Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
xyzzy123
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
32 ms
·
541.
▲
by
xyzzy123
3y ago
Ok I guess the team with the largest LLM workload in the world and billions in funding won't understand how to optimise a chip for the exact workload they have and near future ones.
542.
▲
by
xyzzy123
3y ago
I wonder how much effort is warranted though. If people want to mess up traffic they can drop a concrete block off an overpass, run a heavy chain across the road, change signage, use tire shredders or do any number of antisocial things. Man
543.
▲
by
xyzzy123
3y ago
Also: The Futurological Congress.
544.
▲
by
xyzzy123
3y ago
With a vacuum pump ;)
545.
▲
by
xyzzy123
3y ago
Fair, tho a load bearing cronjob pet at the beating heart of a $20bn company with 30m+ users is further than I would dare take this advice if I had to carry the pager. It's very common though and I guess the blog is more evidence that
546.
▲
by
xyzzy123
3y ago
Interesting they went straight from "1 box running shell scripts with flock" to "mega custom thing" w/out going thru something like kube crob jobs or an off-the-shelf scheduler in between. I find jumps like this hin
547.
▲
by
xyzzy123
3y ago
The context is often the most interesting part of the art though (while I admit pretentious wording also often rubs me the wrong way). Like the shroud of turin is just a dirty cloth if you don't appreciate the historicity of it.
548.
▲
by
xyzzy123
3y ago
Thanks! I have a big rabbit hole to go down now :) I don't get what causes the proof to fail if I provide the wrong bytes to the zkvm when it tries to read from inside the hashed area after the hash & signature are verified (this m
549.
▲
by
xyzzy123
3y ago
The selective disclosure is super cool, I wonder how it works since smthing like a hash of DG1 is what is actually signed, how can you selectively disclose verified data from "inside" the hashed area? It does not sound very feasib
550.
▲
by
xyzzy123
3y ago
Yeah I admit this is pretty situational - the U.S. I don't know much about - but NZ, Australia and most of Europe, consumer sentiment is that we are getting slugged - I don't have a survey to hand but I'm confident I could di
551.
▲
by
xyzzy123
3y ago
This is factually true but my claim is that what motivates people is a desire for greater energy security (EDIT: where improved price stability is one aspect of that, as konschubert points out).
552.
▲
by
xyzzy123
3y ago
It's a trust issue. People don't trust their utilities, residential solar is (among other things) a hedge. Basically, "how much would you pay to limit how much $powerco can screw you in future?". Human nature means a lot
553.
▲
by
xyzzy123
3y ago
How many sides does a circle have? (from 5yo) Triggered quite a fun discussion about how we might decide what "side" means, what different plausible answers might be (undefined / zero / one / infinity) and which per
554.
▲
by
xyzzy123
3y ago
I think yes, but the LED code is designed for arduino so from a code reuse / prototyping perspective it would be simpler to do initial build with 2 units.
555.
▲
by
xyzzy123
3y ago
New issue is AI. Even sites with no ads will want limits on scraping / apis / composability.
556.
▲
by
xyzzy123
3y ago
Warning about pvc is good - I feel like sterner caution about leather might be warranted - e.g. veg tan leather is fine (if stinky) but most leather you see is gonna be chrome tanned and lasering it will blast heavy metal vapour / toxi
557.
▲
by
xyzzy123
3y ago
Is it done or isn't it?
558.
▲
by
xyzzy123
3y ago
Size of backlog is inversely proportional to PM turnover. For old PM the backlog is a memory prosthesis of all the old product conversations (a graveyard of broken dreams). New PM will look at backlog in horror "what even is this I don
559.
▲
by
xyzzy123
3y ago
I think the content is largely AI generated with some human editing. It starts off feeling pretty human but then descends into robotitude from my POV. It could be entirely AI generated in which case the injection of references is pretty imp
560.
▲
by
xyzzy123
3y ago
ODROID-H3/H3+ has 2 sata ports, emmc and an m.2 but you have to be willing to do a tiny bit of hacking to get 10gbe out of it. Supports a fair bit of ram (64GB) but no ECC which is a dealbreaker for some. The "hacks" with the
561.
▲
by
xyzzy123
3y ago
I think it's easier to see if you jump up 1 level of abstraction from transistors to logic gates. Imagine an adder made up of logic gates. The gates aren't inherently synchronous - they don't have a clock input - signals appe
562.
▲
by
xyzzy123
3y ago
Is this a good faith question? Or is it intended as a "gotcha"? I'm fine to spend a few minutes if you find this sort of thing fun. I think I could, given the universe of people P, a set of belief frameworks E and a mapping B
563.
▲
by
xyzzy123
3y ago
Right, but I don't feel this is a problem with bayesian reasoning. If you want to _convince_ people you are in the domain of psychology, focus groups, marketing, social engineering etc. An argument being convincing is not the same as
564.
▲
by
xyzzy123
3y ago
How random is the third party? What epistemic frameworks do they find acceptable? (sorry if I am not getting what you are asking). A theological argument (no matter how good) is unlikely to convince an atheist of anything. In a perfect worl
565.
▲
by
xyzzy123
3y ago
A lot of the time you could say that it's structured handwaving , but the structure gives you a number of benefits over regular handwaving ;) Pretty quickly we can tell if we disagree about the whole framing of something (the overall
566.
▲
by
xyzzy123
3y ago
People want to make sure their loved ones are taken care of / will be financially secure, even after they die. This is a fairly core desire. Family businesses, family farms, family homes - if you hit people there you have to expect an
567.
▲
by
xyzzy123
3y ago
> yet is somehow completely acceptable even among hardcore libertarians. Why? A common area of concern for libertarians is the balance of power between citizens and government. In this case, the power of the government to take your (or y
568.
▲
by
xyzzy123
3y ago
Right now it's not even economic to prove that non-trivial software projects are "safe" let alone AI. AI seems much worse, in that it's not even clear to me that we can robustly define what safety or alignment mean in al
569.
▲
by
xyzzy123
3y ago
Or apply increasing "drag" as wealth accumulates, i.e, wealth taxes, maybe starting at a few bps once you have $100m. Throw in some social accolades / parades / public acclaim for people who pay the most taxes.
570.
▲
by
xyzzy123
3y ago
Piketty's r > g is pretty close to a coherent argument that capital accumulation has an "escape velocity". It's a political topic, so, vehemently contested.
More ›