Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
llm_trw
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
46 ms
·
901.
▲
by
llm_trw
2y ago
Just remember that the reason why Ted Kazinsky said he bombed the federal government was because he was subjected to MK Ultra experiments in college. I'd say you got a good deal getting two dollars instead of life long trauma.
902.
▲
by
llm_trw
2y ago
I've eaten a bag of marshmallows today while coding till 4am on a Friday night. Your proxy tests for self control have no power here.
903.
▲
by
llm_trw
2y ago
Birth rates did not fall below replacement until 1972. Blaming the depression and word war is a bit of a stretch.
904.
▲
by
llm_trw
2y ago
Let me tell you about fall out and how the majority of people who die in a nuclear attack will feel the meat peel off their bones over two weeks of unimaginable agony with nothing that can be done for them.
905.
▲
by
llm_trw
2y ago
Worrying about nuclear war was probably worse for mental health since you'd die quite quickly (if you were lucky), rather than some nebulous future were something's going to happen.
906.
▲
by
llm_trw
2y ago
The majority of people lived in villages until 1920 in the US. US birth rates started dropping off not a generation later.
907.
▲
by
llm_trw
2y ago
Yes and?
908.
▲
by
llm_trw
2y ago
>we don’t yet know best practices for not blinding our membership short of building a completely enclosed separate room for it with door interlocks. That is the only way to not blind your membership.
909.
▲
by
llm_trw
2y ago
Do they have a local model? I keep getting burned by APIs having stupid restrictions that makes use cases impossible that are trivial if you can run the thing locally.
910.
▲
by
llm_trw
2y ago
The animation, not the movement decisions. This is as naive as calling an industrial robot 'just a calculator'.
911.
▲
by
llm_trw
2y ago
Is there any models that can do diarization well yet? I need one for a product and the state of the art, e.g. pyannote, is so bad it's better to not use them.
912.
▲
by
llm_trw
2y ago
>For all of the hype around LLMs, this general area (image generation and graphical assets) seems to me to be the big long-term winner of current-generation AI. Let me show you the future: https://www.youtube.com/watch?v=
913.
▲
by
llm_trw
2y ago
I think we saw how it plays out in the last few days. >The worst IT outage ever! >>The worst IT outage so far.
914.
▲
by
llm_trw
2y ago
The description of updates is painfully true. A long time ago I worked at a broker trader where all communications, including servers communications, had to go through zscaler as a man in the middle. What had been routine all of a sudden be
915.
▲
by
llm_trw
2y ago
One of these days I'll move to Cuba and run all the long list of exploits I've been told aren't exploits by overpaid idiots. https://www.youtube.com/watch?v=pQMuSotWCEs
916.
▲
by
llm_trw
2y ago
I'm reminded of Darpa's old plans for Ada. I expect we'll see the same issues come up as the last time they tried this.
917.
▲
by
llm_trw
2y ago
Because it is half the range expressed in 1 bit of exponent, the same way that [1,3) is half the range expressed in 2 bits of exponent. I'd used [0,2) and [2,4) but that would confuse people used to thinking in base 10, which includes
918.
▲
by
llm_trw
2y ago
Floating points numbers have a fix precision mantissa, and a fixed precision exponent. So you have xxxxx E xxx as an example of a 5 bit mantissa and 3 bit exponent. You have 2^5 floating point numbers for each possible exponent. So no, you&
919.
▲
by
llm_trw
2y ago
People just don't know any better.
920.
▲
by
llm_trw
2y ago
This is a very misleading article. Floats are not distributed evenly across the number line. The number of floats between 0 and 1 is the same as the number of floats between 1 and 3, then between 3 and 7 and so on. Quantising well to intege
921.
▲
by
llm_trw
2y ago
Cheap, easy, good. Pick two.
922.
▲
by
llm_trw
2y ago
Feed it grounding text that's about as long as the output text you expect it to produce. They are called transformers for a reason.
923.
▲
by
llm_trw
2y ago
A bigger audience isn't a better audience. A violin solo in a football stadium with the usual football audience is not an improvement over one in a recital room with the usual audience.
924.
▲
by
llm_trw
2y ago
Given what OS X has become it's un-obsoleted itself again. It's kind of ironic that you're using a post from 20 years ago to invalidate an essay from 25 years ago, about an OS that's been substantially dumbed down in the
925.
▲
by
llm_trw
2y ago
>Months as in 2 months is pretty normal. FAANG processes often take 6+ weeks. What sort of barrel are you scraping that you can spend that long on a hire? The people I've had to hire were on the market for two weeks every five years
926.
▲
by
llm_trw
2y ago
What value does a google search provide?
927.
▲
by
llm_trw
2y ago
Yes, all those things mean I get to find out what _groups_ buy my products. The plural of anecdote is data after all.
928.
▲
by
llm_trw
2y ago
>In the long term what matters is the people's quality of life and diversity. I think you'll be surprised to find out just how many Europeans disagree with the diversity part. We are quite close to ethno-state Europe, just like
929.
▲
by
llm_trw
2y ago
>A food startup who can't afford to hire those people might argue that they are at a disadvantage and therefore the regulations are onerous and favor large companies. That is literally the point of food regulations. They are not pre
930.
▲
by
llm_trw
2y ago
You'd need to ban targeted marketing, not just advertising. If I were selling widgets I still greatly care about knowing who buys a billion widgets a year and will pay good money to find out.
More ›