TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Facebook's AI lab

238 点作者 vkhuc超过 11 年前

22 条评论

swalsh超过 11 年前
The AI facebook has today is already freaky to me at times... while I love any new advancement in AI, these advancements will probably not be used in my interests.
评论 #6876417 未加载
评论 #6880373 未加载
评论 #6876112 未加载
评论 #6876414 未加载
评论 #6879012 未加载
评论 #6876617 未加载
yangtheman超过 11 年前
Am I the only one who is skeptical about this announcement? I am sure the amount of data Facebook has will be huge asset to any sort of AI development, but precisely because of the amount of data and the kinds of data they have, it&#x27;s just scary what this could be used for...<p>AI that has personal information of 500M+ people, using it to manipulate people...(first to click on advertisements, and then much more). In the hands of government, I shudder to think what&#x27;s possible. With NSA already snooping around, perhaps it&#x27;s not all that distant.
评论 #6876643 未加载
评论 #6877638 未加载
demonicus超过 11 年前
&quot;We were foolish to think Google was the entity that would become Skynet, when it was Facebook all along.&quot; - John Connor
评论 #6876244 未加载
turing超过 11 年前
Wow. With Hinton at Google and LeCun at Facebook, industry has definitely been pulling some big names away from academia.
评论 #6875932 未加载
评论 #6876856 未加载
TrainedMonkey超过 11 年前
This is huge. LeCun is one of the leading AI experts in the world. With resources of Facebook I expect great things. I also see this as a technology bet by Facebook on long term strategy<i>.<p></i> I own Facebook stock, so I might be biased.
评论 #6876841 未加载
superfx超过 11 年前
Rob Fergus is joining Facebook too:<p><a href="https://plus.google.com/103199870243230279182/posts/JaqgU1sXkdN" rel="nofollow">https:&#x2F;&#x2F;plus.google.com&#x2F;103199870243230279182&#x2F;posts&#x2F;JaqgU1sX...</a>
lucb1e超过 11 年前
I&#x27;m not sure what I should think of this. What are your thoughts with regard to the amount of personal information that Facebook possesses? Regardless of whether they are good or evil, what do you think the potential implications are? I&#x27;m curious about HN&#x27;s thoughts.
评论 #6876197 未加载
aioprisan超过 11 年前
I took LeCun&#x27;s class in undergrad at NYU and it definitely wasn&#x27;t easy. He&#x27;s a fantastic speaker, definitely made it more enjoyable.
评论 #6876193 未加载
loceng超过 11 年前
This is all about finding patterns for businesses to better target and sell to people. Valuable information is likely NOT going to reach or become usable by say government organizations or other organizations that are trying to problem solve big problems - at least not be affordable, because they&#x27;ll be competing with for-profit businesses for bidding on access those users.
mturmon超过 11 年前
Star power at NIPS:<p>&quot;Facebook CEO Mark Zuckerberg, CTO Michael Schroepfer and I are at the Neural Information Processing Systems Conference in Lake Tahoe today. Mark will announce the news during his presentation at the NIPS Workshop on Deep Learning later today.&quot;<p>It seems Zuckerberg will have an apres-ski Q&amp;A, and be in a panel discussion at day&#x27;s end (<a href="https://sites.google.com/site/deeplearningworkshopnips2013/schedule" rel="nofollow">https:&#x2F;&#x2F;sites.google.com&#x2F;site&#x2F;deeplearningworkshopnips2013&#x2F;s...</a>). Hope the room is big enough.
评论 #6876223 未加载
bborud超过 11 年前
In the eyes of Facebook, we are all undressed. Now it will be like they are throwing in a free colonoscopy while they&#x27;re at it.<p>I bet the guys over at Fort Meade are beside themselves with happiness over this :-)
zerooneinfinity超过 11 年前
Seeing Peter Norvig congratulate him in the status gave me chills. Great times to be living around these titans indeed.
DanBC超过 11 年前
When Google announces AI research it makes sense because self driving cars and trying to understand what a user means by typing stuff in a search box.<p>But Facebook? They could create a million virtual DanBCs and then A&#x2F;B test something against them, and then present me with the irresistible ad, perfectly pitched to draw me in?
评论 #6876615 未加载
评论 #6877081 未加载
评论 #6876989 未加载
dhammack超过 11 年前
Can you imagine the datasets FB has to work with? Statuses, pictures, locations, social graphs. It&#x27;s incredible. It&#x27;s hard to blame LeCun&#x2F;Hinton&#x2F;Ng for moving towards industry with data like that. I&#x27;d bet that the only place with more data than FB&#x2F;Google is our good friends at the NSA.
评论 #6875911 未加载
jnardiello超过 11 年前
This is awesome and incredibly creepy at the same time.
wcoenen超过 11 年前
The combination of AI and social networks reminds me of Friendship is Optimal[1].<p>It&#x27;s a story about an AI who has the job &quot;to satisfy everybody&#x27;s values through friendship and ponies&quot; in an MMORPG but breaks out and starts <i>optimizing</i> the real world.<p>[1] <a href="http://www.fimfiction.net/story/62074/" rel="nofollow">http:&#x2F;&#x2F;www.fimfiction.net&#x2F;story&#x2F;62074&#x2F;</a>
ezioamf超过 11 年前
Google has won the search indexing game. Who will win the next level when robots understands the content?
sureshv超过 11 年前
I hope these labs are as productive as Google or the old DEC labs: <a href="http://www.hpl.hp.com/techreports/Compaq-DEC/" rel="nofollow">http:&#x2F;&#x2F;www.hpl.hp.com&#x2F;techreports&#x2F;Compaq-DEC&#x2F;</a>
X4超过 11 年前
Is any of the conference videos available online, or is that only for paying customers?
ayubskhan超过 11 年前
Cool! Will soon see online&#x2F;offline AI in action!
achompas超过 11 年前
<i>random thoughts on this announcement</i><p>I took Yann&#x27;s ML class during the first semester of my program at NYU two years ago. It was a terrible experience, given that I hadn&#x27;t worked with linear algebra or mathematical statistics since about 2008. I could barely write Python at the time, so my Lush code was some ungodly non-functional crap. That said, I&#x27;m very happy I took the class in 2011 despite hyperventilating my way through the final--it was only offered once more, and I suspect it will never be offered again given today&#x27;s news.<p>Yann really knows his stuff. His convolutional nets (and their application to MNIST and image segment classification) represented a significant improvement in computer vision, and he demoed some incredible low-latency image segmentation stuff for us. He runs one of the best neural net labs in the world (up there with Hinton&#x27;s and Bengio&#x27;s), and he has some incredible students at NYU. I can see this shaping up as a delayed acquihire of sorts...he will not struggle to find excellent candidates.<p>There&#x27;s a lot of discussion around the privacy implications, but I think everyone&#x27;s rehashing old points. Facebook already hires excellent researchers and data scientists -- John Myles White and Sean Taylor just moved out west -- but they don&#x27;t focus on images just yet, from what I know. This hiring represents an investment in image analysis on Facebook&#x27;s behalf that matches what they put into unstructured textual data and graphical inference. If you&#x27;ve already stayed with Facebook through the &quot;graph search&quot; announcement, this shouldn&#x27;t surprise you either.<p>As someone interested in this type of work, it&#x27;s an exciting time to live in New York City. Finance and adtech have been here forever, but things have expanded. Many startups (Foursquare, Tumblr, Knewton, Etsy) have invested in applied statistics and machine learning, hiring excellent researchers and engineers. Columbia and NYU have announced data science initiatives in the last 6 months. Very smart people (and others, like me ;) ) are very active in the community here.<p>There are some obvious applications of Yann&#x27;s work to Facebook&#x27;s advertising goals:<p>&gt; Identifying strong friendships through co-occurrence in photos<p>&gt; Digit or character recognition applied to marketing in photos (shop signs, brands, etc)<p>&gt; Image segment classification (e.g. beach, park, road) for use in predicting a photo&#x27;s location (for those uploaded after the fact)<p>&gt; All the Instagram photos. I mean seriously. They&#x27;re committed to making ads seamless--why not use image segmentation + likes to identify photographic structures people are attracted to?<p>Can anyone think of others?
评论 #6879157 未加载
1il7890超过 11 年前
Looks awesome for once!