Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
sanxiyn
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
20 ms
·
301.
▲
by
sanxiyn
3y ago
250 K at ambient pressure is still revolutionary, as it can be reached by dry ice. The highest critical temperature at ambient pressure had been something like 150 K: still above boiling point of liquid nitrogen, and won Nobel Prize in Phys
302.
▲
by
sanxiyn
3y ago
Some background: there were two Chinese teams publicly pursuing LK-99-derived room temperature superconductor, which I arbitrarily named "north China team" and "south China team". North China team was headed by Hongyang
303.
▲
by
sanxiyn
3y ago
This is a derivative, with slightly different chemical formula (namely, this includes sulfur, while LK-99 didn't).
304.
▲
by
sanxiyn
3y ago
> LOLWUT is also going to be a tradition starting from Redis 5. At each new major version of Redis what the command does will change completely. FYI, here is LOLWUT for Redis 6: https://twitter.com/antirez/status
305.
▲
by
sanxiyn
3y ago
Grid connection is a real problem. Solar power in US waits years for grid connection. https://www.cnbc.com/2023/04/06/outdated-us-energy-grid-tons...
306.
▲
by
sanxiyn
3y ago
"In 1989, Korea began construction on their first domestically developed OPR-1000 design... Twelve reactors of this standard design began construction between 1989 and 2008, and their costs declined in a stable manner... representing a
307.
▲
by
sanxiyn
3y ago
Yes it will, but experience from South Korea says it won't be cheap enough to matter. See https://news.ycombinator.com/item?id=30380897 .
308.
▲
by
sanxiyn
3y ago
Eh, this is an independent confirmation (from Chinese team, of Korean team).
309.
▲
by
sanxiyn
3y ago
As I understand, neither EmDrive nor E-Cat could be reproduced by any third party. This paper is a third party reproduction, so it seems pretty different.
310.
▲
by
sanxiyn
3y ago
Why do you think this?
311.
▲
by
sanxiyn
3y ago
That's Jicheng Liu, the first author of this paper, and a different person.
312.
▲
by
sanxiyn
3y ago
At least she immediately commented on this paper. https://twitter.com/iris_IGB/status/1736947651126415426
313.
▲
by
sanxiyn
3y ago
Since the explainer is pretty long (but it is much more fun to read than the paper, because it is written for the lay people and the paper for experts), here is a sample. It is a description of discovery of memory effect referred to in the
314.
▲
by
sanxiyn
3y ago
What modifications? About the only thing they write about synthesis in this paper is "We follow the procedure of Lee et al. to synthesize our CSLA samples, with 1:1 mole ratio of Pb2(SO4)O and Cu3P", which doesn't show any mo
315.
▲
by
sanxiyn
3y ago
A corresponding author (Yao Yao) wrote an explainer of the paper, which you should read together. The paper itself is pretty cryptic. https://www.zhihu.com/question/635259000/answer/3330698364
316.
▲
by
sanxiyn
3y ago
I don't think any conspiracy theory is needed. Since old board needs to agree to new board, some compromise was made.
317.
▲
by
sanxiyn
3y ago
This requires retraining from scratch, so no, you can't use Llama 2 pretrained weight. As far as I can tell you can take Llama 2 modelling code, training infrastructure, training data and apply proposed modification (they provide PyTor
318.
▲
by
sanxiyn
3y ago
Why? As an economist, he perfectly understands what is a public good, why there is a market failure to underproduce a public good under free market, and role of nonprofit in public good production.
319.
▲
by
sanxiyn
3y ago
IRS requires a nonprofit to have a minimum of three board members for such reasons.
320.
▲
by
sanxiyn
3y ago
Old board needs to agree to new board, so I think some compromise is inevitable.
321.
▲
by
sanxiyn
3y ago
Maybe Sam wanted to redirect Ilya's GPUs to ChatGPT after DevDay surge. 20% of OpenAI's GPUs are allocated to Ilya's team.
322.
▲
by
sanxiyn
3y ago
As far as I can tell, this is correct.
323.
▲
by
sanxiyn
3y ago
I am not sure why. As far as I can tell, the board doesn't need to answer to anyone.
324.
▲
by
sanxiyn
3y ago
Emmett Shear is probably the person most friendly to OpenAI board's AI safety agenda among possible candidates. Source: have a look at his Twitter.
325.
▲
by
sanxiyn
3y ago
On the other hand, Anthropic exists, so I am not sure.
326.
▲
by
sanxiyn
3y ago
Groq is planning to do that. https://twitter.com/GroqInc/status/1691436573014020096
327.
▲
by
sanxiyn
3y ago
My understanding is that if you are renting your data center space right now, power still is the limiting factor and you will need to leave some rack space empty to be able to run GPUs. On the other hand, if you are building your own data c
328.
▲
by
sanxiyn
3y ago
There is a list of packages in other languages at the bottom of the README.
329.
▲
by
sanxiyn
3y ago
This reminds me a lot of Nervana Systems' Flexpoint.
330.
▲
by
sanxiyn
3y ago
> The western world's energy and resource consumption isn't sustainable If god wanted us to have infinite energy and resources he would have put a gigantic fusion reactor in the sky. (OH on Twitter)
More ›