In a functioning company, this is how it goes, from the perspective of a content moderator:
1. Oh no, the review queue has loaded a pretty sus looking video. You buffer to the midpoint and find that it's CSAM. A 15 minute timer has started in your head.
Forgot to add that my lawyers suggested getting my name copyrighted because the studio would try to own it because of the show.
They tried doing it w/Reba McIntyre but she owned hers. So I paid money I didn’t have to do it.
A WEEK LATER…the studio tried! I beat them by a week!
I had a sitcom I created about my life.
I was not the highest paid actor.
I was not the highest paid writer.
The show was named after me.
People ask why I haven’t done created another show.
Sometimes I get close but then remember how I was treated and I stick to stand-up.
Browsers use these surveys to prioritize implementations, so take it, and take it seriously! The time you spend filling this out will be an investment (plus, it’s fun!).
Two weeks left to take the 2023 State of CSS survey!
Every voice counts, and it's also a great chance to learn about new and upcoming CSS features all at once!
survey.devographics.com/en-U…
OpenAI and Microsoft are being sued in California for $3 billion over alleged #privacy violations. Here's a short thread on why this lawsuit is important and worth following 🧵👇 #AIEthics
Oh boy, I'm on a philosophical treatment tonight about AI. Mt friend asked me if I thought chatWTF-X was sentient or ever would be. The brought up "cogito ergo sum" as their reason for believing that it is or soon will be.
Descartes didn't argue that thinking proved sentience. He was contemplating his own existence. He wondered if he truly existed or was a different of someone else's imagination. He concluded that non-existing people don't wonder if they exist.
This, as long as he had doubts, he must exist. His declaration "cogito ergo sum" shouldn't be read as, "I can think, therefore I am sentient," but as, "while I doubt my existence, I must exist."
Nothing to do with sentience. Or defining thinking.
Another crappy argument in favor of LLMs: they're scraping the internet for data just like search engines do!
No, search engines point to where the relevant data is, driving traffic (and revenue) to sites. LLMs steal the data and don't give credit or references.
I'm going to relay a conversation I had irl about AI and ethics. My friend brushed aside the problem that LLMs have harvested the vast majority of their data from non-consenting or exploited people because, "the damage is done but the result is here. Why shouldn't we use it??" 1/
You're either going to unethically build it, causing me harm, or ethically build it. If you can ethically build it, why use the unethical one to begin with?
Questions like these surround a lot of unethical studies in medicine and psychology. 4/
If you can't corroborate the data, you really shouldn't be using it.
There are other problems with LLMs but the argument that the harm is only in the past is wrong. And pointless. It's not like past harm didn't happen or repercussions aren't still echoing from it.
There are no words for how painful it is to see MDN chase this trend. I don't know if they're doing some kind of grift or got sucked into the shiny newness but the end result is the same: they're eroding years of trust and harming the people who trust them
MDN’s fake-AI asserts styles which visually hide content, when applied to`s::after` & `s::before, hides the contents of the `<s>` instead.
I added that example to this issue:
github.com/mdn/yari/issues/9…
happy solstice, here's a thread on our universe--and how little we know!
For as long as humans have roamed the Earth, we have sought to find our place in the cosmos. From the city-states of ancient Greece to the soaring capstones of the Egyptian pyramids--
Forget CSS, people don't even know HTML.
Here's why I'm saying this: I've looked through demos made for last week's #CodePenChallengecodepen.io/challenges/2023/j… in order to get a better idea of how people want such a control to look and work.
And here are some numbers: 😭
Talked with a coworker about the diff between the change and input events for text inputs and found out that React has been mapping onChange to input events. It's the only framework doing so, from my testing. Why??
Input events are usually what you want but why rename it?
CSS nerds: how useful would it be to be able to use the current value of a length? Like
outline-width: min(0.125rem, current)
padding: calc(2 * current)
For width and height we have separate properties, and for font size we have em units but we get no reference to other lengths