became very good at fighting enemies in video games and the enemy characters in these games didn’t actually get better at all since they weren’t programmed to be able to learn.
Programming enemies in video games isn’t easy, and they often struggle to keep up with the skills that some of the best gamers have. So to deal with this game developers released games that allow gamers to play against each other online.
This way the best of the best could have an actual challenge.
There are many problems plaguing the world of video games. For one, there’s rampant serious abuse of the game developers from management. For another there’s the problem of gambling mechanics, micro-transactions, and pay to win. There’s also the problem of DRM.
Many games have tried offering items that you can pay real world money for that will either make your character look cool or (in some cases that nobody likes) give you weapons or items, or abilities that give you an advantage over other players (that’s called “pay to win”).
This isn’t entirely a bad thing though. Some games are “free to play” meaning you can download and play them for free and the developers get their money from selling virtual items.
Some games have started adding a particular type of virtual item called a “loot box”. The basic idea is this: after every round of an online game, the player might get a box that contains items they may want. They have to pay real world money to unlock the box and get a random collection of (mostly worthless) items. Only occasionally do they get something of value.
Needless to say, nobody likes this gambling mechanic, and many people have become addicted to it the same way that people become addicted to gambling. For more information, I would recommend this youtube channel.
We’ve seen so far just how widespread the use of computer chips is, and the invention of the Turing machine is arguably one of the most significant inventions ever.
These computer chips are used for basically everything in engineering precisely because they can, in theory, be programmed to do anything. So is it any surprise then that they would eventually be used for medicine?
Sometimes people lose limbs, sometimes people have organs that fail, sometimes people get bored of having an ordinary human body that can’t be upgraded and want something more. This is where cybernetics comes in.
Of course cybernetics is still a young developing field, but it’s come a long way from the days of peg legs, and hooks for hands to today where we have cybernetic arms that have fingers which can move in response to the movements of muscle groups in the arm.
While machines aren’t yet able to decipher the signals that the brain sends to body parts to control them, and they certainly aren’t completely ready to be installed into people’s brains so as to give them the ability to control machines with their thoughts, that technology is closer than most people think. Vsauce2 did a great video on cyborgs here.
As I’m sure you’re well aware, healthcare is expensive, and the thought of getting a brain implant just because you’ve always wanted a computer chip in your brain is a ways off in a political climate that’s skeptical of the idea of giving amputees replacement limbs for cheap. Not to mention the small issue of making computers secure enough for being used like that. As mentioned earlier, software companies tend to only
care as much as they’re forced to.
The idea of a machine that blindly follows orders is exciting; however there is one small problem: the instructions it understands are too low level. We can’t give it a high level command like “Go make me a sandwich” because it can’t understand that unless we provide extremely detailed instructions on how to do that.
One thing we need is to give it the ability to see and recognize what bread looks like. The first part of that challenge has been thoroughly solved. We can hook up cameras to the computer that allow it to see things, not just in the visible spectrum of light, but even in other spectrums such as infrared and ultraviolet. To put it plainly: there are cameras out there that can see in colors that the human eye can’t see in.
The next problem is how to get the computer to understand what it’s looking at. When the computer receives an image from the camera that image is just a 2D grid of numbers. Somehow we need it to identify if the image is of a slice of bread, and where the bread is in the image. That would be a good starting point.
This has proven to be an incredibly difficult task. How is it that the human brain is capable of recognizing nearly any object and yet computers can’t? Somehow something is going on inside the human brain that enables it to do this. So that’s exactly where the research is being focused. Computers can simulate things.
They can keep track of anything whether those things are atoms in a chemical reaction, or the planets in outerspace, computers can simulate them all. In fact computers can even keep track of things that don’t exist. Like for example: the exact electrical signals going through a chunk of brain matter which doesn’t actually exist in the real world.
By simulating a chunk of brain cells we can teach the computer how to recognize bread, peanut butter, jelly, or any number of different objects. We can also use these artificial neural networks to recognize human speech, or certain sounds, and much more.
The only problem with modern artificial neural networks is that they learn very slowly. It takes thousands or even millions of images of bread, and images that don’t contain bread (which must be carefully labeled by humans) in order for the neural network to learn how to tell the difference between the two types of images. There’s active research going on
right now to try and make neural networks that can learn with fewer samples.
Of course AI can be used for more than recognizing objects and sounds, it can also be used to pick up on any pattern that can’t be easily described by computer code, such as finding trends in the stock market, or finding a complicated pattern in human language that enables it to translate from one language to another, or playing a game automatically, or anything that requires it to learn how to recognize patterns.
This ability enables us to take computers from being bricks that you carry around with you (like smartphones, or laptops), and give them the ability to move around on their own. Which brings us to:
As you can image, a machine that can learn patterns can be used for many things that it shouldn’t be used for, but one that seems to want to rear it’s ugly head is crime prediction.
It’s important to bear in mind that these algorithms aren’t magic. They just find patterns in data and latch onto them regardless of whether or not those patterns are valid or just a coincidence.
“Predictive policing” isn’t a good idea because it’s just a more high tech way of having someone look through all the data and see if they can spot patterns.
Vsauce2 made a great video on this very topic that you can see here.
CGP Grey made a great video on what the future may look like when machines take over which you can see here I’d also recommend this link [insert link to wait but why AI article] on what the future will look like when an artificial intelligence is made that’s as smart as a human. To put it plainly robots are coming to take over all jobs.
Automation is inevitable, and that should be a good thing. I’ve often seen people surprised by the idea that robots can do all jobs including the jobs of designing, building, programming, and maintaining other robots, but to me it just seems like the natural conclusion to the engineering process.
I’m not trying to say that machines can be people. I’m saying that people ARE machines. People are machines according to machine theory. There is nothing that a human can do that a robot can’t do.
Humans won’t be able to compete with robots for jobs. Robots are the ideal worker. They don’t demand basic human rights, they don’t demand a paycheck, and they can be left to run 24/7.