invite me and I’ll bring my own alcohol. spread looks delicious!

Another great example (from DeepMind) is AlphaFold. Because there’s relatively little amounts of data on protein structures (only 175k in the PDB), you can’t really build a model that requires millions or billions of structures. Coupled with the fact that getting the structure of a new protein in the lab is really hard, and that most proteins are highly synonymous (you share about 60% of your genes with a banana).

So the researchers generated a bunch of “plausible yet never seen in nature” protein structures (that their model thought were high quality) and used them for training.

Granted, even though AlphaFold has made incredible progress, it still hasn’t been able to show any biological breakthroughs (e.g. 80% accuracy is much better than the 60% accuracy we were at 10 years ago, but still not nearly where we really need to be).

Image models, on the other hand, are quite sophisticated, and many of them can “beat” humans or look “more natural” than an actual photograph. Trying to eek the final 0.01% out of a 99.9% accurate model is when the model collapse happens–the model starts to learn from the “nearly accurate to the human eye but containing unseen flaws” images.

oh nice! I found the same labyrinth after 4-5 taps.

this is so exciting! I imagine in the next few days I might be able to get to the content!!!

I just had a seizure reading trying to read this “article”. what do I even click to escape??

“If you sing at the table you’ll cry before you go to bed.” I thought it was super common until I said it to my kid and my partner thought I was crazy.

Yeah, the problem is how to sanitise effectively. You’ve gotta be able to find a way to automatically strip out “bad” things from your training data (via an “oracle”). But if you already had that oracle, you could just slap it on your final product (e.g. Search) and make all the “bad” things disappear before they hit the user (via some sort of filter).

Fortunately, there’s good news: https://nypost.com/2024/05/18/us-news/rudy-giuliani-served-with-arizona-fake-electors-indictment-during-80th-birthday-bash-in-palm-beach/

I left the first time for Trump… but moved to the UK just in time for Brexit. Should’ve picked Taiwan I guess

“…Americans are planning to flee in droves, and then will realise they don’t have a passport, don’t know any foreign languages, and don’t know how to get around without a car and will ultimately just stay put.”

yeah I don’t understand the “not eating forever” thing. the only question is whether I eat the three bags of chips before or after the 6 servings of drunken noodles arrived

I was with you until the “construction site and under the bridge” bit. It definitely takes a bit of imagination, but I’m not sure not wanting your kids to play on a site which requires the use of hard hats classifies as being “anxious”

someone needs to spend some time on !fuck_cars@lemmy.ml

Just commenting to also get a name in that history book.

“Oh yeah. We knew it was coming. We were just waiting to see which one would finally cause it.”

The struggle here is that you’re talking about money earned after the fact and not including game theory. It would be a tough experiment to conduct, but say you spent $150 million to save $104. What if you didn’t spend that $150M? Would you have an extra $40k in the bank? Or would the $104M in losses actually end up more like $1.2B because, slowly, everyone realised there was no reason to pay a fare?

I don’t know what the case is here, but I imagine some economists have determined that $150M is enough to balance between actually getting people to ride the subway (increasing fare will eventually drive down revenue) and a substantial enough threat to prevent jumpers (no cops in the way means tons more jumpers).