smlacy

7 ms·

1.
by smlacy 5d ago
I guess universities suffer from NIH-syndrome just as much as big tech does.
2.
by smlacy 7d ago
Or Japan? Having digital accuracy and 10 year battery life is pretty nice.
3.
by smlacy 22d ago
Yet
4.
by smlacy 1mo ago
> When you see an elegantly designed building ...was that really made by a human or did the AI generate it as part of a much larger generation step? Does it matter? Why?
5.
by smlacy 2mo ago
The ultimate driving machine... for marginal revenue.
6.
by smlacy 2mo ago
Can we please just make emoji bigger onscreen? They're not even em-height most of the time. Most interfaces don't scale the emojis when scaling the text. There's so much artistry and time & effort put into these, and the
7.
by smlacy 2mo ago
I'm yours!
8.
by smlacy 3mo ago
boycotting?
9.
by smlacy 3mo ago
So this is basically an advertisement for their product?
10.
by smlacy 3mo ago
Wouldn't work ... without additional electronic components.
11.
by smlacy 3mo ago
Hmmmm. Wonder if you could just induct through the glass with coils on each side? Seems perfect for high voltage applications?
12.
by smlacy 3mo ago
So you're implying that someone in a locked Waymo was assaulted at gunpoint from outside the vehicle? These are rolling surveillance machines (in a good way?) and virtually every aspect of this would be caught on probably a dozen camer
13.
by smlacy 4mo ago
IMHO this article misses a couple really important points. First, if the mesh can use Internet or other transports then it will, and it will be built out in a way where these become a necessity. If all you want is a silly new way to text y
14.
by smlacy 4mo ago
Yeah especially given that California is a leader in renewable energy sources.
15.
by smlacy 4mo ago
How about just ditching this disgusting use of a finite resource and switch to an infinitely abundant resource?
16.
by smlacy 5mo ago
IMHO surprisingly low. Still not clear to me why they don't just port these things to ARM or similar?
17.
by smlacy 5mo ago
156 MHz!!!!
18.
by smlacy 5mo ago
The fetishism of "byte count" (here, as "732 byte python script") needs to stop, especially when in a context like this where they're trying to illustrate a real failure modality. Looking at their source code [1] it
19.
by smlacy 5mo ago
PRINCIPLES.md
20.
by smlacy 5mo ago
What would you suggest instead?
21.
by smlacy 5mo ago
> TPU 8i pairs 288 GB of high-bandwidth memory with 384 MB of on-chip SRAM Wow. Just Wow. I presume that's for each chip, and there are 1152 chips in a pod so that's 331TB HBM and 442TB SRAM per pod. Just wow.
22.
by smlacy 5mo ago
Is it me or is that aluminum already developing some stress cracking?
23.
by smlacy 5mo ago
Post actual results, make a blog post. Don't just say "this sucks" without tangible evidence. Otherwise you're doomed to "sample size of one" level of relevance.
24.
by smlacy 6mo ago
The synth engine is an implementation of https://soniccharge.com/microtonic Not sure what "accurate" means in your context though?
25.
by smlacy 6mo ago
Yes, and with virtually zero context, which makes an enormous difference for TTFT on the MoE models.
26.
by smlacy 6mo ago
Total gimmick. I guess we're "making progress", but this is will never lead to any useful application other than "Yes, you're absulotely right" bots. What's needed for real applications is 10000× the inp
27.
by smlacy 6mo ago
This should be the top comment
28.
by smlacy 6mo ago
And with only like a dozen tokens of context. What happens when this thing gets the ~100k tokens of context needed to actually make it useful?
29.
by smlacy 6mo ago
I like the analogy but which 2 is AI coding? Fast & Cheap (but not Good?) - I wouldn't really say that AI coding is "cheap" Cheap & Good (but not Fast) - Again, not really "cheap" Fast & Good (but not Ch
30.
by smlacy 6mo ago
What's your concern?

More ›