evacideIt's always the most mediocre people who are worried about AI replacing them or taking their jobs or becoming sentient and/or murderous.
Sir Real 🇵🇸@evacide mediocre people need to pay the rent too
LisPi@LeftCoast @evacide That is a valid point. We very much need to fix that hilariously broken renter system of ours.
Access to food, medicine and shelter shouldn't be paywalled.
Greg@evacide Maybe it will take mediocre people’s jobs so maybe they’re right to be worried?
benschwarz@evacide this somewhat overlooks people who don’t understand AI and have real and reasonable fear that they’ll get fucked over by society again, like they always have?
Maybe not your intention, but sure sounded bad when I read it
𝕃𝕦𝕔𝕒𝕤@evacide Not the best hot take. I already know of some folks who have been replaced by ChatGPT 🙁
Jeff Turner ⛵First they came for the mediocre .... you know how it goes
https://en.m.wikipedia.org/wiki/First_they_came_...
regphunt@evacide That's exactly right. I think people who rely really heavily on tech making their output acceptable will be the easiest to replace.
It's the ones who create from the heart, that AI can't just load from the net, that will remain valuable.
Basically it was the blue collars that had to worry about robots, but now some of the white collars do too.
It's the ones who create from the heart, that AI can't just load from the net, that will remain valuable.
Basically it was the blue collars that had to worry about robots, but now some of the white collars do too.
THIS ACCOUNT HAS MOVED@evacide an interesting observation. LLMs are by far the best bullshitters on the planet.
femaven
@noplasticshower @evacide it’s about using Markov chains to fake knowledge about wine http://markallenthornton.com/blog/wine-markov/
Daniel Lakeland@noplasticshower @femaven @evacide
LLMs are literally markov chains. They just have really really complicated state and transition functions.
The way they work is... you give them a sequence of N tokens (that's the state) and they predict the most likely next token (that's the transition function). Then you've got N+1 tokens so lather-rinse-repeat... Once the number of tokens exceeds some value M drop the tokens off the front of the text... So basically it's Markov chains but really big.
@noplasticshower @evacide the summary of the Markov chain for wine comes to a similar conclusion per the 2nd article - Overall, the results of the Markov generation process were mixed - some are nonsensical, self-contradictory, or ungrammatical, but many are quite coherent and convincingly human. Ultimately, perhaps they're best considered as the tasting notes a critic might write after finishing the bottle.
BagheeraAltered 
@evacide AI would be friendlier then the government employees I've dealt with
Mark Wang@evacide - hmmm... are they really the "most mediocre" people, though? where are you ranking people brazenly passing off AI as their own work?
for example, this person: https://arstechnica.com/tech-policy/2023/05/lawyer-cited-6-fake-cases-made-up-by-chatgpt-judge-calls-it-unprecedented/
maybe they weren't mediocre enough to be fearful? or maybe mediocre leaves enough space for "objectively bad at their job"... i guess i could go either way.
bengo@evacide it's always the most white people who imply that jobs are assigned based on quality of a person.
there's no such thing as medicore people
TrashPanda@bengo @evacide I mean, I get where you're coming from, but in a bell curve of who is the best at any given activity, there are always people in the middle. Most of them, really. They are literally mediocre at that activity. You, me, and everyone we know are mediocre at most activities. So there's almost no such thing as people who aren't mediocre, if you want to be accurate and honest. Which should actually generate more sympathy for mediocre people, because... It's all of us.
But not all of us depend on something we're mediocre at for our jobs. The ones that an AI will be able to do better than us are in trouble.
But not all of us depend on something we're mediocre at for our jobs. The ones that an AI will be able to do better than us are in trouble.
Dave Wilburn 
@evacide I'm not worried. I want them to take my job. Then I can finally flee into the wilderness to be left alone.
Why worry? The shitty autocomplete definitely has this covered.
@DaveMWilburn @evacide I'm guessing you are actually very confident AI can't take your job, or you have at least 3 months worth of living expenses in the bank. 😉
That was the stage where I got comfortable with getting fired.
That was the stage where I got comfortable with getting fired.
@TrashPanda @evacide I think it's both? Anyway, I've been doing R&D of ML applied to infosec for a few years now, and I'm pretty sure an LLM isn't taking our infosec jobs.
lynn@evacide This is a super uncharacteristic and disappointing take from someone who's judgement I placed a lot of weight in.
Is automated violence okay when there's a human in the loop? Do 'mediocre' service workers not need to be worried about not being automated away from having a living within their lifetime?
Like... Is there an angle or greater context that I'm missing?
@lynndotpy @evacide please don't call losing your job "automated violence". This is a terrible trend I'm seeing in sacrificing"telling the truth" on the altar of "making things sound super dramatic for attention". Real violence is physical and brutal and causes more harm with less chance of escape or help than any "economic violence". Please don't contribute to people leaking to tune out even more serious harms due to the "boy who cried wolf" problem with calling things violence.
@TrashPanda @evacide I'm talking about the "killing people" violence, in *addition to* people losing their jobs.
I'm talking about predictive policing, Clearview AI, using facial recognition to specifically track Uyghur vs Han people, etc.
I'm also talking about HIL but proprietary bots which almost certainly use trained models to perform well. I'm talking about US drone warfare, LAPDs police robots, Boston Dynamic's dog, etc.
This is "killing people" violence which is already happening.
@TrashPanda @evacide Plus a huge amount of research which is likely dual-use.
ICE already uses utility metrics (water, power usage) to identify places where undocumented people might be living (in LA at least.)
Non-intrusive load monitoring which identifies separate appliance load profiles from one sensor. Useful for identifying who and what is using the most energy.
I worked on NILM, but then realized it would help ICE do their job (violence against undocumented people) more effectively.
Rebeca@evacide people are not worried that AI will be "better" than them, they're worried that corporations prefer bad and cheap over good and (comparatively) expensive.
https://www.washingtonpost.com/technology/2023/06/02/ai-taking-jobs/
Ben Rosengart@evacide It’s really too bad that people not as smart as you want to eat and pay rent. We should understand we are surplus to requirements and walk into the sea.
(Or we can reject this kind of capitalistic morality and value one another as human beings.)
Dynom@evacide isn't that everybody, in at least some speciality they don't excel in?
mupan 📚@evacide The powerful decide who will be replaced. Brains are rarely powerful. Doing a good job rarely is a much relevant criterion for decisions about your job's existence, your duties, and your job role's power. A thumb rule to determine the important factors to be successful and fairly safe on the job and in public roles would be: 40% Self-marketing, 40% reputation, 10% chance, 10% your work and education. So, no ground for classist self-assurance. — I'm not alleging you were classist, I don't know. Your statement could be read classist, think about it.
@TrashPanda @evacide It's true for the average office and production job, so, the majority. Several studies resulted in about these shares, and my own and friends' experience. Thumb rules aren't exact numbers. They show tendencies.
I also found that many believe elbows were part of the demanded job profile. So if self-marketing and explicit progression of my own reputation are embraced soft-skills regarded as part of job qualification, about 90 % own work and education »make the man«.
The concurrency situation is often compared to baboons fighting for higher places in their rock, nearer to the male chief with the biggest red ass, and finally to replace him. You can decide for yourself for teamplay. Then you will most likely stick at about your rank or maybe be pushed over the edge. You can accept the importance of the 80 % as part of your self-defense, while not believing in it, then you might be fairly safe until you are scrapped.
Jack Aponte@evacide My non-techie, non-white, non-cis-men family and friends who are worried about AI are mediocre people? the fuck?
(Yeah I know you responded to a reply here saying you're making fun of cis white men but come on.)
@evacide Look Sarah Connor wasn't a mediocre person, but all her fears were validated about the second part.
Then again she admitted the Terminator was the best dad John ever had so maybe she was kind of prejudiced by a bad early experience, so let's all keep an open mind.
Then again she admitted the Terminator was the best dad John ever had so maybe she was kind of prejudiced by a bad early experience, so let's all keep an open mind.
Andrew@evacide 99% of us are economically mediocre these days. Btw, thanks for that Hyper-Capitalism. Sounded like you meant in relation to skills/competency though.
@evacide luddites, where they all wrong? stop the progression of people kind and machines or ex deuce machina