snitty

This is a really fun argument we're having where the various parties don't agree on the definition of the thing they're arguing about.

They're being very cagy, but it's quite likely they're working on "fast" maskless technology. Basically going from FPGA to chip in low enough volumes that maskless makes sense. It'd be a small, but lucrative, niche if they can get a good system at a decent node size.

That line graph is bad and it should feel bad.

As someone who is familiar with LoRa, but not LoRA, I was confused as to what they could have possibly been using LoRa for to finetune an LLM. The internet has a lot of acronyms, and a lot of jargon, and a lot of people doing seeming absurd things, okay?

LoRA, not LoRa.

I was VERY confused for a minute.

It doesn't. You need to generate models for use on the neural engine, which apple did for Stable Diffusion, but this is just taking advantage of lots of fast RAM and lots and lots of threads, if I understand it correctly.

>We are very far from understanding biology well enough to be predictive (except on TV)

I'm very excited for the next season of Small Drug Discovery on NBC. I hear the first 5 episodes this season are just screening candidates.

This is very interesting, and I think massively underestimates the variability in how small molecular drugs are affected by the body. Small differences in molecules can have drastic effects on where, how, and how quickly they're adsorbed, and what systems they affect.

More than anything, I think this just underestimates the unpredictability of it all. To extend the author's metaphor, if you measure the 100 data points from shooting 10 different projectiles out of 10 different devices, on 10 different celestial bodies, you're still not going to have a lot of predictive power that can be generalized.

There's an acronym in lexicography that I can't remember right now, but the effect of it is that people try to ascribe a nautical origin to everything. I worry this might be a little of that, but it's a fun story regardless.

Rolex are +/- 2 seconds per day.

>The main reason I've switched to Linux was because it was an operating system that was so transparent, so (relatively) simple to debug and experiment with…

>However, [Free/Open BSD] have shortcomings of their own, mainly being so conservative that they've almost been left behind by the overall open-source community.

I feel like the author doesn't understand that these are connected…

Compressive maintenance rather than percussive.

Years ago I heard some random guy talking about how he wanted to skydive with a paintball gun and shoot at the ground as he was in free fall and my first thought was, "I bet he could end up getting hit by his own paintballs."

I did that math that evening, and yeah. Absolutely. While paintballs get shot out of a marker at ~190 mph (plus 120 mph terminal velocity for a person), their terminal velocity is closer to 50 mph. So the paintball will rapidly decelerate, and you keep chugging along at over twice its speed.

>If scientists are going to complain that's its too hard or too expensive to replicate their studies, then that just shows their work is BS.

1 mg of anti-rabbit antibody (a common thing to use in a lot of biology experiments) is $225 [1]. Outside of things like standard buffers and growth medium for prokaryotes, this is going to be the cheapest thing you use in an experiment.

1/10th of that amount for anti-flagellin antibody is $372. [2]

A kit to prep a cell for RNA sequencing is $6-10 per use. That's JUST isolation of the RNA. Not including reverse transcribing it to cDNA for sequencing, or the sequencing itself. [3]

Let's not even reach things like materials science where you may be working on an epitaxial growth paper, and there are only a handful of labs where they could even feasibly repeat the experiment.

Or say something with a BSL-3 lab where there are literally only 15 labs in the US that could feasibly do the work, assuming they aren't working on their own stuff. [4]

[1] - https://www.thermofisher.com/antibody/product/Goat-anti-Rabb... [2] https://www.invivogen.com/anti-flagellin [3] https://www.thermofisher.com/order/catalog/product/12183018A [4] https://www.niaid.nih.gov/research/tufts-regional-biocontain...

>and applications (like levitating a train) require too many amps for our poor wire to remain a special wire. (Most super conductive materials lose conductivity when amps are too high).

I was wondering if there was a current limit on superconductors.

1) Is there any understanding as to why superconductivity breaks down at higher amperage? 2) If so, is there any explanation as to why that doesn't require a PhD in physics?

YBCO superconducts as high as about 93K, so we're talking 17K difference. Which is useful but not wildly so. LN2 boils at 77K, for reference.

There's a wild book from 2006 called Overdose (by Richard Epstein), that makes a more or less libertarian argument for deregulation of the drug industry. What the book misses is that before the FDA and the Pure Food and Drug Act is that anyone who thought they could make a buck by poisoning as many people as possible did exactly that. And, especially given the "tort reform" prevalent in red states, there is fairly minimal disincentive for corporations to kill people in the US.

Ooooh, I'd forgotten about that one. Thanks!

I just open a file stream and write the hex straight to the storage device. Anything else is cheating.

I learned Forth when it was the only language that had an interpreter than ran on PalmOS. So you could write and execute it on your PalmPilot while you were bored in class in the 90s.

6-7 year is not at all unusual in the sciences, at least in biological sciences. 7.5 is _very slightly_ longer than usual.

I cannot unzip the download on a Mac…

And like 90% of E. coli genome is protein coding. I intentionally wasn't limiting it to humans because humans make up a very small portion of total DNA in the world.

I learned C++ in the late 90s and didn't touch it until recently and all the new stuff basically melted my brain. It feels like a different language now.

Can you elaborate on that? A significant portion of DNA in organisms literally encodes for protein sequences. It also has functional parts (binding sites for proteins, promoter sequences). Some RNAs are not translated because the RNA itself has function, but I don't see that same argument for DNA.

RISC-V is already gaining steam in embedded systems. Apple, for one, is replacing some embedded ARM cores in their products with RISC-V cores.

As others have pointed out, we've got SBCs aplenty, and the first laptop systems are coming out later this year.

I get that as humans using computers we're eager for a RISC-V system that matches performance of other modern processors, but it's naïve to think that it isn't a viable platform until it competes with an Ryzen 7.

That makes sense. I initially read it as "absolute value." Which, admittedly, doesn't make sense for an array, but is a reasonable interpretation at first blush.

>function Gcd(a:nat, b:nat) : (g:nat,x:int,y:int))

I hope this is a typo. Otherwise it…requires an unmatched close paren?

That's a new level of cursed syntax.

EDIT: Also, is |array| a common way to get the length of an array in languages? I've never seen that (or some of this other syntax) before and it took me a minute to figure out what was happening.

DSL == Domain Specific Language?

>They're paid to solve problems in lawspace, not in the real world.

Crime, family law, wills, industrial espionage, bankruptcy, insurance disputes, personal injury…

Shame there aren't any lawyers who help folks out with the real world situations.