dnautics

14 ms·

271.
by dnautics 4mo ago
i thought the new NEM screws the homeowner because the city sells electricity at market price and buys electricity back strictly at the worst possible price, even if you're producing at peak demand.
272.
by dnautics 4mo ago
> it balloons the token cost how!!?? you write a test, which is one extra function. and maybe a paragraph or so per feature ("i made a RED test"... "i made it GREEN"), everything else is the same between normal devel
273.
by dnautics 4mo ago
they're not going to. solar and wind producers in texas are rich republicans.
274.
by dnautics 4mo ago
> not-so-great sequence are the limits on not-so-greatness for sidewinder known?
275.
by dnautics 4mo ago
12kb, you can get three gblocks in a week and change and have them assembled in two days.
276.
by dnautics 4mo ago
cool, but i dont feel like dna synthesis is a rate limiting step. unless im missing something, its sort of a micro-optimization that is likely to cost more since you still need an oligo synthesizer (and you will need to write out more base
277.
by dnautics 4mo ago
no, but renewables do speak for themselves in dollars and cents, even if they dont have subsidy. now should petrochem subsidies end too? probably yes.
278.
by dnautics 4mo ago
> the U.S. administration has chosen to actively fight against it the biggest producer of renewables is Texas, by a longshot. and the state of california just created insane NEM laws that favor the pockets of pg&e (and are shit for
279.
by dnautics 4mo ago
it depends. some places it makes sense, some places it doesn't. it will be adopted when the money speaks. the good news is that money is a reasonably close proxy to "environmental benefit" (balancing the environmental costs
280.
by dnautics 4mo ago
is human cognition not an algorithm? is it just woo? or vibes?
281.
by dnautics 4mo ago
i recall a hinton lecture where he talks about the mechanism of memory formation in the brain and it's not fully sussed out but it's also not backprop, its some sort of forward prediction and immediate reinforcement loop. There&#x
282.
by dnautics 4mo ago
> ever increasing in their manufacturing of good sent to the US. no, factories are closing and workers are protesting due to factory shutdowns in china. if you go to target, most things are made in bangladesh, pakistan India, maylaysia,
283.
by dnautics 4mo ago
we know the brain does not use gradient descent. (i agree it doesn't matter)
284.
by dnautics 4mo ago
> An LLM is not a human brain, true > and does not try to emulate one citation please. something like the universal approximation theory comes to mind, transformer architecture clearly has the shape of a universal algorithm approximat
285.
by dnautics 4mo ago
It's arguably not particularly inconvenient for the us political class? the us has been on a tear reducing per capita ghg emissions (also trade corrected ghg emissions). this has been going on for five decades now (consistently 2.5
286.
by dnautics 5mo ago
I think, for me, the thing is that when you tutor undergrads in abstract math, you discover that students will very often find data pattern that fit the goal but does not correspond to a real mathematical principle. sometimes humans making
287.
by dnautics 5mo ago
when i was programming elixir by hand i was making typing errors about 1 every half year or so. none took production down, most were caught and corrected quickly from logs. now im doing mostly llm elixir, almost all typing errors are caug
288.
by dnautics 5mo ago
theres nothing in common between humans and llms or llm training sets!
289.
by dnautics 5mo ago
its statically typed, but not runtime typed!
290.
by dnautics 5mo ago
very cool. would be even cooler if you could disguise type annotations as dialyzer annotations :P
291.
by dnautics 5mo ago
> is there any advantages of using untyped programming language without any evidence, i claim the corpus might have higher quality variable names and conventions that are "human crutches" around not having types. LLM knowledge
292.
by dnautics 5mo ago
i think the design can push people into writing unnecessary matches/guards just to trigger the typechecker. that said, I'm a fan
293.
by dnautics 5mo ago
is groq still using 6 racks to serve Llama3-70B or is that old news?
294.
by dnautics 5mo ago
yeah but you dont have to proof check every cycle, anyways rust safety is not the slow step
295.
by dnautics 5mo ago
Definitely not python. at least in the immediate future, context is the limiting factor, and not knowing if a function call might mutate your object is not good for context (the llm must either dangerously assume or look it up, both of whi
296.
by dnautics 5mo ago
i mean spacex filing reads more like an investor prospectus than an s-1 so, its a few standard deviations off the norm
297.
by dnautics 5mo ago
sure but it would be really weird if there wasn't one
298.
by dnautics 5mo ago
It's in the article. > you can implement your own version that internally checks a build-time flag, approximating C/C++ behavior
299.
by dnautics 5mo ago
> too broad of a brush yes, that was my point. as domeone who uses ai extensively to write zig (and someone who has made very small non-AI cobtributions to zig in the past), rejecting ai is currently a strategically good decision for c
300.
by dnautics 5mo ago
tigerbeetle does not, but the others do.

More ›