doubtfuluser

12 ms·

91.
Ask HN: What are the success factors for AI development?
1 points by doubtfuluser 4y ago|0 comments
92.
by doubtfuluser 4y ago
While I agree with other commenters and also with the article, that this is not like „boom, DeepMind magically speeds up everything and you have to look in more detail at numerical stability etc depending on your use case, there is still so
93.
Amazon Hiring Slowdown(bloomberg.com)
2 points by doubtfuluser 4y ago|0 comments
94.
Ask HN: Books for Reevaluation of Life
31 points by doubtfuluser 4y ago|34 comments
95.
by doubtfuluser 4y ago
Usually you have a cliff of 1 year, getting something like 25% of the shares then. Now include the fact that you got a signing bonus paid for the first 1 or two years, you can get a new signing bonus in a new company… I’ve seen people doing
96.
Meta shows model to predict based on EEG and MEG what a human was listening to(ai.facebook.com)
2 points by doubtfuluser 4y ago|0 comments
97.
'Blockchain is bunk': Crypto critics find their voice(protocol.com)
2 points by doubtfuluser 4y ago|0 comments
98.
by doubtfuluser 4y ago
In mice.
99.
by doubtfuluser 4y ago
I don’t think this is yet the beginning of the end - the company has proven to be resilient in the past (Google+, mobile). But yeah declining MAU from last quarter and overall numbers not too great. Now war-time Zuck will come.
100.
by doubtfuluser 4y ago
How do the micro kitchens look nowadays. That’s where people will realize how the company is really doing
101.
by doubtfuluser 4y ago
Excellent question and surely you have a point. The thing I’m looking for: is there a way I can use my profession for something good? I’m already helping out on weekends in the local equivalent of “soup kitchen”. But I’m rethinking life and
102.
Ask HN: As an AI expert, where to work to help humanity?
5 points by doubtfuluser 4y ago|17 comments
103.
by doubtfuluser 4y ago
In a way you are right, but on the other hand there are two parts of explainability: one is the model explanation, and one is the class explanation. Often times the explanation of the class is what people really need (I classified this as a
104.
by doubtfuluser 4y ago
As an EU citizen, I’m wondering how things like the “right to be forgotten” would work on this? How can we remove personal information in web3? I’m still new to the topic so I guess there are some easy solutions already…
105.
by doubtfuluser 4y ago
While this doesn’t feel totally surprising - I have worked with people who were using benzos for many years, still this was: #in_mice
106.
by doubtfuluser 4y ago
Well… straightforward in a way, yes. But the scale of learning is huge especially with this diverse set of tasks. Not totally unexpected, but certainly not clear that it would work with current networks and sizes.
107.
by doubtfuluser 4y ago
Edit: in addition to the part below, I now see major flaws: Of course there is a positive correlation between masks and deaths. But the causality is important: when I see more deaths in my country, of course will I wear my mask more often.
108.
by doubtfuluser 4y ago
Not surprised on this, I think the rise and demise of clubhouse predicted both behaviors. Long form audio “rooms” require constant attention to follow. I listen to podcasts when running or driving, but not as a short “distraction”. Extrapol
109.
by doubtfuluser 4y ago
I have the feeling, that the paper is flawed and missing an important experiment. All their results seem to rely on skills being used. There is indeed a need then for Amazon to prevent user tracking in skills (like for example Apple does an
110.
by doubtfuluser 4y ago
#InMice… All said
111.
by doubtfuluser 4y ago
What they did there feels like trying to replay Kevin Mitnick in the current times
112.
by doubtfuluser 5y ago
A new one for https://twitter.com/justsaysinmice
113.
Putin brings nuclear options into play(bbc.co.uk)
22 points by doubtfuluser 5y ago|1 comments
114.
by doubtfuluser 5y ago
Putting on my Paranoia hat: what if some aggressor indeed was able to introduce code into the Nvidia drivers, which - if put on enough systems - would cripple the ability to (re-)train Ai systems which might be used in military defense syst
115.
Anonymous takes down official Russian websites(twitter.com)
6 points by doubtfuluser 5y ago|0 comments
116.
by doubtfuluser 5y ago
How does it actually compare to fasttext [1] in performance. Building an interface to that in GO shouldn’t be too complicated. The claim that all language identification (lid) relies on ngrams is bold and there has been a switch to pure neu
117.
by doubtfuluser 5y ago
While Keshav‘s paper was already recommended, the beginning of this talk by Andrew Ng is along the same ideas: https://m.youtube.com/watch?v=733m6qBH-jI
118.
by doubtfuluser 5y ago
At least on the monetary side I liked Nassim Taleb‘s paper [1]. I don’t agree with everything he writes but for crypto-currencies he has some points that ring true. For crypto in general I see definitely the supply chain application, but st
119.
by doubtfuluser 5y ago
Thank you for reminding me of it. This caused me to delete my account there. I am no fan of apps that rely on synchronous experiences, and I also see much higher value in podcasts and their curation (having show notes to know in advance wha
120.
Facebook Clones Snapchat’s Goggles?(tech.fb.com)
1 points by doubtfuluser 5y ago|0 comments

More ›