astrange

7 ms·

61.
by astrange 3mo ago
Bulletproof has 0.3mg capsules, but it's a pricier brand. On the other hand, I'm not sure which other brands actually have the dose they say they do.
62.
by astrange 3mo ago
Therapists have specific techniques and programs that work. Well, some of them do anyway. If you have a real serious issue like PTSD or BPD it can be mostly fixed with the right sort of therapy. It's like saying you only work with a pe
63.
by astrange 3mo ago
I'm trying ADHD therapy now, but the thing about ADHD is that you can still do complex tasks like programming yet can't do simple things. So the therapy is partly focused on emotional management, and partly about skills to do seem
64.
by astrange 3mo ago
Their bio does say they are, but they certainly forgot some things. (Some people presenting with depression should not be on antidepressants because they actually have bipolar disorder.)
65.
by astrange 3mo ago
> Atomoxetine's effect size is surprisingly similar to methylphenidate, though in my experience it treats different aspects of ADHD than stimulants. Some doctors (mine) are open to combining it with a long-acting stimulant, which ac
66.
by astrange 3mo ago
If you feel you're getting tolerance to amphetamines, the first thing to try is magnesium glycinate/threonate supplements, and then hydration and diet and sleep.
67.
by astrange 3mo ago
> Intuniv gets a bad reputation because the patient has to be titrated up to the final dose and it takes a long time to see the full effects. I went up, but then backed down because even 2mg was seriously affecting my sleep quality. I&#x
68.
by astrange 3mo ago
Well, you see, all things are empty of independent origination. (You "have ADHD" if a psychiatrist was willing to say you have it, gave you medication, and it improved your life quality both at work and at home.)
69.
by astrange 3mo ago
> Me as well, which is a really "fun" time when my specific blend of neurodivergence also causes me to immediately resent said authority and external structure and view it as removing my agency. I was never very bad at that, bu
70.
by astrange 3mo ago
> I'm still trying to understand this new reality as I'm not used to it at all. I'm having a similar issue with ADHD meds, which is that we randomly settled on one that made me feel incredibly "normal" (ie not ne
71.
by astrange 3mo ago
> Also very important, try to sleep the best you can and do exercise... I remember reading here about exercise in some cases exceeding the benefits of ADHD medications... Exercise is very true. I used to take walks but it didn't hel
72.
by astrange 3mo ago
I opened it and the TTS mispronounced the first sentence on screen. (ie it ignored the furigana and misread the first word)
73.
by astrange 3mo ago
The train has probably already hit them and they need to be cleaned off it because you don't want to give your train employees PTSD or you run out of employees.
74.
by astrange 3mo ago
I thought the point of an LVT was to force land sales for more productive uses.
75.
by astrange 3mo ago
I doubt deleting income tax is a popular view among economists. Income and payroll taxes have the feature that you always have the money when it's time to pay the tax. That also means they're disinflationary. Whereas property taxe
76.
by astrange 3mo ago
A VP is not a designer, and doesn't have a standard anything.
77.
by astrange 3mo ago
You'd be stuck on how to update "the color of a TV tuned to a dead channel". Was never sure if that meant a starry sky or a bright blue one.
78.
by astrange 3mo ago
> LLMs seem particularly bad at writing prompts for other LLMs for this reason Claude is terrible at this! Probably for the same reason that its writing style in prose is so annoying and full of claudisms.
79.
by astrange 3mo ago
It might need the longer answer to think about the question, so one approach would be to ask it normally and then ask it to repeat itself shorter.
80.
by astrange 3mo ago
If that was the case in a non-trivial way you'd see mode collapse, but you don't, they come out differently.
81.
by astrange 4mo ago
https://www.the-scientist.com/chatgpt-and-alphafold-help-des... https://arxiv.org/pdf/2503.15204 > They already fold proteins with simulations. I like how you're trying to argue with the Nobel
82.
by astrange 4mo ago
Bad pelicans are in the training set because it's read his blog post. Including a good pelican in midtraining wouldn't help the problem because you'd just produce that every time.
83.
by astrange 4mo ago
Query streams aren't really useful data in any sense. Just like nobody else is actually profiting from "selling your data".
84.
by astrange 4mo ago
It is not a sponsored article and he writes one of these every time a new model releases. Why would a professor at Wharton need to write sponsored Substack articles.
85.
by astrange 4mo ago
> Super catchy names like Opus, Mythos and Fable trying to get you to think that these software products are actually super-human life changing experiences. They're originally named after the blends at a nearby coffee shop. https:
86.
by astrange 4mo ago
How can a patch be "reproducible"? The testcases are reproducible.
87.
by astrange 4mo ago
Scraping the internet isn't a copyright violation. Using it for LLM training is much more transformative than Google and Internet Archive, which are legal.
88.
by astrange 4mo ago
They all looked like real CVEs to me.
89.
by astrange 4mo ago
Do you actually have evidence this works and doesn't degrade performance?
90.
by astrange 4mo ago
That's how a base model would work. An assistant model is simulating a human and behaves the same way a human would if you screamed at them. https://www.anthropic.com/research/emotion-concepts-function

More ›