Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
neatze
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
25 ms
·
241.
▲
by
neatze
5y ago
This was the reason why I switched to Linux more 10 years ago, in addition to copy/past, history of what has been done and taking notes is something I don't think is replaceable with GUI, furthermore eventually I realized all comm
242.
▲
by
neatze
5y ago
Interesting to see how much slower is JS when compared to Fortran for matrix operations, no idea how to compare React to Fortran.
243.
▲
by
neatze
5y ago
What makes you say this ?
244.
▲
Apollo 14 and 16 employed mortar with explosive on the Moon
(en.wikipedia.org)
1 points
by
neatze
5y ago
|
0 comments
245.
▲
by
neatze
5y ago
Not sure what you mean by goals, because to a degree you don't need dynamic goals (eg. goals that change throughout lifetime of a system) for reactive behavior.
246.
▲
by
neatze
5y ago
My only question is; why such project was connected to the internet at all ?
247.
▲
by
neatze
5y ago
In analogy, it would be similar to pointing bright flash light directly into the eyes, I guess.
248.
▲
by
neatze
5y ago
Yes, sorry about typo and thanks for the link.
249.
▲
by
neatze
5y ago
Birds abilities to navigate is outright insane, for example albatross can travel ten thousands of miles, I believe it is only bird that circumvents earth, I am not convinced that magnetoreceptor is the only answer to how birds able to navig
250.
▲
by
neatze
5y ago
How and why are you better then onerep.com ?
251.
▲
by
neatze
5y ago
It is false to compare dog or any other animal intelligence (arguable even single cell organisms) to today's most if not all AI approaches, because adaptive and evolutionary dynamics of natural intelligent systems is in a different com
252.
▲
by
neatze
5y ago
While I agree with you to a large degree, from my limited perspective there is one problem with server side cheating detection, such detection would be inherently reactive (eg. it will take few cheating actions, before it is detected by ser
253.
▲
by
neatze
5y ago
Interesting choice of LGPL license with exception.
254.
▲
by
neatze
5y ago
this is just false.
255.
▲
by
neatze
5y ago
My first thought what an outstanding pdf, my second though; this should be in latex on github/gitlab, so people can contribute and maintain, I would love to add numerical psedo code to as much equations as I possibly can with my primit
256.
▲
by
neatze
5y ago
It is not only that, awareness plays crucial role in safety, there are numerous factors that negatively effects awareness, some factors might not be related to job at all, it might be "simple" commutative cognitive fatigue. Arguab
257.
▲
by
neatze
5y ago
store of value with 50% weekly swings? :)
258.
▲
by
neatze
5y ago
Seems like author makes argument of composition over inheritance in some form, without specific use case and most importantly possible evolution of given use case, arguing about such topic to me is useless, to very large extent.
259.
▲
by
neatze
5y ago
> The dataset comprises imaging data that covers roughly one cubic millimeter of brain tissue, and includes tens of thousands of reconstructed neurons, millions of neuron fragments, 130 million annotated synapses, 104 proofread cells, an
260.
▲
by
neatze
5y ago
After skimming through, awesome overview, but to me article should have mentioned; modularity, diversity, and specialization as an critical aspects of self-organization.
261.
▲
by
neatze
5y ago
What is not dangerous ?
262.
▲
Programming Languages Drag Race: Calculating Primes
(github.com)
1 points
by
neatze
5y ago
|
0 comments
263.
▲
by
neatze
5y ago
What really happened (to the best of my understanding): - Researchers submitted paper to IEEE. - Researcher twitter about it. - Tweet was deleted, because people pointed out it was bad humans subject researcher. (consent and deception) - Ot
264.
▲
by
neatze
5y ago
Can you provide specifics what do you mean exactly by unfounded accusations ? Really curious who makes unfunded accusations ...
265.
▲
by
neatze
5y ago
> researchers could be totally nice and ethical people. A thorough review of the IRB documents revealed potential problems in the description of the experiments, and concluded that insufficient details about the experimental study
266.
▲
by
neatze
5y ago
Dual Gigabit Board: https://www.seeedstudio.com/Rapberry-Pi-CM4-Dual-GbE-Carrier...
267.
▲
by
neatze
5y ago
There are multiple copies of deed within multiple organization, what kind of person just trusts single copy without due diligence ?
268.
▲
by
neatze
5y ago
If new product/service is not lead by Mark Zuckerberg you should not trust it ether, since you have no track record, furthermore past performance is not guarantee of future results.
269.
▲
by
neatze
5y ago
Irrespective of real/fake name, behavior is most critical factor in trust, furthermore behavior changes, so interactions is your only information for degree of trust. Your real name is just label nothing more. Distrust is cognitively t
270.
▲
by
neatze
5y ago
You can trade bitcoin futures.
More ›