IrritableOcelot

This gives strong "Lovecraft describing things he doesn't understand as noneuclidian" vibes.

🎶 Saturday night and we in the spot, don't believe me just watch 🎶

Chuck mangione soothes my soul

WELL ACKSHUALLY its a clay tablet, you just press into it with a little stick, then its fired...

Gotta love the low-quality-copper memes

While I agree that publishers charging high open access fees is a bad practice, the ACS journals aren't the kind of bottom-of-the-barrel predatory journals you're describing. ACS nano in particular is a well respected journal for nanochem, with a generally well-respected editorial board, and any suspicions of editorial misconduct of the type you're describing would be a three-alarm fire in the community.

I will also note that this article is labelled "free to read" -- when the authors have paid an (as you said, exhorbitant) publishing fee to have the paper be open access, the label used by ACS journals is "open access". The "free to read" label would be an editorial decision, typically because the article is relevant outside the typical readerbase of the journal, and so it makes sense both from a practical perspective (and more cynically for the journal's PR) to make it available to everyone, not just the community who has institutional access.

Also, the fact that the authors had a little fun with the title doesn't mean its low-effort slop -- this was actually an important critique at the time, because for years people had been adding different modifications to graphene and making a huge deal about how revolutionary their new magic material was.

The point this paper was trying to make is that finding modifications to graphene which make it better for electrocatalysis is not some revolutionary thing, because almost any modification works. It was actually a useful recalibration for expectations, as well as a good laugh.

Edit: typo

Not somebody who knows a lot about this stuff, as I'm a bit of an AI Luddite, but I know just enough to answer this!

"Tokens" are essentially just a unit of work -- instead of interacting directly with the user's input, the model first "tokenizes" the user's input, simplifying it down into a unit which the actual ML model can process more efficiently. The model then spits out a token or series of tokens as a response, which are then expanded back into text or whatever the output of the model is.

I think tokens are used because most models use them, and use them in a similar way, so they're the lowest-level common unit of work where you can compare across devices and models.

Hmmmm milk is slightly acidic, and concrete will dissolve if the pH is lowered from its normal high alkalinity, so given a large enough volume of milk...I suppose milk would dissolve concrete substantially faster than water would.

There's going to be a temperature range somewhere between "fridge" and "corona of the sun" where that milk is the foulest-smelling thing in the universe.

I think its because while its under water it doesn't have a chance to diffuse into a larger volume of air -- normally farts are pretty dilute by the time it makes it to anyone's nose.

My favorite overheard undergrad story:

I was walking past the lecture hall right after an organic chemistry midterm, and there was a cluster of 4-5 students talking about the exam. One asked about question 8b, and another one said "you're not supposed to mix nitric acid and ethanol, that makes TNT, right?" I had to stifle a chuckle as I walked by.

So close, and yet so far! Nitrated acetone is explosive, and TNT (trinitrotoluene) is also made with nitric acid, but toluene is a much more complex molecule than acetone. If those undergrads could figure out how to turn acetone into TNT efficiently, they'd get a Nobel!

Agreed! I'm just not sure TOPS is the right metric for a CPU, due to how different the CPU data pipeline is than a GPU. Bubbly/clear instruction streams are one thing, but the majority type of instruction in a calculation also effects how many instructions can be run on each clock cycle pretty significantly, whereas in matrix-optimized silicon its a lot more fair to generalize over a bulk workload.

Generally, I think its fundamentally challenging to generate a generally applicable single number to represent CPU performance across different workloads.

Ehh, its not actually a big jump, and its oversold here. It's useful as an alternative to hand-designing genes, but its just a summarization tool for the gene sequences we've already annotated -- and a ballpark of 10-20% of those annotations are wrong, as it's actually very hard to annotate genes' functions correctly. I would be wary of trusting that this will work outside the most-studied protein families.