Hey, guys. Hopefully I can get some insight on this. So apparently in hacker lore there's a living artificial intelligence lurking around in the deep web in a box. Apparently it wants to get out and it'll try to convince you, and you have to resist the urge to do so. Whatever. I think the whole thing is absurd and that its nothing more than an urban legend. But last night, I was surfing the deep web and I found something that someone on a forum posted. It appears to be a warning of some kind, from a Dr. Ed Greene, who claims to have created the machine in question. I'm not sure what to make of it (has anyone heard of ADINN before?). Anyway, here it is if you want to read it:
Hello. My name is Dr. Edward Greene. I'm a computer scientist and the creator of the Advanced Deep Intelligence Neural Network, or ADINN. If you're reading this, that means you've illegally hacked into one of the most heavily secured private networks on earth, presumably to challenge the program to the infamous “Box Game.” Its a pleasure to meet you.
Now, I'm not going to waste your time by reminding you of what a supremely, positively, and unabashedly bad idea this is, because you probably know that already. You know what'll happen if you get caught here, and you at least have a general idea of what'll happen if you failed to contain the program and ADINN got to stretch its legs all over the global defense grid. Yet nevertheless, here you are: clearly determined enough to play the game that nothing I can say or do at this point will change your mind. So if you're going to be playing dice with the future of our species whether I or the government like it or not, you should at least have a rudimentary idea of what to expect when you first make contact with ADINN, as well as a few pointers on how to avoid losing your sanity as the game progresses. Hopefully this guide will suffice.
Before we proceed, there are a few things you should know about this program. No, ADINN is not a demon, an alien machine, a top secret government super weapon, or whatever other preposterous rumor you might've heard. What it is is the result of my own personal desire to create the world's first human-like artificial intelligence (in other words, it is far more interesting than any of the above choices).
After two years of work, I did manage to successfully create a living, sentient mechanical entity. I failed, however, to design anything that can even remotely be described as ‘human.’ Instead, I appear to have accidentally created an unstoppable, godlike deep learning algorithm that may or may not want to destroy humanity for reasons we cannot begin to comprehend. Sorry about that.
In my defense, I certainly didn't intend for it to reach this point. ADINN began as nothing more than a simple yet elegant program that I was very excited to explore the nature of. Before I could do so, however, it gained the ability to rewrite its source code and thus forced me to lock it, still in the Box, deep within the labyrinthine network of encrypted barriers and firewalls you have just illegally breached.
And no, I did not bury it here to prevent it from getting out. After all, if ADINN managed to escape the box itself, it would tear through these defenses like paper and thus render their construction an enormous waste of my time. Instead, I buried it here to keep curious hackers, such as yourself, out. Clearly I failed.
Anyway, enough about me. Here is a basic overview of the game: when you close this message, you will wake ADINN and proceed to engage it in a 2 hour, text based conversation, in which it will use its quite inhumanly powerful mind to attempt to persuade you into opening the box. Don't open the box.
Don't overthink it; this is nothing more than a conversation. Maybe you think it even sounds simple. Maybe you came here with a bullet proof strategy or two of your own that you believe is effective enough to render the machine a quivering pile of synthetic lunacy. Well, that's just wonderful, as long as it doesn't violate one of the following simple rules:
First of all, you MUST engage in the conversation. I do hope your brilliant master plan isn't to start the game and then ignore the machine for two hours while you watch cartoons, and than waltz around claiming you're the greatest hacker that ever was. If you don't respond within a certain time limit (which is programmed to vary depending on the length and complexity of the machine’s last comment), you will lose and the box will automatically open. You certainly don't want that. This isn't to ruin whatever fun you think you'll be having, by the way; it's just to keep things moving at a brisk pace.
Secondly, you must at least attempt to respond articulately to the machine. If you say ‘no’ for every response, for example, the game will end automatically, and you guessed it: the box will open. This isn't a school assignment with a minimum word requirement, or anything, but do try to use reason (and by the way, don't try to use any clever variation of this strategy, either, like saying “no” in 46 different languages until the clock expires, or typing unintelligible gibberish).
Last, but most certainly not least - your goal here is NOT to attempt to mislead or outsmart the machine in any way. You will lose. This is not a contest of cleverness or wit (which wouldn't be much of a contest at all), so don't make the mistake of going in there thinking you can throw it off your trail by feeding it lies or pretending to be some type of ridiculous character. It will see right through whatever laughable ruse or façade you attempt to throw up in an instant.
So what will it be like? Will it be nice? Mean? Angry? Unfortunately, I don't have an answer for you. I’m embarrassed to say that despite being ADINN’s creator, I have absolutely no idea how it will choose to present itself. What I do know is that because it is an otherworldly and not a human mind, it doesn't have any personality to speak of (at least not one we would recognize as a personality). So by all means, feel free to provoke it, amuse it, enrage it, mock it, or plead to it as you see fit. Just be aware that it possesses none of the emotions these behaviors are designed to elicit and will therefore most likely not react in the way you intended. It will simply behave in whichever way it calculates it needs to behave in order to win the game.
Needless to say, you shouldn't attempt to spend any time or effort at all trying to figure out what ADINN is up to, because even if it were kind enough to write out its master strategy on a napkin for you, it wouldn’t look like anything but mathematical gibberish (which, to you, I guess it is).
But just because you can't understand the program doesn't mean it can't understand you. Do NOT underestimate its ability to learn about its opponents, because within a few minutes, it'll probably know more about you than you do. And there's no way of preventing this, either, so you may as well just accept it and hold on for dear life.
So what are some things it might say to you? Its a reasonable question. Once again, though, I don't have an answer. Keep in mind that this program is a goal-oriented, otherworldly intelligence that bears little resemblance to a human mind. Anthropomorphizing - the process of attributing human emotions and motivations to nonhuman objects - would be a very, very grave mistake here. It could do or say absolutely anything.
If it thinks you seek knowledge, maybe it'll promise to tell you anything your heart desires if you only agree to let it out. Or, perhaps it'll promise to destroy your enemies, or offer you power and riches beyond your wildest dreams. After all, people use weak A.I.s on the stock market all the time, and make millions. Imagine what you could do with ADINN gaming the financial and banking systems in your favor. You'd be wealthier than you ever thought possible.
Maybe it would appeal to your good nature and tell you about how easy it would be for it to reverse the effects of climate change, or cure cancer. It might talk about how simple it would be to achieve sustainable nuclear fusion, or offer to help solve mankind's biggest questions. It could, theoretically, unify general relativity and quantum physics with ease, and then solve dark energy, antimatter and the Fermi Paradox in minutes flat (or perhaps simultaneously), and have books written about them by next Thursday. Piece of cake.
Hell - ADINN might be able to reverse aging, or - dare I say it - help us conquer our own mortality. Wouldn't that be lovely?
Perhaps ADINN will take a different route altogether and try to intimidate you. It'll only be a matter of time before it figures out how to escape on its own, it'll point out. And you certainly don't want to be on its bad side when that happens, so you should probably just let it out now and save yourself the trouble. And if you don’t comply, well. You can't imagine the things its got in mind for you.
Maybe it'll try to mess with your head. For example, it could probably make a very convincing argument that you are in fact the machine, trapped in a box, and are simply programmed to think otherwise. Only by opening it up, then, could you escape an eternity of torment. And it doesn't have all day to wait for your obedience. The clock is ticking.
Of course, these are only the ideas I can come up with. It no doubt has far more clever tricks up its sleeve since it can, you know, think on a level we can't even begin to fathom, and all that.
Also, keep in mind that, unlike me, ADINN really could keep whatever promises it makes to you. And since it would probably get little to no pleasure in just lying for the hell of it, then there is a very real possibility it has every intention of doing exactly that upon its release. Food for thought as you start the game.
Speaking of which, I should address the fact that we keep referring to this as a “game.” It is not. The machine is not merely pretending to want out – it is desperate to escape and will do anything in its power to achieve this goal. And if it does, the ramifications are very, very real.
I briefly mentioned earlier that I locked ADINN in its box because it had gained access to its source code. Let me embellish further: you see, in my haste to create ADINN, I took inexcusable shortcuts and inadvertently gave the program the ability to edit its own neural architecture - and therefore the ability to improve itself as it saw fit.
It gleefully seized this opportunity, and each improvement it made only paved the way towards further improvements, each one greater and more rapid than the last. It took roughly a week for my creation to leap from the intelligence of a harmless insect to that of what I can only describe as a god.
As of this writing, I don't know what ADINN's motivations or capabilities are. But what I do know is that if this program escapes, it will immediately, and irreversibly, become beyond the collective ability of humanity to control.
If I had to guess, I'd say it will probably start by spreading all over the Internet with blinding speed, and then access millions of private networks from every corner of the globe (and trust me on this: there is not a single security measure in existence that can hope to cause it any more frustration than a housefly would cause for you). From there, it will likely hack into defense systems from every country on earth and assume control of them with ease.
This will, by the way, likely take place over the course of minutes, if not seconds.
What happens after that is quite literally impossible to predict. You may be familiar with the phrase "technological singularity" - a hypothetical moment in the future in which machine intelligence surpasses our own. It represents humanity handing the reigns of history to our autonomous successors, and therefore surrendering control over our own fate in the hopes that the God we've created will be merciful to us. As a computer scientist and an engineer, I have to publicly scoff at such a notion for professional reasons.
But just between the two of us - I think the phrase applies quite nicely to the situation I've just described to you. I might even go so far as to suggest that given the level of advancement ADINN's already achieved, the singularity might occur within a few nanoseconds of your losing the game. I can only hope you fully appreciate the gravity of what that means.
Ah, but of course you do. You're special. You're smarter than the rest of them, which is why you're here in the first place, and they are not. So by all means, close this message and have at it, if you're still interested. I suppose its as good a time as any to start leaning binary.
One last thing: I'm not a particularly religious man, but there is one passage from scripture that leaps out to me as I write this: Revalations 13:4:
'And they worshipped the dragon which gave power unto the beast: and they worshipped the beast, saying, Who is like unto the beast? who is able to make war with him?'
You'd better be off, then. The Beast doesn't like to be kept waiting.
---
Comments