SkyLander
ellejust called ai "the mediocrity machine" in a meeting and a tech bro is twitching so hard he can't even plug this into the mediocrity machine so it can tell him how to respond
Why Not Zoidberg? 🦑@ElleGray Because that's what heroes do.
The Witch of Crow Briar
Noa@ElleGray calling it a mediocrity machine is a fun way to dunk on tech bros, but it's really a lightning rod for accountability & is therefore scarier since human labor is gonna fill the gaps in performance.
Efi (nap pet) 🦊💤@ElleGray that is such a mediocre reaction to a brilliant descriptor
Lukas@ElleGray I'm offended. I'm a mediocrity machine. AI can't hit that
NO NAMESay this as you walk out, “Google indexed the internet, and ChatGPT plagiarized it.” - LOL that ought to be the straw that makes tech bro go POP!
ScottinSoCal 🇺🇦 🕊 🏳🌈IT at work wasn't happy when I responded to their "AI is now available!" announcement to remind them AI suggested glue as a pizza topping.
MaybeMyMonkeys@ScottSoCal @ElleGray an AI would never pineapple on a pizza
John McChesney-Young@ScottSoCal @ElleGray And Google's AI is still recommending it: https://www.theverge.com/2024/6/11/24176490/mm-delicious-glue
Andreas K@jmccyoung @ScottSoCal @ElleGray
Again as funny this is not one of the dangerous ones, they are clearly dangerous and faulty.
Gemini reporting in a convinced tone slightly wrong payroll tax rates (see attached screenshots, it claims they are current for this year even, and provides a kind of correct, but top-level URL)
Notice the little whopper, the pension part: 10,25% is the actual employee part, according to a more detailed PDF here: https://www.sozialversicherung.at/cdscontent/load?contentid=10008.784719&version=1703166731 )
@jmccyoung @ScottSoCal @ElleGray
Note:
Gemini answered a German question about Austrian payroll taxes with a elegant, convincing answer, that contained seemingly all relevant information (it left out half a dozen of small ones in the <1% range out), with the wrong data (all the percentages are wrong, but most of them are off only by a small error).
And more believability to the big lie, it added a truly valid date when the update “social insurance thresholds, etc” usually happens.
@jmccyoung @ScottSoCal @ElleGray Plus it added a URL to the correct website of the Austrian social insurance, but only to the home page, so you have to find the correct info by clicking yourself to the info.
Perfect. If I wanted to do a presentation to present fake info about Austrian payroll taxes as true info, this is about how I'd go about it.
I guess Gemini AI Premium would add the colourful and extra manipulative graph slides that I might add to it.
@jmccyoung @ScottSoCal @ElleGray
Now let's analyse this in detail:
a) an Austrian tax advisor or payroll accounting specialist will probably spot the issues with the answer, especially if they read it and not only glance over it.
b) non-Austrian experts, or even Austrian business people, might accept it at face value. Sounds plausible, has references that are a little hard to check, let's run with it for the moment.
c) the general public will generally accept it as authoritative.
@jmccyoung @ScottSoCal @ElleGray
And BTW, I'm a well informed element in set B. Just currently evaluating the situation for a indepth discussion with our family's members of set A.
And the post showed up in my timeline while I have the social insurance data open in a tab, and Gemini open in a tab, thus the experiment.
Ned Yeung@yacc143 @jmccyoung @ScottSoCal @ElleGray
This is the difference between searching on a search engine and receiving a top quoted result, including the source and a quote pulled from the source. If I don't get enough context surrounding the quote, I can click on the link and see the entire context. I can easily find locations, dates, etc. to gauge relevance. Nothing made-up, and I can use my own experience to know if the source is legitimate.
This is how humans used search engines effectively.
@yacc143 @ScottSoCal @ElleGray I appreciated the more accurate term proposed here: https://undark.org/2023/04/06/chatgpt-isnt-hallucinating-its-bullshitting/ to replace "hallucination." The whole mechanism of LLMs is focused on plausibility; truth isn't a factor at all.
@jmccyoung @ScottSoCal @ElleGray
As I like to point out, truth is a hard to nail down concept. Yes, the easy questions sound easy, but even these can have surprising twists.
Now you might want to discuss sharks and batteries with some MAGA acolytes. The insights you'll gain might surprise you.
But without somehow a way for a computer to measure truth to guide the training of the network involved, it's unfair from us to expect them to be truthful.
@jmccyoung @ScottSoCal @ElleGray
And technically, it's not even “plausibility”, generally the training is measured against the test data set, which normally is a split of the training set, e.g. how similar the generated text is to the test part of the corpus. So it's “similarity to the training data”. Which presently is assumed to be mostly human written text.
(But you might see how this can turn ugly when the Internet fills up with AI-generated bull shit.)
@jmccyoung @ScottSoCal @ElleGray
Yeah, there is no quick way to get rid of such cool stuff from the outputs of a LLM:
You can try to add to the prefix rules that are added to the prompts that are sent to the LLM. That's hit or miss.
You can try to manipulate the training set of the LLM. Now that's hit or miss, and the compile time for a LLM (it's called training time) makes compiling complex C++ on a 64KB 8-bit box look like a great and speedy idea.
Estarriol, Terrorist Dragon@ElleGray mediocrity, they wish their machines could reach mediocrity, I mean that would be about 9 steps up from their current position.
@Thebratdragon fair point
Coyoty@ElleGray AIs will make @lowqualityfacts unemployed.
Gemma ⭐️🔰🇺🇸 🇵🇭 🎐@ElleGray I said the exact same thing yesterday, after watching the Apple WWDC keynote presentation. The ultimate effect of these "AI" tools will be the reduction of humanity to the least common denominator, an inherently reactionary drive toward conformity to mediocrity that the users of these tools will never even be able to perceive. LLMs and generative adversarial networks are really just elaborate Bayesian probability filters. Garbage in, garbage out.
@ElleGray In fact, I predict that there will very quickly come a point when most people will start thinking creations by expert human creators are "incorrect" (grammar, composition, etc) because they don't conform to expectations generated from exposure to mediocre AI output.
That was actually my very first thought when Apple demonstrated the "Professional" rewrite of an email. Corporatespeak is the new fluency. Imagine the quality level of Microsoft Word's grammar checker, everywhere.
JoeMcconIf we put enough money and gear and effort into it, there's a chance that, in time, it could achieve mediocrity. With luck. But probably not.
Jargoggles@ElleGray
He's going to respond much later with a lame comeback, Costanza-style, except George didn't need to use as much electricity as a major city to do it.
He's going to respond much later with a lame comeback, Costanza-style, except George didn't need to use as much electricity as a major city to do it.
The Lichtenbergian@ElleGray You're doing the lord's work.
@ElleGray AI is the average of everything it learns on, so it will always be average.
Thomas Fuglseth@ElleGray I'm stealing that.
Works well in Norwegian as well, come to think of it. "Middelmådighetsmaskin".
@Uglesett haha well I'm stealing this ⬆️
Ovan
Yendolosch
Cat Lady Kydia Music 🇺🇦“But…but..mediocrity is so cheap and convenient!
Why strive for originality, creativity, brilliance, and greatness when that takes so much time, money, and effort? Plus it really cuts into our sittin’ around, absorbing mediocre content time!”
😩
Gulthaw 
@ElleGray as someone who loves IT and has worked on it for almost 30 years, I know my colleagues see me as a "tech bro".
Nothing further from the truth!! As I usually say "I've worked as a programmer, I know the shit they would like me to put there, I wouldn't touch it with a six feet pole"
Nothing further from the truth!! As I usually say "I've worked as a programmer, I know the shit they would like me to put there, I wouldn't touch it with a six feet pole"
paul
Hudson@ElleGray I’m as excited about AI as I am about crypto.
Eric Borisch@ElleGray @dangillmor I’m pushing for “Regurgitative AI” to better represent what this generation of “Generative” actually does.
Kinene⭐🐻@ElleGray AI: Something about a monkey with a typewriter and infinite time producing a Shakespeare play. Thanks, no thanks.🙈🙉🙊
ChookMother 🇦🇺🦘@ElleGray Oh, good for you. Do keep it up.
This report "Disruptions on the horizon" by the Canadian Government lists "People not knowing what's true or false" as an imminent threat to the nation. Also "AI develops rapidly and its usage becomes pervasive. Society cannot keep up, and people do not widely understand where and how it is being used."
Holds true for the rest of the world.
https://horizons.service.canada.ca/en/2024/disruptions/#Prominentdisruptions
argonaut@ElleGray it is a more precise definition than you probably think.
it also speaks volumes of us from how taken we are by it and how much attention we dedicate to it.
@ElleGray @dangillmor Wonderful description of ai.
nantucketlit@ElleGray "The mediocrity machine", I like that.
Gilgwath@ElleGray You win the internet for today!
Simon Frankau@ElleGray Oh, "the mediocrity machine" is painfully accurate.
I mostly see AI being used now by those who can't reach "mediocre" otherwise, or would happily go for "mediocre, but cheaper" over "good".
It just cuts to the core of the current situation.
Nini
Yaoi Box Balloon@ElleGray This is genius.
@ElleGray I've long said LLMs are basically the unearned confidence of a mediocre white man coded into software.
ArneBab@ElleGray that’s still a positive take. Most of the internet is filled with low-grade bullshit — just not too bad to make money — and this is what AI plagiarizes.
Travis Mooney-Evans@ElleGray Love it. I just keep muttering GIGO.
Mike@ElleGray @dangillmor "mediocrity" is generous
I consider my self a mediocre cook and would never suggest glue on pizza 🤷♂️
Mike Aparicio@ElleGray that was his nickname in college.