mappingbabeljc

4 ms·

1.
by mappingbabeljc 7y ago
Hiya, I work at OpenAI. I think the Grover paper is a good place to read about some of this: https://arxiv.org/abs/1905.12616 We're likely publishing more on detecting fine-tuned outputs in the future, also.
2.
by mappingbabeljc 8y ago
This does not have anything to do with OpenAI (though it's a cool initiative.) Source: I work at OpenAI.
3.
by mappingbabeljc 8y ago
Hi there, I work at OpenAI. This isn't related to OpenAI in any official sense, though there are some people involved who either have had links to us in the past, or currently have links to us.
4.
by mappingbabeljc 9y ago
Correct - we saw a similar phenomenon of rapid capability gain via self-play in our Dota 2 work: https://blog.openai.com/more-on-dota-2/
5.
by mappingbabeljc 9y ago
(Disclosure: I write Import AI). Link: http://jack-clark.net/import-ai
6.
by mappingbabeljc 9y ago
Hi there, the weights and model are here: https://github.com/openai/generating-reviews-discovering-sen...
7.
by mappingbabeljc 10y ago
I write a weekly AI newsletter called Import AI which is also cross-published to this WP blog. I try to cover a mixture of fundamental research papers and applied stuff. It also includes some OpenAI updates: https://jack-clark.ne
8.
by mappingbabeljc 10y ago
Hiya, (I'm Jack Clark) - this is a good point. The child probably gets about 50 to several hundred distinct 'frames' of the chicken. Still, a remarkably small number of examples.
9.
by mappingbabeljc 10y ago
You can create environments - it's coming! We'll be releasing many components over next few months.
10.
by mappingbabeljc 10y ago
There was an interesting academic research paper that showed you could train in GTA and transfer over to the KITTI dataset and do ok: https://arxiv.org/abs/1610.01983
11.
by mappingbabeljc 10y ago
We wanted people at NIPS in Barcelona to have something nice to read over their morning coffee and such. [I work at OpenAI - @jackclarksf on Twitter]
12.
by mappingbabeljc 10y ago
It just instituted a bunch of changes which will radically change the magazine and may make it break even. "A person familiar with the matter said the publication, which Bloomberg acquired in 2009, was on pace to lose between $20 milli
13.
by mappingbabeljc 10y ago
Even more interesting - an OpenAI paper on the roughly same subject/technique came out a little earlier (RL2 fast reinforcement learning for slow reinforcement learning https://arxiv.org/pdf/1611.02779.pdf ). I thi
14.
by mappingbabeljc 10y ago
Hiya, I wrote the article. What might not be captured here is that this is above all a community of people all working on AI so it's not like anyone is really emotional and/or aggressive about this issue. It's more that it&#x
15.
by mappingbabeljc 10y ago
Hi there, I wrote the article. We also looked at data on things like number of forks and contributors on GitHub as well as massive rise in questions on StackOverflow (thanks to Delip Rao's great post here http://deliprao.com
16.
by mappingbabeljc 10y ago
they plan to publish more details in a research paper, AFAIK.
17.
by mappingbabeljc 10y ago
One area where Google does seem to avoid this secrecy is in AI: both Google Brain and DeepMind are consistently publishing papers & in many cases putting them up on Arxiv pre-publication.
18.
by mappingbabeljc 11y ago
I'm an optimist in this area - people like Wait But Why and XKCD (eg - https://xkcd.com/thing-explainer/ ) - show that most complex subjects can be explained in simple language that most people with an average educ
19.
by mappingbabeljc 11y ago
Hi there, author here -- mostly trying to be accessible. And I think in the tech community it is well understood that robotics is hugely difficult, but it seems like general public pretty much equates progress in software AI with progress i
20.
by mappingbabeljc 11y ago
Hiya. Reporter here. On the press briefing call Hassabis said that the single node version won 494 out of 495 games against an array of closed- and open-source Go programs. The distributed version was used in the match against the human and
21.
by mappingbabeljc 11y ago
Hiya. Author here. I'd class some of Apple's main peers as: Google, Microsoft, Facebook, Samsung, Baidu, and, to a lesser extent, Amazon. These are all large consumer technology companies that are trying to develop quite intimate
22.
by mappingbabeljc 11y ago
Hiya. For those interested, the RankBrain approach of converting words and phrases into vectors ties directly to Geoff Hinton's more ambitious ideas about AI. He speaks about it a bit from 32 mins in, in this video from the Royal Socie
23.
by mappingbabeljc 11y ago
Hiya. They wouldn't explicitly confirm that it is word2vec, but everything we discussed indicated it's likely doing something roughly equivalent to word2vec, and is also doing similar conversions for sequences which is likely conn
24.
by mappingbabeljc 11y ago
Thanks for the feedback. As a general news organization we struggle with definitions/scoping for stuff within AI as it's such a new area and we try to write for a broad, albeit informed, readership. I'll keep this in mind for
25.
by mappingbabeljc 11y ago
Hiya, author here - yes, you are correct. RankBrain is one of hundreds of distinct signals that go into the results page. It just happens to be one with a great deal of influence, which I think demonstrates the surprising generality/ad
26.
by mappingbabeljc 11y ago
Tried to get at this in the article (author here), but it's using vectors (think word2vec and seq2seq) to distill meaning and embed words and phrases into a single space that the computer can then use to reason about. From my understan
27.
by mappingbabeljc 11y ago
Hiya. Author here. They wouldn't tell me. Asked a lot.
28.
by mappingbabeljc 11y ago
Hiya. One of the authors of the article here. We also reported this week that Azure alone likely did about $400m in revenue for Microsoft in June quarter, versus for Amazon on $1.8 billion. http://www.bloomberg.com/news&#x2F
29.
by mappingbabeljc 11y ago
Hiya - I wrote the article, it wasn't from PR. As far as I know the supplier stuff, revenue numbers, signup numbers, are new.
30.
by mappingbabeljc 11y ago
Disclaimer: I wrote the article. I feel the HN community may find some of the business numbers for DigitalOcean and Backblaze interesting.

More ›