ImaCake

9 ms·

61.
by ImaCake 1y ago
> clean code is not the true goal Its not, but it does matter. LLMs, being next word guessers, perform differently with different inputs. Its not hard to imagine a feedback loop of bad code generating worse code and good code generating
62.
by ImaCake 1y ago
I just use VSCode with copilot and don't worry about these re-skins. I don't get a lot of time to write code so I certainly don't have time to learn a new IDE for a small boost to my productivity when vscode gets me most of t
63.
by ImaCake 1y ago
It's a well known "dirty" secret that aerosols drive (short lived) cooling effects and that this effect is very significant [0]. In fact, the climate models used in the OP nature paper would not be useful if they didn'
64.
by ImaCake 1y ago
I suspect it's jumping on the hype train. Especially since its from a big Uni. Funding in research is all about marketing and latching onto the right keywords (just like VC really) so the most successful researchers are those who can m
65.
by ImaCake 1y ago
Those flood maps are common across the first world, sometimes called a "flood overlay" and there is usually also a "bushfire overlay" if you live in Australia. They help describe risk, in the insurance sense, and are pre
66.
by ImaCake 1y ago
Thanks for pointing out that it is counter signalling, but I would also say that it is good advice regardless. It's like an efficient highway - the road is straight and unadorned because looking "scientific" and sensible is h
67.
by ImaCake 1y ago
Thanks for the clarification, these click-bait titles pop up again and again around very interesting technical climate science, causing not only pointless panic but allowing denialists to drive doubt by pointing out the BS. Its doubly frust
68.
by ImaCake 1y ago
Brilliant way to phrase this idea! I think its incredible that we can even manage an imperfect way to escape our "social" brain. It's clearly very powerful - math, and thus all of science, only exist because of how weird pred
69.
by ImaCake 1y ago
>if you read Popper Thanks, I might actually go do this :) I recently got exposed to a very persuasive form of "rationalism is a social construct" by reading "Alchemy" by Rory Sutherland. But a theme in these comments
70.
by ImaCake 1y ago
>The problem with trying to reason everything from first principles is that most things didn’t actually came about that way. This is true for science and rationalism itself. Part of the problem is that "being rational" is a soc
71.
by ImaCake 1y ago
While their calls sound like chalk on a blackboard their body language is complex. You can communicate a lot via your actions and expressions!
72.
by ImaCake 1y ago
>My AI skeptic friends are all nuts This is dead right; my AI skeptic friend are ideologically opposed to LLMs and GenAI. Their arguments have nothing to do with how good or bad LLMs are: - IP issues (for code in a jupyter notebook which
73.
by ImaCake 1y ago
>the dependence. Sure, but that is not the point of the article. LLMs are useful. The fact that you are dependent on someone else is a different problem like being dependent on microsoft for your office suite.
74.
by ImaCake 1y ago
Maybe this means that LLMs are ultimately good for small buisness. If large buisness is constrained by being large and LLMs are equally accesible to 5 people or 100 then surely what we will see is increased productivity in small companies
75.
by ImaCake 1y ago
It's also just bad optics, unfortunately. A coal power plant kills people living nearby slowly over the course of decades, a battery fire might injure a few people but it does so in a way that makes headlines. No one cares about boring
76.
by ImaCake 1y ago
I am no expert on this, but my understanding was natural gas is useful for its flexibility and efficency to handle gaps in supply. The big batteries are amazing but they can't (yet) cover longer gaps the same way natural gas can.
77.
by ImaCake 1y ago
He blogs a lot on the OP link. I would imagine he has a few words about LLMs in there. One of his older works explores the risks of software similar to LLMs but a little more advanced.
78.
by ImaCake 1y ago
We seem to be pretty good at inventing jobs both useful and pointless whenever this happens. We don't need armies of clarks to do basic word processing these days but somehow we still manage to find jobs for most people.
79.
by ImaCake 1y ago
The price increased for sending stuff to the USA. It didn't for every other consumer market in the world.
80.
by ImaCake 1y ago
>I doubt anyone realized it could happen in one year I mean China has been modernizing their academics for a long time. See "Double First-Class Construction" [0]. But it's worth remembering that they did a lot of damage d
81.
by ImaCake 1y ago
"Don't do this or I quit". If you don't have enough money to do that, immediately start looking for a new job. If you are feeling lucky you can try weaponised incompetence to try to prevent this from being shipped.
82.
by ImaCake 2y ago
Agreed with this. A lot of people seem to think they suck, and sometimes they do, but I find them to be more useful than not. It’s actually really nice having a prose digest of the most plausible links for my search query.
83.
by ImaCake 2y ago
Not all stations are built equal, but I do agree that this is mostly symbolic. But it could be a harbinger of much more problematic censoring of US air quality data.
84.
by ImaCake 2y ago
Air Quality network usually refers to just having at least one air quality monitoring station. There's no special network here, they usually just use 4G. The stations themselves run between $100 to $50,000.
85.
by ImaCake 2y ago
It's not quite the right aesthetic but what about having solar panels and an electric car? Or even just an electric bike or scooter. There's definitely a few practical solarpunk-esque tools available to us, probably will be more i
86.
by ImaCake 2y ago
You can do this with github pages too, including your own subdomain. ChatGPT could walk you through the steps but there is no requirement for an LLM. There's been big efforts to simplify this stuff for years. Lots of good tooling out t
87.
by ImaCake 2y ago
I've never thought of Simpson's Paradox as a political problem before, thanks for sharing this! Arguably this applies just as well to Bayesian vs Frequentist statisticians or Molecular vs Biochemical Biologists.
88.
by ImaCake 2y ago
It might depend on how much you struggle with writers block. An LLM essay with sources is probably a better starting point than a blank page. But it will vary between people.
89.
by ImaCake 2y ago
I suspect SSRIs would work a lot better if we could select a better cohort for them. My sister comment notes how depression can be a symptom of a more treatable/precise disorder and so Depression becomes a broad church of several disor
90.
by ImaCake 2y ago
I suspect SSRIs are prescribed because they are easy to prescribe. I think they don't work because Depression, like Alzheimer's, is poorly defined. From a population genetics perspective we can find clear genetic signals for other

More ›