I expect Google will one day try to sell us a similar service built from our Gmail accounts.
Will first step be an upgrade to their autocomplete perhaps?
A simulation of me: fine-tuning an LLM on 240k text messages
91–100 of 145 posts
Re: A simulation of me: fine-tuning an LLM on 240k text messages
#92Earlier quoted context omitted.
It has been tried. Joshua Barneau tried it with GPT-3, training it on his deceased fiancees text- and facebook messages. It’s a moving read: https://www.sfchronicle.com/projects/2021/jessica-simulation... It ‘worked’ for him for a while.
Shocking a biochemical reaction ceases when external stimuli is removed. Technology is like acupuncture and astrology of the modern west.
Unless there is an acupuncture powered smartphone and an astrology powered orbital telecommunications network I am unfamiliar with, the only similarity between them and technology is in the minds of the users, not the stuff itself.
Re: A simulation of me: fine-tuning an LLM on 240k text messages
#93Re: A simulation of me: fine-tuning an LLM on 240k text messages
#94I’m far from the first to think of this. Several people — perhaps inspired by creepy Black Mirror episodes — have tried to fine-tune an LLM on their SMS or WhatsApp history in an effort to create a simulation of themselves. It's a much older concept than Black Mirror. Ever since Markov chain IRC bots got popularized in the late 90s and early 2000s, people have been trying to train their virtual doppelgängers. I'm sur…
There's this recent cool SF novel by the Polish (and undertranslated!!!) SF author Jacek Dukaj called 'The Old Axolotl'. In it, people used gaming systems to copy their 'beings' into machines before a planetwide wipeout event of biologic life. The machines, with the human simulacra within them, survive for hundreds of years. What's interesting about the books is that the machines feel human and have all the human emo…
It's a great book, and totally agree the author is under translated!
Re: A simulation of me: fine-tuning an LLM on 240k text messages
#95This post motivated me to do the same thing on my imessage data. I used GPT-3.5 for fine-tuning (and haha GPT-4 to help me build it :D). I guess, I should run fine-tuning jobs to create instances of my correspondence partners? Since, for the first system prompt I tell the model to pretend to be me, but on other fine-tuning attempts I can reverse the input and output, and then tell the model to "pretend" to be my coun…
This is a lot more risky than doing it on a self-hosted model surely? Do you really want to upload your entire message history to a third party (i.e. not Apple)?
Re: A simulation of me: fine-tuning an LLM on 240k text messages
#96Why keep humans around at all? With only a little extra data and training effort it should be possible to create an entire internet of immortal bots living out shiny self-actualising Apple-perfect virtual lives, while the surviving hairless monkeys live out their short, pointless, ugly, imperfect existence slaving away in the lithium mines.
Re: A simulation of me: fine-tuning an LLM on 240k text messages
#97> The LLM would generate conversations that seemed to be surreal parodies of my life. My work colleagues would endlessly reschedule meetings earlier and later; my partner and I would relentlessly debate our lunch options.
Or as I call it, life.
Re: A simulation of me: fine-tuning an LLM on 240k text messages
#98At least in Western magic and mythology, doppelgangers have been harbingers of personal doom for centuries - so beware of your evil twin.
Just saying, so that no one is surprised when they start to doubt themselves...
Re: A simulation of me: fine-tuning an LLM on 240k text messages
#99Earlier quoted context omitted.
Have you ever watched the original Superman movies and Clark talks with the AI hologram of his biological father from his home planet? I believe in the future our descendants will likely find it normal to talk with AI versions of their ancestors, perhaps not us parents right now, but I suspect my children will willingly train and generate AI versions of themselves for future generations to interact with.
> Have you ever watched the original Superman movies and Clark talks with the AI hologram of his biological father from his home planet? You get that this only played out positively in the movie because the writers/director wanted it to play out that way, right?
In the Smallville TV show, the Jor-El AI was shown to have its own agenda, which at times was very much at odds with not-yet-Superman's goals and view points.
Re: A simulation of me: fine-tuning an LLM on 240k text messages
#100Earlier quoted context omitted.
> it might be therapeutic to have a conversation with a deceased loved one. Would it? I'm not a specialist, but it would seem a bit off to me to approach it like that. I'd think best is to accept that your deceased one has gone, and it seems like having a fake one might just prolong the grievance process and possibly also develop into an obsession of trying to replace your deceased one with the simulation, upgrading…
Have you ever watched the original Superman movies and Clark talks with the AI hologram of his biological father from his home planet? I believe in the future our descendants will likely find it normal to talk with AI versions of their ancestors, perhaps not us parents right now, but I suspect my children will willingly train and generate AI versions of themselves for future generations to interact with.
We can speculate that this was not a mainstream thing in Kryptonian society. Then again, their society was on the verge of collapse when the Els committed the heresy of conceiving a "general purpose" child in a society where bloodlines were bred for particular roles in society. Creating a life-like AI resemblance subject to the same restrictions as Jor-El himself seems almost "normal" in comparison.