sgillen

70 ms·

481.
by sgillen 8y ago
not a paper that came out this year but Emergence of Locomotion Behaviors in Rich Environments by Google deepmind: ( https://arxiv.org/abs/1707.02286 ) made me realize that although RL is brittle and slow it can still be
482.
by sgillen 8y ago
Can also put two distance sensors on and fet the angle using some trig. And when you’re ready to get a little more serious you can get an imu and gps. Maybe even get a more beefy computer and add a camera!
483.
by sgillen 8y ago
Don’t have too many recomentations for the printer, but you may want to look into printing services rather than buying one. If you live near a university they may offer this service for relatively cheap, same for local maker spaces. I’m sur
484.
by sgillen 8y ago
You’re basically describing function or program approximation. This is the central problem that machine learning is trying to solve. A lot of people are talking about this.
485.
by sgillen 8y ago
I’m not sure it would be as significant as you are thinking. We already have really really cheap low power computing. The von bottleneck is also something people have been working on for a long time, I mean basically the entire memory hiera
486.
by sgillen 8y ago
I definitely do this too. But for some acedemic papers it might take me (a first year PHD without much research exp) a few hours to actually understand everything the author is doing. For these cases I definitely try and skim the article to
487.
by sgillen 8y ago
Yup, but I’ve had some issues pop up using that before. I’m a longtime emacs user and this seems to be a common theme for me. I try some feature in emacs, it works great but breaks on some edge case or something, and then I have to find a w
488.
by sgillen 8y ago
Also be careful, if you add something that contributes nothing you may receive downvotes.
489.
by sgillen 8y ago
The monster group
490.
by sgillen 8y ago
Yeah I here you there I’m sure that happens all the time. I’ve mostly done R&D work in my career so I guess my real question to you is what are all the problems that happen when zip ties are used in production.
491.
by sgillen 8y ago
What’s so evil about zip ties? They probably shouldn’t end up in a real deliverable product but for prototypes and proof of concepts I’ve always found them to be very useful.
492.
by sgillen 8y ago
I wish they would give some numbers and methods here.
493.
by sgillen 8y ago
What exactly do you mean by computational challenges? That seems like the exact kind of thing a bot would be much better at compared to a human.
494.
by sgillen 8y ago
Yeah I think people want their “robots” to be able to move around. I guess the washing machine is just that, a machine.
495.
by sgillen 8y ago
Is the flag always a message of “I think this should be taken down”. Maybe the person who flagged it was thinking “this has the potential to get toxic quickly, definitely something to monitor”... Just playing devils advocate here.
496.
by sgillen 8y ago
The other thing they are very good at is marketing, both to consumers and to potential employees. I think this is largely thanks to the cult of personality around Elon Musk. I don’t really mean that in a bad way either. If they can retain t
497.
by sgillen 8y ago
Also researchers who need fast approximate physical dynamics. For example the Mujoco simulator or Drake.
498.
by sgillen 8y ago
I have heard it called the most important algorithm of the 20th century. Not sure I buy that, I would say it's the most important algorithm in EE though.
499.
by sgillen 8y ago
Second Khan Academy as a good resource. The most important thing is that you practice doing math a lot, don't worry too much about reading the right book or anything.
500.
by sgillen 8y ago
Typically when I need to develop inside a VM I do my editing inside that VM as well. So my tools would be: 1. Ample memory and power 2. OpenBox (or equivalent lightweight DE) 3. Emacs (or equivalent “light weight” editor/terminal Also
501.
by sgillen 8y ago
The whole "cybernetics" field is pretty weird to me, I think the notation and lingo have sort of moved on (I could easily be wrong and biased without realizing). However the theorems you are talking about have useful, well studied
502.
by sgillen 8y ago
I think it's most accurate to say that Apple did it and ARM helped. Apple designed the chip, and ARM designed the platform and instructions for them to build with.
503.
by sgillen 8y ago
There are a few successful examples of graphical programming languages used by engineers and scientists everyday. Simulink and Labview are what I'm thinking of, there may be a few other similar ones I don't know about. I actually
504.
by sgillen 8y ago
Could also just be a circadian rythm thing. I’m sure there are many otherwise normal people who’s biological clock would prefer to sleep early and wake up at 4.
505.
by sgillen 8y ago
Things are not this simple. I think these questions have more complicated answers than you think.
506.
by sgillen 8y ago
>> Why are programming environments still so primitive? I think adding all the features the author is requesting and have it work seamlessly is much much harder than be seems to realize.
507.
by sgillen 8y ago
I like that idea too(and as people point out several companies are experimenting with this idea). It would also be interesting to set up "remote drivers" for the last mile. The truck itself can still be unmanned and someone drives
508.
by sgillen 8y ago
I do think that the realm of experimental particle physics and astrophysics have sort of fallen behind the theory. It will take some new technology or breakthrough I think before we see more big strides in these fundamental areas. As eigens
509.
by sgillen 8y ago
This is starting to happen I think. The economics are starting to make sense, for some companies, due to insane rent and cost of employees.
510.
by sgillen 8y ago
Yeah sorry I was referring to people. Although the US is a massive landmass too which I also forget sometimes.

More ›