paytonjjones

7 ms·

91.
by paytonjjones 4mo ago
This can usually be solved by scale alone (in all three contexts: RL, evolution, and IRT / psychometric testing) The co-occurence thing is often not a bug of the algorithm but a genuine part of the stochastic landscape that must be sol
92.
by paytonjjones 4mo ago
You could say the same thing about employees! And to be fair, the ability to fire employees and hire new ones is pretty important for that reason. In cases where you can't easily fire employees (e.g. unions), you encounter the very pro
93.
by paytonjjones 4mo ago
Tools like Playwright and Maestro can already give you a small taste of what that would look like. But overall I agree, LLMs are currently awful at being beta testers. They miss the most basic stuff that any human would immediately catch as
94.
by paytonjjones 4mo ago
This is, in short, the big current problem with AI. LLMs are built for scale so they've given up on the kind of online learning / "long term memory" processes that would individualize them. The LLM is permanently locked
95.
by paytonjjones 4mo ago
Don't worry they've got a dude named Claude they keep in the back just for occasions like this
96.
by paytonjjones 4mo ago
It's fun when subreddits or Discords are premised on a specific subtype of weirdness. There, you'll get mobbed if you're normal.
97.
by paytonjjones 4mo ago
Apply the same question to jobs and it's easy to see: why is a lossy [interview] filter better than just [firing] those who can't make it? This has enormous costs to the institution, the teachers/mentors, and of course to the
98.
by paytonjjones 4mo ago
That's called out in the README under Attribution too, hopefully it didn't seem like I was trying to bury that! > Derived from [pganssle/cim]( https://github.com/pganssle/cim ) by Paul Ganssle. Rebuilt
99.
by paytonjjones 4mo ago
Done! It was linked under Attribution but it doesn't hurt to add a second link.
100.
by paytonjjones 4mo ago
Thanks for commenting and for doing all the hard work here! (For reference, the story here is that I was using pganssle's web app as a PWA, but encountered some snags on mobile that led me to make a PR, which eventually turned into a f
101.
by paytonjjones 4mo ago
I would take a look at https://github.com/pganssle/cim , which this is very closely based on! My brief story: I was using that as a PWA, but encountered some snags when using it on mobile. I cloned it and fixed those,
102.
by paytonjjones 4mo ago
This is super interesting, thanks for commenting. You've listed a lot of negatives of absolute pitch, and they all make a lot of sense. But you also said you wouldn't trade it away. Why not? What are the positives that outweigh th
103.
by paytonjjones 4mo ago
"Sight singing" is the classic exercise to develop strong relative pitch. There are lots of resources on this — there was even a sight singing class at my college. It might be a little too challenging (and boring) for a young kid
104.
by paytonjjones 4mo ago
I 100% agree that perfect pitch is less important than these other things. The only reason I think perfect pitch is worth prioritizing early is because of the developmental gate.
105.
by paytonjjones 4mo ago
Appreciated! It's edited.
106.
by paytonjjones 4mo ago
I would be happy to edit but I don't think I yet have the ability. Cheers.
107.
by paytonjjones 4mo ago
The only age-sensitive part is the ability to map the chords to the colors. Eventually, you would want to teach them to map the color to the chord name and recognize the root of the chord. But that can be learned any time. Also keep in mind
108.
Show HN: Teach your kids perfect pitch(github.com)
217 points by paytonjjones 4mo ago|146 comments
109.
by paytonjjones 4mo ago
This suggestion is scarier than the thought someone might say a slur. You're talking about using a machine to detect social undesirables then quarantining them in the matrix.
110.
by paytonjjones 4mo ago
Often it's not the app itself doing tracking or cross-premise tracking, but data is passed to installed third party SDKs that do.
111.
by paytonjjones 4mo ago
Fun concept. Unsurprisingly a lot of these are common jokes 'plagiarised' directly from the training data. I wonder if there's a way to integrate a search tool to rule out those with near-exact matches on the web. It would be
112.
by paytonjjones 4mo ago
I think Apple has convincingly demonstrated that the look matters A LOT for high-end technology consumption.
113.
by paytonjjones 4mo ago
Whether you legalize social media or not (or smoking, or drinking, or riding in a car without a seatbelt), that's making a decision for other people's kids. Choosing parental liberty is still choosing. I sympathize with the libert
114.
by paytonjjones 4mo ago
It's a collective action problem. The psychological impacts come from secondary network effects. The studies suggest that taking social media away from just your kid doesn't do anything, because the culture wherever they go will s
115.
by paytonjjones 4mo ago
> and you actually see huge difference Correlation, causation, and all that
116.
by paytonjjones 4mo ago
Learning by doing is often more effective because it naturally involves forced recall. But it's not the only way to learn and not even the only way to do a lot of forced recall. It's downright crazy to say you aren't learning
117.
by paytonjjones 4mo ago
Pandas has lots and lots of problems. Performance is definitely one of them, but it also has inconsistent and duplicated methods, inconsistent defaults (e.g. some methods are inplace by default), copy by reference issues, I could go on. It
118.
by paytonjjones 4mo ago
Por que no los dos?
119.
by paytonjjones 4mo ago
I've never once turned on Fox News, but I have been in elite academia and I'm sorry to say the person you're replying to is dead on. That's...exactly the point of DEI statements! They are used to filter for "cultur
120.
by paytonjjones 4mo ago
If they saw it on Twitter then actively went and tried it, that wouldn't be very 'accidentally'

More ›