glalonde

Do you also stop in your tracks and write blog posts when you see suicide hotline signs on the GGB?

This is genuinely insightful.

Probably many reasons but here’s one demo: https://en.wikipedia.org/wiki/Braess%27s_paradox

Every actor acting rationally in their own interest can get to a worse equilibrium than if they were coordinated(in this case, to completely ignore the new edge). There are many other examples of this in game theory, you should look it up.

"extrapolation" literally implies outside the extents of current knowledge.

*Overstated

"Optimize the kernel (in KernelBuilder.build_kernel) as much as possible in the available time, as measured by test_kernel_cycles on a frozen separate copy of the simulator." from perf_takehome.py

which makes sense in a portfolio, but a company only has one CEO in their basket. But I suppose if CEO elections represent stock holders with many companies, then perhaps it may still be optimized like a portfolio with multiple CEOs.

Is that true and is that sustainable? My understanding has been that only the relatively low level libraries/kernels are written in cuda and all the magical algorithms are in python using various ML libaries. It's like how intel BLAS isn't much of a moat -- there are several open source implementations and you can mix and match.

How is CUDA so sticky when most ML devs aren't writing CUDA but something several layers of abstraction above it? Why can't intel, AMD, google w/e come along and write an adapter for that lowest level to TF, pytorch or whatever is the framework of the day?

it had a gps

Atlas is also electric. An electric motor tied to a gear pump presumably.

have to appreciate the irony in worrying about the bitrate of simulated lo-fi crackling

internal internal combustion engine engine

Presumably early waymo. Ian probably worked for bot & dolly which was bought by Google.

The weird thing about this article is that it's only comparing annotation performance, which is important, but not what you should ultimately care about. If you trained a visual model using annotated lidar for ground truth, then you might expect better performance from the model than from human annotations of the image alone, and certainly better than a model trained on those annotations.

Yeah I'm kinda surprised by this. You would think there would be some industry support for emulating graphics/gpgpu pipelines to make cloud CI cheaper and simpler, but maybe it's just a much harder problem than I imagine -- to the point where mocking opengl or using dedicated hardware makes more sense.

Probably no one is arguing that SV programmers have particularly hard lives in the larger picture, but if you get in the habit of punishing collusion insufficiently because, 'eh, I've got enough flex in my budget', you're being a little selfish. That's like boomers supporting an unsustainable social security system: 'Look, it works well enough for me today. And I'm not going to be in the system when it turns sour.'

The article links to that very post.

now they just need to bolt it onto Glass or some other HUD. IRL subtitles. Could put a mic array on the frame too, and then get an indication of where the speaker is so you can be alerted even if you aren't facing the person.

The world at large doesn't drive SUVs. People in France don't complain about F150s and SUVs because they don't exist at nearly the frequency. They get by with twingos, why can't suburban California? Even the farms only rarely have trucks and even then they're taco sized, not F-series.

>I'm sure sociologists have a name for this, but to me it represents people who have a big hole to fill regarding self-worth.

Fundamental attribution bias? "Everything bad that happens to me is because of external factors beyond my control. Everything bad that happens to you is your fault." And the opposite: "I'm skilled, you're just lucky."

Nothing teaches this better than a game of Dota2.

Why isn't it a distance squared relation?

newly translated. He wrote it 3 years before TBP

Anyone know how this would compare to IPOPT for constrained, sparse nonlinear optimization?

Everyone was cartoonish, even the good guys. It was like a small army of glimmering MD PhD's with perfect physiques vs. the lazy bums.

It's a work of fantasy, why does anyone point to it and say "This is what I base my beliefs on."? That would be like using LotR as your moral compass. Except with worse plot holes like gross misrepresentations of how small mountain village economies could possibly function.

Exacerbated. I think, is the word you were going for.

>I'm quitting fb for real this time.

>Oh, and look at all these thousands devoted followers I'm walking away from

>I just don't want to always put up a front of my existence for others to see

>I don't want to live for how others perceive me.

>Better share this transcendental experience on my blog.

k.

What about pose estimation? e.g. Given a well defined coordinate system, like the origin is the nose on a face, determine the pose of the face. Is this still best done with classic optimization formulations like ransac/ICP and a supplied model, or have these been bested by learned models somehow?

The human would probably be able to blink/cancel if they were expecting it and therefore able to focus on a single thing prior to an event. The bot can focus on everything simultaneously and doesn't really need to expect anything. It gets a signal, it responds within 200ms, no problem. You could program that analytically.

So I would say the superhuman-ness isn't in the number of actions taken, or in the response delay, but in the massive attention bandwidth. I believe they've attempted to even the playing field in the first two, which are easily quantifiable, but I don't know about the latter.

>Native binaries

Is this common practice? I've looked for guides on this without much success, but is it basically, just static link your binary and then write a thin js wrapper to invoke it and hope it is the right architecture? Anyone know how multithreading would work? Is it more cost effective to multithread or just spawn separate lambdas?

hold up, how does that figure? I thought you could only get up to the amount invested back, which is why it is a really not great idea. Investing $3k now might net you up to $3k if the stock hits 0, but it could triple in which case you lose $6k.

Bounded upside, infinite downside. I'm sure it's more complicated but what am I missing?