ericjang

16 ms·

91.
by ericjang 6y ago
Interesting! Is there some kind of meta-objective for "AI optimal" that could have replaced the 6 or 7 iterations you did with human R&D? For instance, if you had real human playtesters interacting with the prototypes, is ther
92.
by ericjang 6y ago
proof-of-work based blockchains are technically also an arms race between good and bad actors. It still works because extending the chain costs compute, so as long as the good actors have the majority of compute, all is well. Training a GAN
93.
by ericjang 6y ago
OP here. I hope that's not the takeaway readers glean from my article - the point I was making was just that it doesn't make sense to shoehorn a biophysical learning mechanism into a DNN, rather we should use a DNN to find a bioph
94.
by ericjang 6y ago
This is a good question. I think many "biologically plausible" neural models are willing to make some approximations for the benefit of computational power (e.g. rate coding instead of spike coding, point neurons and synapses inst
95.
by ericjang 6y ago
Deep Learning has nothing to do with biophysical neuron simulation, even though there is a confusing overloading of the term "neural network". A good introduction to deep learning is this chapter: https://mlstory.org&#
96.
Don't Mess with Backprop: Doubts about Biologically Plausible Deep Learning(blog.evjang.com)
91 points by ericjang 6y ago|53 comments
97.
by ericjang 6y ago
Thank you for sharing the photos ! Some of the poses are pretty funny :)
98.
by ericjang 6y ago
Which is fair. I don't think the interview is meant to be read as an authoritative source of truth on Amazon culture and sales and so forth, but rather just relaying what they've heard. They're more plugged into the Amazon ru
99.
by ericjang 6y ago
Fair enough. I could imagine that if the work were billed by the hour or said research firm hired multiple people it would be easy for costs of the work to run up to $75k - it's within O(20k). I'm not qualified to price these thou
100.
by ericjang 6y ago
Rather than the exploiter setting an arbitrary price (which would be closer to blackmail), I think parent comment was saying that the fair market value of disclosing such a bug was worth closer to $75k given the unique skill set required. S
101.
by ericjang 6y ago
Interesting. I assume other multinational apps run into this problem as well. Is Apple considering support for apps with country-dependent payloads in the future? Otherwise seems like a lot of bloat across all major apps.
102.
by ericjang 6y ago
You've asserted the existence of a physics based answer, but you haven't answered what the "physics-based" criteria of an egg being "cooked" is. I assert that a straightforward short definition does not exist.
103.
by ericjang 6y ago
Are you sure that making a bowl of cereal is that much easier? Are you sure that the act of making cereal doesn't involve some physical impediment that is trivial for humans but actually quite difficult for any non-humanoid robot morph
104.
by ericjang 6y ago
at what point would you consider an egg "cooked"? how would you define this as a measurable physical quantity?
105.
by ericjang 6y ago
Often defining a function that classifies whether a task is accomplished or not requires a fair bit of semantic precision as well...
106.
by ericjang 6y ago
Something like this? https://www.theburnin.com/technology/toyota-ceiling-mounted-... Thanks for enumerating the use cases. They are fun to think about. I wanted to dive into the first application you mentioned. As ment
107.
by ericjang 6y ago
Humor me - what would you use such an arm for? What does the design look like? How many times a day would you use it? A lot of time in homemaking (a full time job for some people) is spent cleaning, unpacking things, and packing things away
108.
by ericjang 6y ago
This happens all the time at Google, though the mundane details of internal tech churn are not often publicized. Source: am Google employee.
109.
by ericjang 6y ago
I think that's probably the way to go if a business wants to tackle the problem of a useful home robot in the next decade. I totally agree with your point about home developers being forward-looking and designing homes with robots in m
110.
Software and Hardware for General Robots(blog.evjang.com)
42 points by ericjang 6y ago|32 comments
111.
by ericjang 6y ago
are you sure that would happen? Make no mistake, I get that Western powers do have cultural and economic agendas (McDonald republics and such). But if you look at Japan and South Korea, they didn't really have significant protests or a
112.
by ericjang 6y ago
This poem is dismissive of the progress humanity has made in computing and information technology. Tools like Pandas and Arduino and computer graphics were not mature in 1970s, and in order for data science, hobbyist electronics, or modern
113.
by ericjang 6y ago
Thank you, dang! Having to read so many pages of ceaseless bickering and techno pedantry must take a toll on the psyche. I appreciate all that you do!
114.
by ericjang 6y ago
It is widely believed by neuroscientists that spike timing in animal brains (whether by aggregate rate or by individual spikes) does the majority of information encoding.
115.
How many AGIs can dance on the head of a pin?(togelius.blogspot.com)
1 points by ericjang 6y ago|0 comments
116.
by ericjang 6y ago
Curious, why does this specific arxiv submission keep on cropping up popularly over the years? Is this common for any top post?
117.
by ericjang 6y ago
The report, while a fun thought exercise for any aspiring academic, is not a novel insight, and is obviously untrue for most definitions of market efficiency. If we are going to play the academic one-upmanship game, a more general result th
118.
by ericjang 6y ago
thanks, I was able to figure it out.
119.
by ericjang 6y ago
If you highlight the name, the highlighted text is {{14}}. I don't think the lead investor's name is necessarily 6 characters. FWIW, I was deeply curious about who the investor was and spent some time investigating who lead the fu
120.
by ericjang 6y ago
I searched for the "famous paper" in question and could not find it - can you point us to a link? So far I find two of Dijkstra's talks/writing on anthropomorphism [1] https://www.cs.utexas.edu/users&#x2F

More ›