only mike can judge me

Minty | 29 | they | ace | @bruhgender's garbage SV sideblog | nowadays mostly just reblogging anything that reminds me of the show

[snowberry-pie]

this has been discussed before but reducing female characters to the girlboss braincell holder in the name of combating misogyny in fandom is ironically also a form of misogyny

[snowberry-pie]

people love love loooove to go on and on and on about whiteguy #135336843’s internal torment and the nuances of his writing and then when it comes time to talk about women it’s all um actually she can do no wrong and holds the braincell and babysits the boys like a mother hashtag feminism. i hope you fucking blow up

[st-just]

This is just very funny to me.

[kiloueka]

Not even his robot child likes him

[mollyjames]

I feel like the big push for AI is starting to flag. Even my relatively tech obsessed dad is kinda over it. What do you even use it for? Because you sure as hell dont want to use it for fact checking.

There's an advertisement featuring a woman surreptitiously asking her phone to provide her with discussion topics for her book club. And like... what. Is this the use case for commercial AI? This the best you could come up with? Lying to your friends about Moby Dick?

[lesbiananimeism]

One of the big pushes tech companies are making for AI is entirely in the tool of convenience. Take Gemini for example, one of Google's really big pitches for it is in features like Help Me Read and Help Me Write, which are like the lowest tier use case for deep learning models but are also the two AI features that the average consumer will actually care about. Sure they advertise the GenAI stuff Gemini Advanced is able to do, but they've woken up to the idea that the average consumer does not care about GenAI and non-AI Bros fundamentally loathe GenAI.

Every company with a language model got sucked into the venture capital pitfall of AI and now have to market the one set of features the general person actually cares about.

[rageagainstthechristine]

I work in advertising and the culture shift surrounding AI even from January until now (end of March) has been drastic. At the beginning of the year, the company I work for was using AI to design most of their assets. Clients started coming back and requesting that we no longer use AI generated images or videos for copyright liability reasons. Basically, there's no way to tell whose art or photography was scalped to make an image, so as companies who are trying to make a profit using potentially stolen images, it puts them in a gray area, legally.

Also, companies do look at their comment sections. Anti-AI commenters on social media ("this is not a real image" "I don't trust companies who use AI" etc) are seen by higher ups of a company. Basically, keep bullying brands who use AI, it's working. Now my company uses almost no AI for deliverables, which is a huge win.

[parkchanwoohoo]

[ineffable-writer]

Asking AI for information is like asking your drunk uncle for information. Usually wrong, definitely untrustworthy, and a little bit racist.

[grimeclown]

Its like. It's it's really something that the richest man in the world desires internet clout more than anything. Like he wants specifically the kind of clout you get on reddit or 4chan by being a poster like he wants to be a high tier Video Gamer he wants the specific kind of respect you get from being a competitive smash player who posts really good online. But he's just not good. At any of these things . And it's really something. That he has more money than god. But he can't be happy because he can't get a genuine laugh out of the twitch chat.

[cognitohazardous]

im surprised no one is talking about how elon musk paid people to make high level hardcore characters for him so he could claim it was all his work on livestream only to be immediately exposed as he couldn't even play the game right

[cognitohazardous]

like holy shit

[bigandretired-deactivated202501]

[oneheadtoanother]

[closet-keys]

[algorithmist]

propaganda for joining the war on drugs on the side of drugs

[probablyasocialecologist]

AI worse than humans in every way at summarising information, government trial finds

A test of AI for Australia's corporate regulator found that the technology might actually make more work for people, not less.

Artificial intelligence is worse than humans in every way at summarising documents and might actually create additional work for people, a government trial of the technology has found. Amazon conducted the test earlier this year for Australia’s corporate regulator the Securities and Investments Commission (ASIC) using submissions made to an inquiry. The outcome of the trial was revealed in an answer to a questions on notice at the Senate select committee on adopting artificial intelligence. The test involved testing generative AI models before selecting one to ingest five submissions from a parliamentary inquiry into audit and consultancy firms. The most promising model, Meta’s open source model Llama2-70B, was prompted to summarise the submissions with a focus on ASIC mentions, recommendations, references to more regulation, and to include the page references and context. Ten ASIC staff, of varying levels of seniority, were also given the same task with similar prompts. Then, a group of reviewers blindly assessed the summaries produced by both humans and AI for coherency, length, ASIC references, regulation references and for identifying recommendations. They were unaware that this exercise involved AI at all. These reviewers overwhelmingly found that the human summaries beat out their AI competitors on every criteria and on every submission, scoring an 81% on an internal rubric compared with the machine’s 47%.  Human summaries ran up the score by significantly outperforming on identifying references to ASIC documents in the long document, a type of task that the report notes is a “notoriously hard task” for this type of AI. But humans still beat the technology across the board. Reviewers told the report’s authors that AI summaries often missed emphasis, nuance and context; included incorrect information or missed relevant information; and sometimes focused on auxiliary points or introduced irrelevant information. Three of the five reviewers said they guessed that they were reviewing AI content. The reviewers’ overall feedback was that they felt AI summaries may be counterproductive and create further work because of the need to fact-check and refer to original submissions which communicated the message better and more concisely. 

3 September 2024

[spiribia]

we do need to revisit the wording of "you can't have your cake and eat it too" because i don't think it clearly enough conveys that it's more that you can't simultaneously retain a cake and also get to consume it (which would render you cakeless). for years i was like But why not....it's my cake....?

[ferrettoy]

this fucking problem is how they caught the unabomber

[m0thmancore]

hey you should uh. elaborate. for my own personal satisfaction

[ferrettoy]

the unabomber was pedantic about idiomatic phrases like "have your cake and eat it too" and rephrased it to "eat your cake and have it too" (which to be very fair makes sense). fast forward to when he starts writing manifestos. he uses the phrase word for word in his pedantic style and his brother (who has been keeping his eyes on the unabomber shit for obvious reasons) notices the phrase and is like "oh fuck that's my fucking brother no one else fucking says that" and calls in an FBI tip

[tainbocuailnge]

what embarrassing porn did elon musk get caught liking on twitter dot cum after renaming it to a porn site

[cerastes]

If you actually wanted to know

[unprettyextra]

What the heck is "&udm=14"? A new search engine strips AI junk from Google results

It's the work of Ernie Smith, who describes &udm=14 as the "disenshittification Konami code" for Google.

[unpretty]

here’s a direct link to the site:

A quick way to get an AI-free search without any extra work.

here’s his explanation of how it works and ways you can change your browser settings to make the default google search be text only:

How I Made Google’s “Web” View My Default Search

Forget AI. Google just created a version of its search engine free of the extra junk it has added over the past decade-plus. You just need o

[the-real-zhora-salome]

Yikes

[brucebocchi]

elon musk admitted under threat of perjury that this is his burner account

[youarewhatyougeek]

The more I learn about musk the more I confused I get - how does he live like this??

[lizardsfromspace]

ELON MUSK'S GRAND PLAN FOR TWITTER

Buy it for billions of dollars

Be so surrounded by crypto and NFT shills and yes men that you start to see Twitter's ordinary users as the real bots

Scrap the site's pre-existing bot programs bc you were mad a report showed that the site didn't have as much of a bot problem as you thought

Make impersonation trivial, and let everyone who pays you $5 have all their tweets boosted no matter what

Make the site a barely functional husk run by a skeleton crew and reduce moderation in the name of free speech

Have your site flooded with bots, impersonators, and spam

???

Make them pay you $1 as a effort to fix spam, even though most of them already paid $5 for the checkmark that incentivized them to spam

Genius. The real Tony Stark. A pioneer on par with the Wright brothers and Zefram Cochrane. Where do I sign up for my One-Way Ticket to Mars NFT

[lizardsfromspace]

I love how he's obsessed with fixing bots on Twitter, but not just fixing it, he has to fix it.

A normal person buying a company would react to a report saying the site's bot problem isn't as bad as expected with "oh, great, that frees up resources we don't have to allocate there". He reacted by firing the staff responsible in anger, bc if there weren't millions of bots for him to ban, spending all that money would be a waste. This is the management style of a guy who demands a rescue team ditch their equipment to use his sub midway through, and who said he could fix world hunger then rejected an itemized list of what he could spend money on to end some world hunger bc it wouldn't fix all of it

[lizardsfromspace]

When Elon bought Twitter, he was given a report saying the bot problem wasn't as bad as he thought. He fired the people responsible and started declaring that he would solve the bot problem

Today, the majority of traffic on Twitter is now bots

It all makes sense, though, when you realize Elon is not only really stupid and bad at everything, but also he didn't even realize the bots on Twitter were bots, bc of course someone who surrounds himself with crypto shills was going to end up seeing the site's normal users as the suspect ones

He literally ran the site into the ground due to an obsessive need to "fix the bot problem", and he ended up creating the worst bot problem in the history of social media, over 73% worse than the next-worst site. And that's before you consider the possibility that big brain genius Elon Musk is trying to fix Twitter's ad revenue problem by making bots that engage with ads. Incredible

[lizardsfromspace]

Fave part of this, from the story itself, is that the cybersecurity firm monitoring this almost refused to publish the data, bc they didn't believe it could really be that bad

Apparently, before he did all this, the number of bots last year at the same time was...2.81%. That was the number Elon was desperate to fix, and he turned 2% into 75%

[forgottenflickr]

2007