Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
mizzao
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
94 ms
·
601.
▲
by
mizzao
9y ago
More like the 21 yard line?
602.
▲
by
mizzao
9y ago
Can you please elaborate?
603.
▲
by
mizzao
9y ago
Thats usually a sign of getting someone trying to get fake verified reviews.
604.
▲
by
mizzao
9y ago
Nice. Another account created just to attack this thread. They really do have an army of either trolls or heavily invested people (see link above)
605.
▲
by
mizzao
9y ago
Straight out of the Trump / Nunez playbook...just leak something that questionable or even contradicts your point, but tell your supporters that it exposes fraud...and most of them will conveniently apply confirmation bias to believe y
606.
▲
by
mizzao
9y ago
That thread is full of IOTA fans who are convinced that this leak "exposes" malfeasance and conflict of interest on the part of DCI. One has to wonder if they actually read the whole transcript, or just jumped on the assumption ba
607.
▲
by
mizzao
9y ago
Next up...self-driving car dashcam videos!
608.
▲
by
mizzao
9y ago
Tensorflow Playground, more of a self-driven place to play with machine learning: http://playground.tensorflow.org/
609.
▲
by
mizzao
9y ago
There's a difference between a bimodal and a distribution with a long tail (like a lognormal). The new hire distribution doesn't really give a sense of this because there's an implicit cap to how much companies are willing to
610.
▲
Ask HN: How do I replace Google Feed recommendations with Hacker News stories?
7 points
by
mizzao
9y ago
|
1 comments
611.
▲
by
mizzao
9y ago
I work at CTRL-labs, a startup focused on electromyography (EMG) based control devices. This article has a bit more technical detail about what we do: https://www.wired.com/story/brain-machine-interface-isnt-sci... EEG
612.
▲
by
mizzao
9y ago
...and you're exemplifying confirmation bias right there, which is the whole reason this happened in the first place.
613.
▲
Reading the brain from the outside
(economist.com)
2 points
by
mizzao
9y ago
|
0 comments
614.
▲
Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions
(google.github.io)
3 points
by
mizzao
9y ago
|
0 comments
615.
▲
by
mizzao
9y ago
That only works with diet coke.
616.
▲
by
mizzao
9y ago
And $2000 in electricity bills
617.
▲
by
mizzao
9y ago
Why would someone microwave blueberries?
618.
▲
by
mizzao
9y ago
Scary: Facebook probably has even more facial recognition power, and data, than the Chinese government.
619.
▲
China’s CCTV surveillance network took just 7 minutes to capture BBC reporter
(techcrunch.com)
2 points
by
mizzao
9y ago
|
1 comments
620.
▲
by
mizzao
9y ago
Data science is a horribly overloaded term. Ask a databases person and they will tell you about ETLs and relational joins. Ask a statistician and you will hear about estimators and the bias-variance tradeoff. Ask a social scientist and they
621.
▲
by
mizzao
9y ago
The halting problem has been solved!
622.
▲
by
mizzao
9y ago
Obvious in the way that Paypal, Square, Instacart, Dropbox and many other companies were obvious. You could have started any of those, right?
623.
▲
by
mizzao
9y ago
Snap is paying Google $400m a year for hosting costs... https://techcrunch.com/2017/02/02/snap-commits-2-billion-ove... Seems like at that point it might be cheaper to run your own datacenters?
624.
▲
by
mizzao
9y ago
The biggest name usually gets all the fame on academic work - a "Matthew effect". Moreover, the last author is usually the most senior person or someone who funded the project; usually perceived as important or more important than
625.
▲
by
mizzao
9y ago
https://common.net/ is a startup delivering this kind of service.
626.
▲
Microsoft is Dead (2007)
(paulgraham.com)
5 points
by
mizzao
9y ago
|
1 comments
627.
▲
by
mizzao
9y ago
Aside from the correlation !=> causation issues, there's a ton of endogeneity and selection bias in this data. The initial programming quiz has code snippets only in Javascript, Ruby, and C. The interviewers are usually more familia
628.
▲
by
mizzao
9y ago
This is true in my experience as well - most papers to these vanity journals are rejected without review by the editor ("desk reject"), and a prior relationship between a PI and the editor basically guarantees you will sail past t
629.
▲
by
mizzao
9y ago
Chrome's "own ad blocker" mentioned in the article sounds good in theory, but seems like a slippery slope to eventually making it simply show only Google's own ads.
630.
▲
by
mizzao
9y ago
I find that the copy/paste terminal stuff and keyboard shortcuts are much better by using an X-based terminal emulator, e.g. xfce: https://askubuntu.com/a/867696/86161
More ›