relevant xkcd

We will (still) allow children’s cartoon characters to be mutilated and put into explicit situations and then push them as child friendly, but how dare you use a swear in the first 15 seconds of a video or say the work kill.

To me, it is the loss of meaningful work.

Alot of people have complained “why take arts and coders jobs - make AI take the drudgery filled work first and leave us the art and writing!” The problem is: automation already came for those jobs. In 90% of jobs today, the job CAN be automated with no AI needed. It just costs more to automate it then to pay a minimum wage worker. Than means anyone who works those jobs isn’t ACTUALLY doing those jobs. They are instead saving their employer the difference between their pay and the amount needed to automate it.

Before genAI came, there were a few jobs that couldn’t be automated. Those people thought that they not only have job security, but they were the only people actually producing things worth value. They were the ones that weren’t just saving a boss a buck. Then genAI came. Why write a book, code a program, or paint a painting if some program can do the same? Oh, it is better? More authentic? It is surprising how much of the population doesn’t care. And AI is getting better - poisoned training and loss of their users critical thinking skills not withstanding.

Soon, the only thing proud a worker can be about their work is how much they saved their employers money; and for most people that isn’t meaning enough. Somethings got to change.

Depends on who programed the AI - and no, it is not Kyoto

Do they still sell those novelty toy mics that broadcast and speakers pick it up?

I’m not telling you what to do with that, but there are many options.

Thirty is cordless and light, because fuck 50 pound vacuums that you have to lug around because ‘it gets better suction’

Forty is a robotic vacuum that doubles as an alarm clock when it wanders into your room at 7 am, eats a sock, and yells at you to unclog it.

Probably there is a lower ceiling in text generation to things that are considered “progress” by everyone. People can verbalize exactly what is wrong with pictures and video, but for text, if it is coherent and gets the point across what is progress beyond passing the Turing test?

Well, maybe sooner than 2030… https://spacenews.com/musk-calls-for-deorbiting-iss-as-soon-as-possible/

Most people aren’t murders

I do that with my email. Email has a search function - if I don’t know a key word in the email or approximately when it came in, how would I know that I found it if I ever found it in some other way?

Let’s change our universally recognized trademarkible name and change it to a non-trademarkible blowdriver setting! Wait, no one recognizes our name anymore? Undo!

Shouldn’t block people. You can make good money by arguing

deleted by creator

Hushpuppy underground!

There was a fad of running things through google translate and back several times until you get something funny.

Has that fad returned with a ChatGPT expand -> contract loops?

Tried to get one at 42 - they required a cola-guard screening first. Requested a cola-guard screening and insurance didn’t cover it until 45. Tried to manually buy one - couple thousand dollars. So basically you can’t start getting screened until 45.

Wonder how much of Windows 10 was written by Stack Exchange?

Wikipedia had bots writing US census gathering-place articles in 2002, 20 years before LLMs were a thing. They’ve got decades of regulations in place, so I am not scared that the quality is going to drop.

Dude, it’s been 2 centuries, take a hint

That is exactly why I said it

If you open up your camera app and spin around and take a picture, 99% of the picture will be garbage.

If you boot up a AI art program and type in a random prompt, 99% of that will be garbage.

Photographer have specialize lenses and choices of FOV that affects how the pictures look. Ai artists have specialized weight and loras that affect how the picture will look.

Photographer don’t just take pictures at random. They set and frame the scenes - doing prep work and framing. AI artist can use base pictures instead of random noise to bias the outcome (image to image).

With live subjects, photographer can either give no guidance, or direct the subjects (think “look at the camera and say cheese”, only more nuanced). With AI art, there is a whole subfield of prompt engineering l which is akin to this.

After a photographer take pictures, they do minor touch ups and photoshoping to clean up parts that didn’t come out right. So too with AI artists.

And with both, you can get 100s if not 1000s of pictures of a subject. The photographer and the AI artist true test is being able to pick from those thousands the one or two good shots.

Yes there is a bunch of legal and copyright problems with AI art. When the camera was first invented, people argued that you couldn’t take pictures of crowds without getting everyone’s concent, nor could you take picture of other people’s property with out breaking the law. That the legal realities around photography weren’t settled didn’t mean those taking picture back then weren’t artists, and it doesn’t mean that people doing AI art today aren’t artists. AI generators are like camera in that you get out better results depending on how much work you put it.