landon32

13 ms·

61.
by landon32 9y ago
I don't know Facebook's specifics, but typically in research they will train a wide variety of examples, e.g. English -> French French -> English English -> German French -> German
62.
by landon32 9y ago
It'd be fine to say something like "demonstrated exceptional reliability and proficiency in operating their machinery" could work. I know nothing about factories though
63.
by landon32 9y ago
>If someone has made a project in TensorFlow to paint in Picasso style then you should hire them regardless of education. Hmm, I definitely don't agree. Once the Style Transfer paper was released, it's pretty trivial to impleme
64.
by landon32 9y ago
Yeah that's interesting, it does have some kind of memory with all the genes we have. I mostly just meant that Nature is not an intelligent agent with an agenda (unless you believe that in a spiritual way, which is cool too). >If we
65.
by landon32 9y ago
> For Nature to learn how to develop a nervous system capable of flexibly interacting with the environment, culminating in our brains, took hundreds of millions of years. Nature never learned anything. Nature is not a force that chooses
66.
by landon32 9y ago
Yes for sure. I am very math-averse (almost failed calculus 1, skated by pre-calc in high school... not my intelligence area) and though I don't have a full time job doing something in AI, I do feel like I can build models that solve r
67.
by landon32 9y ago
Also used for robotics: https://blog.openai.com/roboschool/ and other control problems: https://vmayoral.github.io/robots,/ai,/deep/learning,/rl,/re... I think not so long ago
68.
by landon32 9y ago
Doing extra-ordinary well, or even just "doing great" or "doing very well". Usually, "it" in "killing it" refers to some kind of game, so they're "killing the game", or doing a really g
69.
by landon32 9y ago
Hmm, I think there's a number of people who have some limited AI experience (side project, internship, etc.) but don't have the money to invest in a huge amount of cloud computing hours to do novel research. It doesn't seem l
70.
by landon32 9y ago
Their architecture is definitely novel. What are you referring to that made it sound like real AI we have could already do this?
71.
by landon32 9y ago
Right now we create feed forward neural networks that have a very specific architecture so that we can easily train them. If we want a really "general purpose" sufficiently complex neural network we would need it to use way, way,
72.
by landon32 9y ago
You're right, but I think this paper is the first step on a (potentially very, very) very long road to building machines that could "imagine no possessions"
73.
by landon32 9y ago
This title is a very strong exaggeration. It's actually feasible for someone to start working on Artificial Immortality right now, but this article is just talking about making a chatbot that sounds like his dad.
74.
by landon32 9y ago
>Also, this article reeked of AGI ideas. Deep learning isn't trying to solve AGI. Reasoning and abstraction and high level AGI concepts that I don't think apply to deep learning. I don't know the path to AGI but I don&#x27
75.
by landon32 9y ago
That's really interesting though, you might be able to combine something like Random Forest with Neural Nets to make them more robust to adversarial images.
76.
by landon32 9y ago
have we tried using DL to construct 3d space, using either video or 3d images? That seems closer to what humans do than just the 2d images
77.
by landon32 9y ago
Wow, I had never considered superintelligence that wasn't at least at some level modeled after the human brain. That is crazy to think about. We could be at the very low end of the spectrum of intelligence I guess.
78.
by landon32 9y ago
Could actually be their data, right? Imagine if you had only had experience with software engineering. The only data you use when engineering software are the data you learn when using the product or writing tests, it's all the algorit
79.
by landon32 9y ago
Hmm, sometimes I think that we won't get super close to AGI until we can actually model something the size of a Human Brain (in terms of neurons or Synapses). Human Brain has 1B+ Neurons, or 4Qu+ synapses. So that's 12.5 GB all at
80.
by landon32 9y ago
This is totally true, but I think it's still important to note that while something like Artificial General Intelligence is still way beyond the state of the art, the state of the art still has a huge impact on the world. A tiny slice
81.
by landon32 9y ago
Yeah, I think the author's just priming the pump for a few more posts in this series that show a new way of abstraction/new NN architecture that solves some new problems. Doesn't seem like he's trying to claim anything p
82.
by landon32 9y ago
I don't think he's talking about literally mimicking how the human brain works. It seems like he's just talking about making neural nets more effective in certain tasks by allowing for more types of abstraction, just like a h
83.
by landon32 9y ago
I know I don't have all the information, but just on first glance-- I think it's pretty inappropriate for a VC to ask an entrepreneur who is at demo day (presumably pitching) to babysit their kids, unless they had known each other
84.
by landon32 9y ago
Yeah, I think a service like this becomes much more effective when both positive and negative reviews can be posted. Also would help a lot if you could see analytics on the review, like average rating from each gender. Then it would be much
85.
by landon32 9y ago
I would love to hear more as well. First thought is that if pretty experienced founders who are starting a company in a space YC isn't experienced in(e.g. biotech 5 years ago) went through it they might not get much out of it.
86.
by landon32 9y ago
Wow, this is wild to me. At first I thought this was only in a professional context, but it's not. I cannot believe a plurality of people think that having dinner 1-on-1 with someone of the opposite sex is inappropriate if they're
87.
by landon32 9y ago
I don't think high speed rail would help here. We already have normal speed trains that are affordable, and I'd imagine high speed rails would just be more expensive. Unless you're talking about the highly possible near futur
88.
by landon32 9y ago
re: contract work, the reason why they would hire you instead of put an employee on it is that you will be able to get it done faster and will probably do a better job. Sure, they could have one of their employees do it, but that would take
89.
by landon32 9y ago
I disagree that it's clickbait Might have some silly opinions toward the end, but I think the summary and analysis in the first half-2/3s is nice.
90.
by landon32 9y ago
This is cool. Is it mobile only?

More ›