Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
andreyk
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
81 ms
·
871.
▲
Weak Supervision: A New Programming Paradigm for Machine Learning
(ai.stanford.edu)
3 points
by
andreyk
8y ago
|
0 comments
872.
▲
OpenAI’s GPT2 – Food to Media Hype or Wake Up Call?
(skynettoday.com)
2 points
by
andreyk
8y ago
|
0 comments
873.
▲
by
andreyk
8y ago
Well, that's not necessarily true... self driving in particular usually uses planning, which is basically all about 'what if'. But it's true this is relevant to AI in general.
874.
▲
In Favor of Developing Ethical Best Practices in AI Research
(ai.stanford.edu)
2 points
by
andreyk
8y ago
|
0 comments
875.
▲
by
andreyk
8y ago
Fun project, though it'd be nice to split up newsletters and blogs - there are a lot of quality blogs that deserve their own section (this does not have a link to Distill right now?)
876.
▲
by
andreyk
8y ago
Some other fun recent-ish developments in QA datasets: - Natural Questions ( https://ai.googleblog.com/2019/01/natural-questions-new-corp... ) - SearchQA ( https://arxiv.org/abs/1704.05179 ) Many
877.
▲
Beyond Local Pattern Matching: Recent Advances in Machine Reading
(ai.stanford.edu)
80 points
by
andreyk
8y ago
|
18 comments
878.
▲
In Favor of Developing Ethical Best Practices in AI Research
(ai.stanford.edu)
2 points
by
andreyk
8y ago
|
0 comments
879.
▲
In Favor of Developing Ethical Best Practices in AI Research
(ai.stanford.edu)
2 points
by
andreyk
8y ago
|
0 comments
880.
▲
In Favor of Developing Ethical Best Practices in AI Research
(ai.stanford.edu)
1 points
by
andreyk
8y ago
|
0 comments
881.
▲
Job loss due to AI – How bad is it going to be?
(skynettoday.com)
2 points
by
andreyk
8y ago
|
0 comments
882.
▲
Job loss due to AI – How bad is it going to be?
(skynettoday.com)
2 points
by
andreyk
8y ago
|
0 comments
883.
▲
by
andreyk
8y ago
This is a pretty good overview, though rather hand-wavy (eg the overview of RL here gives a VERY course idea of how this stuff works in practice). And the characterization of existing similar overviews as either overly technical or overly f
884.
▲
Job loss due to AI – How bad is it going to be?
(skynettoday.com)
2 points
by
andreyk
8y ago
|
0 comments
885.
▲
Job loss due to AI – How bad is it going to be?
(skynettoday.com)
2 points
by
andreyk
8y ago
|
0 comments
886.
▲
Job loss due to AI – How bad is it going to be?
(skynettoday.com)
1 points
by
andreyk
8y ago
|
0 comments
887.
▲
Job loss due to AI – How bad is it going to be?
(skynettoday.com)
11 points
by
andreyk
8y ago
|
1 comments
888.
▲
by
andreyk
8y ago
Also worth noting that it starts by imitation learning from pros. I'd be curious to see if the macro can be learned without imitation; a much harder challenge. Also, playing with full visibility as was mostly the case in the demonstrat
889.
▲
by
andreyk
8y ago
Note that this approach is very dissimilar to Alpha Go or Alpha Go 0. So it's true that they did not generalize...
890.
▲
by
andreyk
8y ago
StarCraft is also a FAR bigger challenge than Go. So much so that even saying AlphaStar is similar to AlphaGo is quite a stretch (see https://deepmind.com/blog/alphastar-mastering-real-time-stra... - the architecture a
891.
▲
Sophia the Robot, More Marketing Machine Than AI Marvel
(skynettoday.com)
1 points
by
andreyk
8y ago
|
0 comments
892.
▲
A summary of notable news about AI from 2018
(skynettoday.com)
3 points
by
andreyk
8y ago
|
0 comments
893.
▲
Summary of notable AI news from 2018
(skynettoday.com)
1 points
by
andreyk
8y ago
|
0 comments
894.
▲
by
andreyk
8y ago
While it's true that the basics of Deep Learning (eg ImageNet) are not that conceptually interesting, it is also worth noting people are working on tons of interesting directions within this framework combined with other ideas from AI
895.
▲
by
andreyk
8y ago
See also "NLP's ImageNet moment has arrived" ( https://thegradient.pub/nlp-imagenet/ ) by one of the researchers involved in the papers surveyed in this post.
896.
▲
by
andreyk
8y ago
This blog post certainly is pretty flawed on attribution of ideas - it attributes word2vec as the first to introduce word vectors (which is.... very wrong). "Word2Vec showed that we can use a vector (a list of numbers) to properly repr
897.
▲
Why Your AI Might Be Racist
(skynettoday.com)
4 points
by
andreyk
8y ago
|
0 comments
898.
▲
Sophia the Robot, More Marketing Machine Than AI Marvel
(skynettoday.com)
2 points
by
andreyk
8y ago
|
0 comments
899.
▲
by
andreyk
8y ago
Yeah, hamburger menu is weird. Also, it'd be nice to have a column for code (just links to github really).
900.
▲
by
andreyk
8y ago
General purpose only if you have access to a simulator and can do MCTS with it :) Just fyi, there are many caveats for how this relates to progress in AI as a whole (self promotion warning): https://www.skynettoday.com/edito
More ›