Drifting from fandom to fandom. Currently obsessed with asian BL. This blog contains 18+ stuff consider this your tag. On AO3 as iffervescent and ko-fi is ko-fi.com/iffervescent
Honestly I’m pretty tired of supporting nostalgebraist-autoresponder. Going to wind down the project some time before the end of this year.
Posting this mainly to get the idea out there, I guess.
This project has taken an immense amount of effort from me over the years, and still does, even when it’s just in maintenance mode.
Today some mysterious system update (or something) made the model no longer fit on the GPU I normally use for it, despite all the same code and settings on my end.
This exact kind of thing happened once before this year, and I eventually figured it out, but I haven’t figured this one out yet. This problem consumed several hours of what was meant to be a relaxing Sunday. Based on past experience, getting to the bottom of the issue would take many more hours.
My options in the short term are to
A. spend (even) more money per unit time, by renting a more powerful GPU to do the same damn thing I know the less powerful one can do (it was doing it this morning!), or
B. silently reduce the context window length by a large amount (and thus the “smartness” of the output, to some degree) to allow the model to fit on the old GPU.
Things like this happen all the time, behind the scenes.
I don’t want to be doing this for another year, much less several years. I don’t want to be doing it at all.
—-
In 2019 and 2020, it was fun to make a GPT-2 autoresponder bot.
Hardly anyone else was doing anything like it. I wasn’t the most qualified person in the world to do it, and I didn’t do the best possible job, but who cares? I learned a lot, and the really competent tech bros of 2019 were off doing something else.
And it was fun to watch the bot “pretend to be me” while interacting (mostly) with my actual group of tumblr mutuals.
In 2023, everyone and their grandmother is making some kind of “gen AI” app. They are helped along by a dizzying array of tools, cranked out by hyper-competent tech bros with apparently infinite reserves of free time.
There are so many of these tools and demos. Every week it seems like there are a hundred more; it feels like every day I wake up and am expected to be familiar with a hundred more vaguely nostalgebraist-autoresponder-shaped things.
And every one of them is vastly better-engineered than my own hacky efforts. They build on each other, and reap the accelerating returns.
I’ve tended to do everything first, ahead of the curve, in my own way. This is what I like doing. Going out into unexplored wilderness, not really knowing what I’m doing, without any maps.
Later, hundreds of others with go to the same place. They’ll make maps, and share them. They’ll go there again and again, learning to make the expeditions systematically. They’ll make an optimized industrial process of it. Meanwhile, I’ll be locked in to my own cottage-industry mode of production.
Being the first to do something means you end up eventually being the worst.
—-
I had a GPT chatbot in 2019, before GPT-3 existed. I don’t think Huggingface Transformers existed, either. I used the primitive tools that were available at the time, and built on them in my own way. These days, it is almost trivial to do the things I did, much better, with standardized tools.
I had a denoising diffusion image generator in 2021, before DALLE-2 or Stable Diffusion or Huggingface Diffusers. I used the primitive tools that were available at the time, and built on them in my own way. These days, it is almost trivial to do the things I did, much better, with standardized tools.
Earlier this year, I was (probably) one the first people to finetune LLaMA. I manually strapped LoRA and 8-bit quantization onto the original codebase, figuring out everything the hard way. It was fun.
Just a few months later, and your grandmother is probably running LLaMA on her toaster as we speak. My homegrown methods look hopelessly antiquated. I think everyone’s doing 4-bit quantization now?
(Are they? I can’t keep track anymore – the hyper-competent tech bros are too damn fast. A few months from now the thing will be probably be quantized to -1 bits, somehow. It’ll be running in your phone’s browser. And it’ll be using RLHF, except no, it’ll be using some successor to RLHF that everyone’s hyping up at the time…)
“You have a GPT chatbot?” someone will ask me. “I assume you’re using AutoLangGPTLayerPrompt?”
No, no, I’m not. I’m trying to debug obscure CUDA issues on a Sunday so my bot can carry on talking to a thousand strangers, every one of whom is asking it something like “PENIS PENIS PENIS.”
Only I am capable of unplugging the blockage and giving the “PENIS PENIS PENIS” askers the responses they crave. (“Which is … what, exactly?”, one might justly wonder.) No one else would fully understand the nature of the bug. It is special to my own bizarre, antiquated, homegrown system.
I must have one of the longest-running GPT chatbots in existence, by now. Possibly the longest-running one?
I like doing new things. I like hacking through uncharted wilderness. The world of GPT chatbots has long since ceased to provide this kind of value to me.
I want to cede this ground to the LLaMA techbros and the prompt engineers. It is not my wilderness anymore.
I miss wilderness. Maybe I will find a new patch of it, in some new place, that no one cares about yet.
—-
Even in 2023, there isn’t really anything else out there quite like Frank. But there could be.
If you want to develop some sort of Frank-like thing, there has never been a better time than now. Everyone and their grandmother is doing it.
“But – but how, exactly?”
Don’t ask me. I don’t know. This isn’t my area anymore.
There has never been a better time to make a GPT chatbot – for everyone except me, that is.
Ask the techbros, the prompt engineers, the grandmas running OpenChatGPT on their ironing boards. They are doing what I did, faster and easier and better, in their sleep. Ask them.
What is or isn’t a slur can be highly contextual, y'all.
“Jonny Sims bummed a fag off my ma” doesn’t contain a slur, but “What are you, some kind of fag?” does.
“Queer studies”, “the queer community” and “I’m queer”? Not a slur. Some bigot calling you a “dirty queer”? Slur.
“Be gay, do crimes” and “He’s gay” ≠ slur, but “Ew, that’s so gay” = slur.
In conclusion, stop buying into this fucking “q slur” bullshit. Queer people talking about the queer community aren’t using it as a slur any more than a gay man calling himself gay is using that term as a slur.
Looks like its time for derogatory pepperoni again
by the way on this the first day of dracula season let me just say that if you are wondering whether you, yes you personally, should sign up for dracula daily this year to see what all the fuss is about, the answer is unequivocally Yes, Do It. dracula is one of the weirdest books i have ever read (if you like i was are only familiar with it through cultural osmosis you are in for basically unrelenting surprise when you dive into the actual text), a horror novel about train schedules, an action movie about archival diligence. it’s an extremely victorian novel that i really do think speaks to our time both in spite and because of the extent to which it’s a perfect distillation of what fears and values the british empire was haunted by in the twilight it didn’t yet see coming. it’s funny by accident but also on purpose - like, really, really funny - and scary and gross and horny and strange and romantic by accident and also on purpose and if i had to choose one word to capture its emotional mood i would say sweet. discovering it in the real-time serialized format offered by dracula daily was honestly a highlight of my year and one of the most fun and rewarding reading experiences i’ve ever had, and its mix of silliness and earnestness i really think makes it a weirdly well suited novel for pondering on this particular website. it’s a love story baby just say yes!
I was on a plane this weekend, and I was chatting with the woman sitting next to me about an upcoming writer’s strike. “Do you really think you’re mistreated?” she asked me.
That’s not the issue at stake here. Let me tell you a little something about “minirooms.”
Minirooms are a way of television writing that is becoming more common. Basically, the studio will hire a small group of writers, 3-6 or so, and employ them for just a few weeks. In those few weeks (six weeks seem to be common), they have to hurriedly figure out as much about the show as they can – characters, plots, outlines for episodes. Then at the end of the six weeks, all the writers are fired except for the showrunner, who has to write the entire series themselves based on the outlines.
This is not a widespread practice, but it has become more common over the past couple of years. Studios like it because instead of paying for a full room for the full length of the show, they just pay a handful of writers for a fraction of the show. It’s not a huge problem now, but the WGA only gets the chance to make rules every three years – if we let this go for another three years and it becomes the norm? That would be DEVASTATING for the tv writing profession.
Do I feel like I’m mistreated? No. I LOVE my job! But in a world of minirooms, there is no place for someone like me – a mid-level writer who makes a decent living working on someone else’s show (I’d like to be a showrunner someday, but for now I feel like I still have a lot to learn, and my husband and I are trying to start a family so I like not being support rather than the leader for now). In a miniroom, there are only two levels – the handful of glorified idea people who are already scrambling to find their next show because you can’t make a decent living off of one six-week job (and since there are fewer people per room, there are fewer jobs overall, even at the six-week amount), and the overworked, stressed as fuck showrunner who is going to have to write the entire thing themselves. Besides being bad for me making a living, I also just think it’s plain bad for television as an art form – what I like about TV is how adaptable it is, how a whole group of people come together to tell a story better than what any of them could do on their own. Plus the showrunner can’t do their best work under all of that pressure, episode after episode, back to back. Minirooms just…fucking suck.
The WGA is proposing two things to fix this – a rule that writers have to be employed for the entire show, and a rule tying the number of writers in the room to the number of episodes you have per season. I don’t think it’s unreasonable. It’s the way shows have run since the advent of television. It’s only in the last couple of years that this has become a new thing. It’s exploitative. It squeezes out everyone except showrunners and people who have the financial means to work only a few months a year. It makes television worse. And that is the issue in this strike that means everything to me, and that is why I voted yes on the strike authorization vote.