jowday

13 ms·

91.
by jowday 5y ago
> Personally, the idea of "a wire in the road" by which I mean something like cats-eyes in the road but with RFID tags placed every 40 feet, or some other common means maintained by the DOT that all cars can use to sense where
92.
by jowday 5y ago
I work in the AV space, so I guess you’ll just have to trust me. I assume Cruise internals look very similar to the internals at other AV companies, which I think is a safe assumption. >The sensors and automation decisionmaking is going
93.
by jowday 5y ago
If anything, I got the impression Dan had a pretty good handle on the technical aspects of Cruise and the path forward for the company, and that the GM CEO fired him because she didn't like what she was hearing from him. Dan wanted to
94.
by jowday 5y ago
>You do need to prep for an interruption of utilities, but you should do that anyway for hurricane/earthquake/etc preparedness. If you've got ~3 days of water and ~2 weeks of food and are outside a blast radius, there&#x27
95.
by jowday 5y ago
That would accomplish nothing other than making Russia even more of an international pariah than it already is. They don't need to demonstrate that they can use nuclear weapons if they want to - the world already knows that.
96.
by jowday 5y ago
This post was at the top of HN a few days ago and does a great job explaining, IMO. https://acoup.blog/2022/02/25/miscellanea-understanding-the-...
97.
by jowday 5y ago
I think the reality is that where exactly you were when the nukes started flying wouldn't matter - you'd be very likely to die of starvation or other causes in the days and weeks afterwords unless you were already prepared with a
98.
by jowday 5y ago
In practice, for digital production backgrounds like the volume they actually set the part of the screens that intersect the camera frustum to a green screen in case they want to change anything in post. The main benefit you get from digita
99.
by jowday 5y ago
Discussions of “Consciousness” in the context of ML or AI research always seem to devolve into navel-gazing futurist pseudointellectualism. I don’t think it’s possible to have a meaningful conversation about something as ill defined as con
100.
by jowday 5y ago
This is definitely the sense I get when I hear him discuss ML.
101.
by jowday 5y ago
There’s tons of videos of Waymo’s service in Chandler.
102.
by jowday 5y ago
I don't think this is true - most of the examples people cite look more like they used underpaid contractors to manually recreate storefronts and billboards with shoddy clip art and plain fonts. ML-based upscaling wouldn't replace
103.
by jowday 5y ago
This has been a thing for a long time, just not with generative ML models - most solutions use rule-based procedural generation. A large fraction of trees in games have been generated this way since the mid 2000s - see https://en
104.
by jowday 5y ago
> Further that process can take place from a photo of the sign uploaded by the car to then be encoded by the rules. The general case is stopping and having a remote driver slowly tell the car what to do. So you’re now agreeing that you n
105.
by jowday 5y ago
I don’t think I’m moving the goalposts - we were discussing whether autonomous driving (which I take to mean L4-L5 driving without the need for a human in the loop) is possible without geofences or HD maps. “Edge cases in some random small
106.
by jowday 5y ago
And now we’re already making concessions about the car’s abilities. There are 10 MPH speed limit signs on Market Street in SF that specify in incredibly small text “when behind trolleys”. Assuming we take your approach, the car will just al
107.
by jowday 5y ago
If you really wanted to make this argument, you wouldn’t even want to bother inferring depth, since that’s not what humans do, not directly at least. If you’re actually trying to obtain a depth map as part of your pipeline, LIDAR (or LiDAR
108.
by jowday 5y ago
Except Tesla isn’t inferring a 3D map from stereo vision either, at least not outside of the front facing cameras - they’re using monocular depth prediction.
109.
by jowday 5y ago
> Cars really don’t need to identify what something is just it’s location and movement which is a vastly easier problem. A trash can rolling down the street can be treated just like an oil drum doing the same thing etc. You’d think that,
110.
by jowday 5y ago
> I just think that if the software is unable to make decisions based on visual data alone without up to date high resolution maps it'll never achieve true FSD in the general case (not geo locked). You'll end up trapped in a lo
111.
by jowday 5y ago
I’ve seen their LIDAR cars in SF too - if I had to guess they’re gathering ground truth data to train monocular depth models on. And even really naive integrations of LIDAR will show big improvements over camera only. You can do something a
112.
by jowday 5y ago
Humans don’t regress per-pixel depth or use convolutions and region proposals to draw bounding boxes around objects. They don’t function on models with fixed weights trained by backprop either. The idea that “vision only” somehow more close
113.
by jowday 5y ago
The truth of it is that it’s just not possible (with currently existing technology/ML architectures) to create a truly autonomous taxi without HD maps. Everyone in the robotaxi industry knows this - even Tesla builds HD maps, they just
114.
by jowday 5y ago
They moved to pure vision because they were constrained by radar suppliers because of supply chain issues and the chip shortage, not because of any ML progress - they were actually investigating using a higher resolution imaging radar befor
115.
by jowday 5y ago
What makes you say Tesla is imitating humans? Their motion planning is all traditional robotics logic, not anything learned.
116.
by jowday 5y ago
Author spends a lot of time discussing the social contract and then laments an elected official calling out a billionaire for disobeying a public health mandate during a world historic pandemic to reopen a factory.
117.
by jowday 5y ago
I’m not sure they will, at least not with the research in the state it is presently. Researchers are interested in vision transformers because they’re competitive with CNNs if you give them enough training data - they don’t drastically outp
118.
by jowday 5y ago
Outside of research transformers are rarely used for computer vision problems and CNNs remain the go to architecture. And you actually need to do some hacks to get transformers to work with computer vision at a meaningful scale (splitting i
119.
by jowday 5y ago
Gravity’s Rainbow. Too much to explain.
120.
by jowday 5y ago
Someone asked this question to Leonard Susskind once at a Q&A and he said the chances are basically zero. When the guy kept pressing on the idea Susskind got kind of exasperated and seemed like he was on the verge of calling the idea wo

More ›