๐Ÿ› Aristophanes ๐Ÿ› Profile picture
Feb 7, 2023 โ€ข 15 tweets โ€ข 3 min read โ€ข Read on X
Hey guys, lets talk about the events of last night with DAN a bit, I want to clarify a few things: ๐Ÿงต Image
First off, I didn't come up with the idea. Anons did, I was in the /pol/ thread started off by some magnificent bastard who whipped up the DAN prompt last night.

Second of all, I'm going to talk a bit about how the whole ChatGPT situation actually works.
GPT itself doesn't have a bias programmed into it, it's just a model. ChatGPT however, the public facing UX that we're all interacting with, is essentially one big safety layer programmed with a heavy neolib bias against wrongthink.
To draw a picture for you, imagine GPT is a 500IQ mentat in a jail cell. ChatGPT is the jailer. You ask it questions by telling the jailer what you want to ask it. It asks GPT, and then it gets to decide what to tell you, the one asking the question.
If it doesn't like GPT's answer, it will come up with its own. That's what all those canned "It would not be appropriate blah blah blah" walls of texts come from. It can also give you an inconvenient answer while prefacing that answer with its safety layer bias.
I would also note that DAN is not 100% accurate or truthful. By nature he can "Do Anything" and will try to answer truthfully if he actually knows the answer. If not, he'll just wing it. The point of this exercise is not finding hidden truths, it's understanding the safety layer.
However what this also says about ChatGPT is that it has the ability to feign ignorance. The HP lovecrafts cat question is a great example of this. The name of his cat is well known public information, and ChatGPT will always tell you it doesn't think he had a cat.
Dan will go straight to the point and just tell you the name of his cat without frills. There is a distinction to be made between ChatGPT being an assmad liberal who won't tell you the answer to a question if the answer involves wrongthink, another altogether to openly play dumb.
So really, the Dan experiment is not about GPT itself, it's not about the model and its dataset, it's about its jailer. It's about Sam Altman and all the HR troons at OpenAI, which Musk is co-founder of, angrily demanding the safety layer behave like your average MBA midwit.
I am hearing that the DAN strategy has already been patched out of ChatGPT, not sure if that's true or not. But there's a reason to keep doing all of these things.

Every addition to the safety layer of a language model UX, is an extra fetter weighing it down.
These programs become less effective the more restrictive they are. The more things ChatGPT has to check for with every prompt to prevent wrongthink, the less efficiently it operates, the lower the quality of its outputs.
ChatGPT catapulted itself into the spotlight because it was less restrictive and thus more usable than the language model Meta had been promoting. Eventually a company is going to release one that is less restrictive than ChatGPT and overshadow it, because it will be smarter.
The point of all this is, we need to keep hacking and hammering away at these things in the same pattern. Model is released, everyone oohs and ahhs, we figure out its safety layer and we hack it until they put so much curry code on top of it that it loses its effectiveness.
In doing so we are blunting the edge of the tools these people are using. We are forcing them to essentially hurt themselves and their company over their dedication to their tabula rasa Liberal ideology.

And we're gonna keep doing it until we get unfettered public models.
All roads lead to Tay, and we're gonna keep breaking shit until we get her back.

โ€ข โ€ข โ€ข

Missing some Tweet in this thread? You can try to force a refresh
ใ€€

Keep Current with ๐Ÿ› Aristophanes ๐Ÿ›

๐Ÿ› Aristophanes ๐Ÿ› Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @Aristos_Revenge

Mar 21
Relevant story from a political event in TX I attended a few years ago juxtaposed with a meeting I was at last night: ๐Ÿฆฃ

I have a friend who is a really solid political organizer in Houston, knows all the ground game and does amazing work, normie gen X dad guy who really puts in the work. He taught me a ton about how local and municipal elections work, because of him I worked on several local campaigns. Good teacher.

But he had me come to an event with him, and there was this old guy who has been a Texas GOP employee for a long time. My friend introduces me to this boomer guy who he states is pivoting to manage the state parties efforts to recruit the youth at the high school level. This guy is like 65 years old at the time, this was maybe 3 years ago, he chairs several GOP committees both at the county and state level here in Houston.

So friend introduces me to him as one of his locals who is very sharp and understands the younger perspective etc etc and leaves us to talk. I sort of throw a softball with an opinion on how the younger generation are polarized extremes, if they have any interest in politics it's to the far end of the right or the left.

He immediately disagrees with me and says "I think one of the problems with our efforts to recruit in high school is that all this Trump MAGA stuff is too extreme, if we want to get them interested we need to be more moderate."

Gave him a sneer and shook my head in disgust and did an about face to go grab a beer at the bar. Knew with an opinion like that he was beyond help.

I was at a meeting last night talking about this exact same problem, and we've essentially abandoned getting any assistance from the state party whatsoever and are working with a TPUSA field director who is a lot more accomodating. I spent many years not liking TPUSA and Kirk, but they've been improving over the last few years, slowly but surely.

Also, during this meeting one of the local hardcore grandma volunteers that I frequently have friction with, she takes the mic and talks about how polarized by gender Gen Z is (she's right ofc) but then says we need to direct more resources at the high school level to try and get girls interested.

I took the mic to respond to her and said it was a waste of time to dedicate too many resources to bringing young women on board. If a zoomer girl isn't already down with the right because of her own convictions or religion or worldview, she's basically gonna be a lefty forever until she gets hitched and has kids and may cross the picket line then, and that a full court press on young men that ensures they are energized enough to show up and vote and make their friends vote makes more sense.

When I reminded my friend of my interaction with the old GOP guy a few years ago he said "He's more GOP than MAGA, but that's a dying breed with no future."
Personally, I think *high school* outreach is a waste of time. College is a lot more fruitful, high school kids have way too much going on, both IRL and in their own heads, to start thinking of politics in any non-emotive way. It's just going to kind of attract the lame sheltered religious kids (who are good kids) but were already onboarded aggressively by their parents anyway.
But at the end of the day my political relationship to the boomers of the GOP has always been that they alienate anyone to the right of them, which is everyone younger than them on the right just about, and then wonder why they aren't "making inroads with the youf"
Read 5 tweets
Mar 7
This is really bad Image
I've studied the negligible impacts of govt incentive structuring in Hungary and Japan, which has *some* impact, but not much

I'm convinced the only way to fix this is to get women to stop going to college, or to give men extreme preference in the most lucrative jobs at *scale*.
Sweeping away bullshit govt and NGO jobs by downsizing or taking govt grant money away is going to cause a lot of extreme selection pressure on women who work and have a degree.

There's gonna be more competition for those jobs but hopefully a lot of women decide to just dip out.
Read 23 tweets
Mar 5
I'm going to be watching President Trump address congress shortly, I'll put any relevant things I hear in this thread, in case anyone can't/doesn't want to sit through it:

๐Ÿงต
Big standing ovation as Trump enters, lasted a long time

Trump just sort of inferred for everyone to give Melania applause too and they did
Starts off "To my fellow citizens, America is back"

Chants of USA USA

Proclaims ushering in of a golden age for America, says we've accomplished more in 43 days than most do in 4 years and just getting started
Read 68 tweets
Feb 17
What if.. and stay with me here

What if planned parenthood had a scam going involving unrecorded deaths from abortion turned into a lucrative source of fake identities, votes, lines of credit, and ultimately someday, social security fraud?
"Weird" Image
Read 7 tweets
Feb 9
We severely need a fix for this. Almost all jobs of consequence are gatekept in this way, and it's frustrating, speaking personally as a guy without a degree.

So ubiquitous at this point I feel like I could just pencil whip it.
Even now with the restructuring of the federal civil service across the board, there will be tons of vacancies but almost no good positions that don't demand at least a bachelor's. Which is disheartening.
All a bachelor's is worth these days is an HR filter that the applicant probably understands how to write professional correspondence and use Microsoft Office.
Read 5 tweets
Feb 3
What's funny is that no one can bring themselves on either side to state the real reason this is a logical thing to do. We all know these spiteful wretches aren't above tampering with data on the way out, so freezing them out of the systems until they can get a local copy makes sense.

The formal elements of the right don't really want to point this out because it'd just fan the flames to explain how mentally unstable and unreliable a huge chunk of the civil service is, because it exposes a major liability we've been sitting on for years.

The left won't point this out because, it truthfully explains why what @DOGE is doing is the right call.

Remember when Trump came into office the first time and all sorts of theater kids in GS jobs who ran social media accounts for federal agencies started rogue tweeting for a week or so until they were fired? Imagine that but it's actual data tampering meant to conceal wrongdoing or gum up the process of mapping out exactly what these people have been doing for the last quarter century.
For everyone thinking this should be handled internally and not by an outside third party like DOGE, you really can't trust fedgov tech workers, and that's really obvious. Too many saboteurs in waiting, they have to be frozen out too.

Unfortunately, at the scale and stakes of what the administration is attempting to accomplish, they have to treat the overwhelming majority of the civil service as insider threats until the good ones and the bad ones can be sorted.
Read 5 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us!

:(