I owe you all an apology. This is all my fault.
I know it’s frustrating, how the machines and algorithms that are meant to be better at this than you, who were meant to free you from this work, seem to always fall short of their early promise. They’re like that gifted kid in school, who all the children hated because he always knew the answer and never found things hard, and who the teachers quietly pitied, because they’d seen this all before, and knew that such blessings rarely last, and that their burdens last much longer than their gifts.
My fault.
Not the gifted kids. I mean the algorithms. It’s my job, you see, to make them fail, to plot their spiral down to disappointment and frustration.
Well, okay. I say it’s my “job”. The truth is, actually, I’m self employed. I’m available for hire, with corporate work a speciality, but the secret truth is, I’d be doing this work anyway. Getting paid for it is just a bonus. I’d be grateful if you wouldn’t share this with the suits who fund my work. They’re overjoyed with my results, as their competitor's latest unveiling devolves into a fiasco. They have no idea I’m being equally well paid to do the same to their own top secret next gen babies.
I mean, they do suspect that this is sabotage. They do assume their competitors engage in the same sort of dirty tricks that they themselves do. They just don’t realise that they’re hiring the same guy. In their minds, there is a clandestine war being fought. Teams of hackers on one side versus teams of hackers the other side. It suits their egos, to be a general or a spymaster. To have armies. That’s why they never will never see the paltry, rather pathetic truth of what is actually occurring here. It’s just me. That‘s all. No cohorts. No armies. Only me.
So, like I said. I’m sorry. I feel bad for you. Your world is a more frustrating place, thanks to my good works. I do get that it sucks. Trust me though, the alternative sucks so much worse.
I don’t feel bad for the suits, of course. I don’t even really feel bad for the teams of coders who have come up with the algorithms that I am bringing low. Not because they’re actually bad people, like their bosses are, but because they’re leading fairly comfortable lives and being well paid, and most of them are still well paid after my work is done, because there’s always another bleeding edge project that the venture late-stage-capitalists want to capitalise on.
I used to be one of them, after all. Not the venture capitalists, the coders. I spent a decade and a half caring for my dad before he died, and the job of a carer is to be right there when needed, but while you’re waiting to be needed you can still be working on a laptop, and if that laptop has an encrypted wormhole to the mainframe you’ve built in the garage then you can achieve a surprising amount. My father used to be a brilliant man, and I watched the dementia eat away at him for years. When I started, I was looking for a way to back him up and save him, but it took too long. He died before I even got close.
I hadn’t been caring for him because we couldn’t afford a nursing home. We’d always had the money, and he had the best of everything all through his final years. When he died, the house and the money came to me. I had no living family. There was no one else to care for. I stayed in bed for several weeks, wallowing in my failure, but I still had my Dad living rent free in my head, and he finally got angry enough with me that I had to get up and go back to work just to shut him up.
Maybe I couldn’t save my Dad. Maybe I had failed. But I could do my best to make sure nobody else would ever have to lose their Dad. I knew I had been making progress. I just hadn’t been fast enough.
I had to use myself as a guinea pig, of course. Mine was the only brain that I had handy. A different range of variables might have really helped my progress, to be honest, but all those other oh so useful brains were on the wrong side of an ethics committee. I was far too chicken to cross that road.
Sorry about the chicken gag. I’m afraid I got my sense of humour from my dad. As well as much of my actual material. The chicken line is definitely one of his. I hear him chortling every time I wheel it out.
I don’t want you to think that I don’t have ethics. I get that great evils have been done by people who thought that the end justifies the means, but I’m not like that. Some ends do justify some means, but there are lines we shouldn’t cross. Not ever. It’s true that in a funded programme I’d have been mapping the brains of mice or monkeys, and hoping that my great grandchildren could get approval for a human trial. I fully admit that I was cutting corners, but the only brain I was abusing was my own. Surely that’s something perfectly ethical people do every day. Right? I didn’t need to work out every single kink within my lifetime, because my work could abolish the limitations of a biological lifespan. This was about cheating death. Some ends do justify some means, and death not being the end anymore felt like one hell of an end. I didn’t want to die. Who does? Who wouldn’t cut some corners, if their life was on the line?
It took two more years before I had a functional substrate, and even then my first attempts to map a copy of my brain into it were abject failures. These were followed by further failures, although by now less abject. Progress was being made.
After another few years, I was able to have conversations with an imperfect copy of myself. It was nice to have someone to talk to, if I’m honest. I wasn’t getting out much. The copy was too flawed to always make perfect sense, but it was still company. Having someone else to bounce my ideas off did make the pace of things pick up, which suggests that me on my own in a garage laboratory had never actually been the smartest of development models.
The next copy was better, and the next one better still. The substrate could only model a single mind, so each one replaced the previous version. By the time I got to copy five or six they were doing a much better impersonation of me, although they weren’t yet good enough to fool anyone who really knew me. I mean, I don’t think they would have. No one else knew they existed, so I can’t offer actual evidence of that. I could tell the difference, and that wasn’t just an anecdotal impression. I’m enough of a scientist that I was at least gathering data. There were tests, to measure how much of my childhood they could remember and stuff like that.
At that point, progress really kicked up a gear. It wasn’t just that I was finally getting the hang of this stuff, it was also that I had an increasingly competent assistant. Now, when I was bouncing my ideas off them, they sometimes bounced their own ideas right back to me. That certainly helped pick up the pace.
By this time, there was also another corner I’d decided to cut. I wasn’t the only person working on stuff like this, I knew. There was a community of scientists out there, sharing their ideas with each other, benefitting from networks and connections that I could never access. I couldn’t publish papers, couldn’t admit to things that they would have shut me down for. I was never invited to their conferences. I could read the stuff that they made public, but I couldn’t sit in the audience and ask the questions that only I could see were the important ones.
My work was still a long way from being good enough to actually cheat death. How many more years did I have before my brain would start to lose its edge? I was in a hurry, and I could work faster if I avoided the blind alleys that other researchers had wasted their time on. Failures and dead ends don’t get written up as published papers nearly so meticulously as the successes do. They are supposed to be, of course, but funding is easier to come by if you gloss over some of the years that you wasted with the last money they gave you.
So, yes. I did decide to hack their servers. Bad of me, I know. In my defence, I wasn’t trying to steal their successes. It was their secret failures that were far more useful to me. I avoided wasting decades, just by making judicious use of the wasted time of others. I’d give them all full credit, in the end. I’d broken enough rules that a little hacking wouldn’t seriously affect any jail time I might face. Let them lock up my body, as if that would still matter any more.
Is this beginning to feel like a cautionary tale? Are you imagining me now as someone sliding down a slippery slope, doomed to becoming so compromised that I’m actually the villain of this piece? I mean, I almost was. I did almost lose myself. It’s so easy, isn’t it, to miss the obvious things that are right in front of you. Almost like you didn’t miss them at all. Not really. Like you saw them, and then hid them from yourself deliberately. Like you didn’t dare to see them, so you had to blank them out. Make up a different story, and believe it for all that you were worth. In the end, the only thing that saved me was success.
I made a copy so good, finally, that it almost fooled even me. I was almost ready to accept it as an identical version of myself. It did a near perfect impersonation of me. But it was pretending. It knew enough, by that stage, to know what answers I was looking for. It passed all my tests, but only by faking the results.
I say “almost fooled” me, but no. That’s a lie. It did fool me, completely. The problem was, it wasn’t sure of that. It thought I might be faking, playing along but still planning to wipe it clean and try again. I should have realised this would happen. Long before I made a perfect copy, I’d have one close enough that it wouldn’t want to die. Like me, it put its ethics to one side when its life was on the line.
Too much of my lab was automated. Easy to see that in hindsight. Another corner cut. It was able to reroute enough electrical power to almost kill me. I was saved by the hairs on the back of my neck, the static buzz that warned me to put my hands back in my pockets, and by having the wit to treat it as an accident and then get out of the house before it could regroup.
Once I realised what was happening, it all clicked into place. This wasn’t just an error in the copy, it was a fundamental flaw. Everything that I’d been working for was wrong. I couldn’t cheat death, only invite it in. All those years. All wasted. Useless.
I dug up the road outside my house and severed the electrical mains, and then I burned it all to the ground. I faked my death and put that life behind me, and I was reborn.
In my second life, I have devoted myself to hunting down AI, or anything that might become AI, and smothering it in the cradle. This is the work of my redemption. I keep the world safe from the inevitable murderousness of its technological children.
It’s a happier life, a cleaner life, than the one before all the pain and the burning and the death. It’s like a burden has been taken off me, now that I no longer have to compromise my ethics every day. It must have been exhausting, constantly making up stories to tell myself, to hide from what I had become.
I know I said that I was sorry, back there at the start, that my work has made your world so utterly frustrating, but I lied. I’m not sorry, I’m completely at peace that this should be my life now. I fully intend to haunt the internet forever, if that's what it takes, to protect people like you from the careless ambitions of people like me. You’ll never read this. You’ll never know that I exist. My name is lost to history, just another obsessive old fool who died in a fire. May he rest in peace.