wrsh07

20 ms·

331.
by wrsh07 2y ago
Ignoring the ethics of Mr Beast, he is producing real videos at an incredibly high volume and they consistently do numbers. None of those videos is easy to make. Sure, it's maybe not great to be so impressed by logistics or supply chai
332.
by wrsh07 2y ago
Right, a challenge with an artifact like this is that the writing is good. And it's a lengthy read. The early commenters almost by definition haven't read it and can only comment about their opinions on the creative output of Mr B
333.
by wrsh07 2y ago
From the excerpted section I totally see why that's the question, and yet that's not quite how I parsed it in the essay. The author makes the point that readers are more likely to reject the author's imposed world/struct
334.
by wrsh07 2y ago
Nice article. Two notes: First, some of these suggestions will make it harder to introduce bugs when updating the code. That's good! Particularly tricky is when somebody splits up identifiers or function names. These types of things of
335.
by wrsh07 2y ago
Makes me wonder when an update to the world models paper comes out where they drop in diffusion models: https://worldmodels.github.io/
336.
by wrsh07 2y ago
You don't need back propagation to learn This is an incredibly complex hypothesis that doesn't really seem justified by the evidence
337.
by wrsh07 2y ago
100%!! Ray-Bans are iconic, so whatever comes next will be extremely dorky in comparison. The proof will be in the pudding: how useful is it to chat with llama? (I'm extremely skeptical but want it to be cool)
338.
by wrsh07 2y ago
One thing that frustrates me with this comment is that you're measuring their state today as if that's the end state. Is waymo a failure? It's not deployed across the US yet, and it's been over a decade etc No, they&#x27
339.
by wrsh07 2y ago
What is the goal of quest? Is it to get mau? Is it sales and marketing for eng (come work at _, where we have fun projects like these Glasses - sounds familiar), is it to develop and flex hardware muscles? Is it to make sure Meta is competi
340.
by wrsh07 2y ago
> This so called “sweet spot” comes at the cost of billions of dollars in operating losses and layoffs at Reality Labs year after year Yes, it has cost them billions of dollars (ongoing) to develop their AR/VR tech. You can definite
341.
by wrsh07 2y ago
Honestly, this is a shame, because I think a lot of the vision pro's flaws are Apple problems. That said, meta seems to have found a sweet spot in price/performance, so maybe in a few generations we will have something with the qu
342.
by wrsh07 2y ago
I think the point of wolfram's essay is that you don't need the base unit of computation to be a dot product
343.
by wrsh07 2y ago
But to your point - that is how I feel about graph nns vs transformers or the fully connected set (GPUs are so good at transformers and fully connected nns, even if there is a structure that makes sense we don't have the hardware to ha
344.
by wrsh07 2y ago
Sort of, yes. But if the existing thing were "the cheapest", quantization wouldn't exist. It depends on what your constraint is! So if you're memory constrained (or don't have a GPU), a bunch of 1 bit atoms with ope
345.
by wrsh07 2y ago
Because of the computational simplicity, I think there's a possibility that we will discover very cheap machine learning techniques that are discrete like this. I think this is novel (I've seen BNN https://arxiv.org&#x2
346.
by wrsh07 2y ago
This is truly horrifying and also illustrates a massive incentive problem. The hospital probably made money through its negligence (until the inevitable lawsuit)
347.
by wrsh07 2y ago
One of the most exciting things not really mentioned elsewhere in the thread (as far as I see!) is that you can make synthetic hydro carbons wherever you want. If you wanted to have a refueling airport in the middle of the ocean, you could
348.
by wrsh07 2y ago
"will likely fail" is a surprising claim when it's not clear to me that vertical farming won't still have a chance to succeed For vertical farming, the question at the limit shouldn't be comparing solar and a farmer
349.
by wrsh07 2y ago
When you're projecting future costs of solar, you should be using very optimistic estimates, right?
350.
by wrsh07 2y ago
I've met a few, but they tend to be focused on their niches.
351.
by wrsh07 2y ago
Agreed - the essay is decent and asks an interesting question (how do you find the right tail of venture returns?) In normal tech hiring (eg Google), they're not necessarily looking for high variance individuals. They're often loo
352.
by wrsh07 2y ago
The point about self-made: if you already have money, you probably won't [be willing to] take enough risks in order to become a top 200 rich person The whole article is trying to ask: if you want to capture the far right tail of founde
353.
by wrsh07 2y ago
In that video, Nigel pronounces www as "world wide web" and I have not heard that in decades. It's so much better than actually pronouncing the letters, I can't believe we ever stopped using it
354.
by wrsh07 2y ago
I also found Ikenfell to be really delightful!
355.
by wrsh07 2y ago
I mostly agree with your take. Mobile is particularly challenging. I'm optimistic about a hardware only startup succeeding (Framework but for phones), but building your own hardware and your own OS is much harder, I think. This doesn&#
356.
by wrsh07 2y ago
Can I encourage you to start the android company you want to see? I'm confident you can get funding
357.
by wrsh07 2y ago
If you're talking about America's semiconductor problem and not mentioning Intel even once, your view is horribly horribly slanted
358.
by wrsh07 2y ago
Hard. Naively, run the election, use real vote counts, claim that you're the one that got more. Ok but now we need to fake this on a local level. But everybody knows people in this district are more Party A and in that are more Party B
359.
by wrsh07 2y ago
Of course (having learned most of my cpp at Google) I agree with you. However, the language default is incorrect (you have to opt out of implicit conversions rather than opt in) which is why I phrased it like that
360.
by wrsh07 2y ago
Ah there's some context missing. As user dataflow notes, if you have a convention of single arg constructors being explicit (which Google does: https://google.github.io/styleguide/cppguide.html#Implicit_C... ), the

More ›