a_wild_dandan

10 ms·

91.
by a_wild_dandan 2y ago
Using either frontier model for basic edge device problems is wasteful. Use something cheap. We're asking "is there a profitable niche between the best & runner-up models?" I believe so.
92.
by a_wild_dandan 2y ago
It almost never is. While HN contrarians endlessly complain about it, the rest of the world happily ships Electron. Discontent with Electron is so comically over represented here that I'm beginning to think it's an insecurity or c
93.
by a_wild_dandan 2y ago
So...reinforcement learning?
94.
by a_wild_dandan 2y ago
Just use supervised testing. It's scalable, battle tested, and pervasive. Even ignoring scalability, orals have problems. KISS!
95.
by a_wild_dandan 2y ago
> - Pressure landorders who rented under the suggested price How did RealPage do this? I'm OOTL.
96.
by a_wild_dandan 2y ago
Exactly! Have you tried the Phi models? To me, they indicate that we can get much more efficient models. In a few years, 70b on gold standard synthetic data + RL might run circles around SotA. It's such an exciting time to be alive.
97.
by a_wild_dandan 2y ago
I run q5s usually, since it's a 40% haircut on model size, with nearly no PPL loss. (Presuming an 8b native model like Qwen.)
98.
by a_wild_dandan 2y ago
Yeah, I would want the 192GB Mac for attempting such hefty models. But I have such basic bitch needs that 405B is overkill haha.
99.
by a_wild_dandan 2y ago
For 70B models, I usually get 15-25 t/s on my laptop. Obviously that heavily depends on which quant, context length, etc. I usually roll with q5s, since the loss is so minuscule.
100.
by a_wild_dandan 2y ago
70B+ models typically run great with my MacBook's 96GB of (V)RAM. I want a Mac Studio to run e.g. llama-405B, but I can't justify the marginal model quality ROI for like $7k or whatever. (But I waaant iiit!)
101.
by a_wild_dandan 2y ago
We don't. That person has no idea what they're talking about. Virtually no AI researcher would require "neuron simulations" for intelligence, nor agree with their AGI claims. (See: recent machine learning expert surveys,
102.
by a_wild_dandan 2y ago
What’s normal? On your Apple silicon.
103.
by a_wild_dandan 2y ago
Hence their qualifier “rarity.”
104.
by a_wild_dandan 2y ago
My sheer productivity boost from these models is miraculous. It's like upgrading from a text editor to a powerful IDE. I've saved a mountain of hours just by removing tedious time sinks -- one-off language syntax, remembering pa
105.
by a_wild_dandan 2y ago
> You are still responsible for what you do regardless of the means you used to do it. And a lot of people use this not because it’s more productive but because it requires less effort and less thought because those are the hard bits. Ye
106.
by a_wild_dandan 2y ago
This tests vision, not intelligence. A reasoning test dependent on noisy information is borderline useless.
107.
by a_wild_dandan 2y ago
I mean, I passed a physics PhD exam and I can’t model the Empire State Building. The jury is still out on whether I’m an intelligence tho.
108.
by a_wild_dandan 2y ago
Sounds like a good job for a new nonprofit named Black Mirror or something.
109.
by a_wild_dandan 2y ago
You could try zx[1]! I prefer it to Bash scripting. [1] https://github.com/google/zx
110.
by a_wild_dandan 2y ago
I wonder if using nested markdown bullet points would help. You would preserve the information hierarchy, and LLMs are phenomenal with (and often output) markdown.
111.
by a_wild_dandan 2y ago
Let’s start calling these stories FAErytales.
112.
by a_wild_dandan 2y ago
> But how do you reason? My theory: reasoning is the application of analogies. Here `analogy = memorization + pattern matching`. Pattern matching is just an associative memory query; remembering an example with fuzzy enough details to ap
113.
by a_wild_dandan 2y ago
We're unsure how minds or LLMs work. So let's also not dismiss potential parallels either. It's okay to not know stuff. We'll get there!
114.
by a_wild_dandan 2y ago
You're blaming TypeScript for self-inflicted wounds. Don't blame the type system that you banished (with `as unknown as T`) for not catching you; or for some React library having bugs, e.g. an incorrect interface; for not defining
115.
by a_wild_dandan 2y ago
Using TypeScript for the entire stack feels like a superpower. The type system is incredible. V8 is fast. The frameworks are phenomenal (Next.js, Material UI, etc). The ecosystem is enormous, with packages for everything . The unified code
116.
by a_wild_dandan 2y ago
By "Colorado as a whole", do you mean "Boulder in particular"?
117.
by a_wild_dandan 2y ago
That doesn’t seem anywhere remotely near obvious to me, but that might say more about my policy ignorance than reality.
118.
by a_wild_dandan 2y ago
There is a "no downstream dependencies" option; it's called writing/auditing everything yourself. Everything else -- be it libraries, monolithic SaaS platforms, a coworker's PR, etc. -- is a trade off between your t
119.
by a_wild_dandan 2y ago
Just a simple 2-queue. I love this, since it's low friction and your essentials are persistent. No more brain-wracking to remember what the grocery list is missing.
120.
by a_wild_dandan 2y ago
Me too! If I must take out some forgettable item, it sits blocking the door.

More ›