felippee

16 ms·

91.
by felippee 9y ago
I'd disagree, there is more here than the no true Scotsman fallacy. It is equivalent to saying that any magic trick ceases to be magic once exposed as a trick. We don't have a good definition of I in AI therefore such arguments go
92.
(Artificial) Intelligence confuses the intelligent(blog.piekniewski.info)
2 points by felippee 9y ago|0 comments
93.
by felippee 9y ago
I'd say it is over-optimization, but I'd agree that this is not the biggest problem. What drives me nuts a lot more, is that the same legs of flight cost completely different amount of money depending where they originate. Try e.g
94.
The Complexity of Simplicity – Balancing on the Occam's Razor(blog.piekniewski.info)
2 points by felippee 10y ago|0 comments
95.
by felippee 10y ago
It is not entirely obvious that such extrapolations would succeed. GAN is not a new idea (new in the context of deep learning) but automatic generation of adversarial samples is a known and exploited subject (even e.g. in the field of evolu
96.
by felippee 10y ago
This style transfer idea, though eye candy and sometimes impressive, seems to be the core application of deep learning these days. François Chollet tweeted something like that two years ago about the prisma app (yes it was 2015). He back th
97.
by felippee 10y ago
Took them only like what - 10 years? Better late then never I suppose.
98.
Self Made Alternative to Apple Time Capsule(blog.piekniewski.info)
3 points by felippee 10y ago|0 comments
99.
by felippee 10y ago
Yes, that is exactly my point. Solving these particular situations is like typing all the possible Winograd Schema sentences into a chat bot. Certainly possible but does not solve a larger problem. In reality there will always be a differen
100.
by felippee 10y ago
Sure they do, agree with that. But word "equally" suggest the susceptibility is the same. I would actually emphasise the difference. It is much easier to fool a machine than a human, particularly if we have a copy of the machine a
101.
by felippee 10y ago
It depends on the context. Humans may make a better decision if there is something else providing the information. Perhaps the manhole cover sits at the curb triggering immediate attention. Perhaps there are cones, only in the wrong place b
102.
by felippee 10y ago
> Humans aren't really good at driving either, I'm not sure where did that come from. One fatality per 200 million miles not good? Seriously, there are millions of people travelling every day. And true, there are accidents, but
103.
by felippee 10y ago
What is the real problem then?
104.
by felippee 10y ago
Yup, there is a boatload of low level knowledge missing from any AI that we build. It is painfully visible with robots (and autonomous cars are robots as well), see DARPA challenge video which was then lead BTW by Gill Pratt: https:/&
105.
by felippee 10y ago
> (Doesn't solve the prank problem, but humans are equally vulnerable to a targeted prank anyway.) I would hesitate using the word "equally". People are actually quite robust. Particularly the second human in row will cert
106.
by felippee 10y ago
Right, there are cases where such a moving stop sign would actually be a real stop sign. The problem is, we try to program in the complexity that cannot be anticipated.
107.
by felippee 10y ago
How would it show up different than a regular pot hole? What if its almost filled with water and the only thing giving away its true nature is its roundness? The point is to explore the limits. If these aren't the limits, then they are
108.
by felippee 10y ago
> This doesn't seem like it would be that difficult to detect. And also, I would expect self-driving cars to have an emergency stop button. That it if the passenger pays enough attention, which they won't. Also note: each one o
109.
by felippee 10y ago
Hey, I just finished another one. A school bus with a swinging open stop sign in the middle of the freeway. Anyway, just a food for though, autonomy really requires a lot of "intelligence" and our technology is not quite there to
110.
by felippee 10y ago
I took the liberty to draw a few situations which AI would have likely no clue what to do, even if it detected the danger. I'll be adding some more as every day on the freeway brings me new ideas (in fact I'm editing one right now
111.
by felippee 10y ago
Holy crap. In some sense it is really surprising there's only been one fatality. Just as well tomorrow there might be 10 more if people really do such stuff...
112.
by felippee 10y ago
Ironically this whole deep traffic exercise of theirs has as much to do with driving an autonomous car as the game chess has with playing soccer.
113.
by felippee 10y ago
I love that part when he asks a question "is driving more like chess or more like soccer". I understand the question is rhetorical? But generally seems like a good intro with good hype busting cautionary statements, particularly t
114.
by felippee 10y ago
Not a single truck driver has yet lost his job to ML. Just a side note.
115.
by felippee 10y ago
Nevertheless he's got the chance to become the biggest comedian :) I think Hollywood would work out better for that...
116.
by felippee 10y ago
He was the biggest hacker, tried to be an entrepreneur, now he is working hard to became the biggest comedian of Silicon Valley. Seems like the comedian role might actually work out.
117.
by felippee 10y ago
Its rather the opposite. It is unsupervised initially, just learns to predict its input. Note there is confusion: the unit itself is using supervised (by the future signal) but all in all nothing needs to be labeled cause the reality just u
118.
by felippee 10y ago
Cars have driven billions on miles on cruise control, but that does not mean that cruise control can autonomously drive a car. Planes fly billions of miles on autopilot and yet we keep the pilots. What matters is not the number of miles dri
119.
by felippee 10y ago
You assume it has already happened. Let me remind you: all the self driving stuff right not is in tests. Still mostly in favourable driving conditions. The whole self driving thing may still not happen, if it turns out (after a few accident
120.
by felippee 10y ago
Why would you want that? Conv net is not some magic. It's just a crude way to reduce dimensionality by loosing spatial location. For some things it works, for some it doesn't. I think the shift in thinking should rather be: instea

More ›