Nor does he say he was the first to create them. The bro has posted hundreds or more pictures functions, many of which are well known... But well-colored and cropped into interesting locales, and described well. Including a good article that he wrote in a 1990 magazine.
Please think about how this can be set up to track upstream Kicad. Hopefully you have a small adaptation layer and some clever build, not wholesale changes to the code.
I teach HS classes where a lot of students end up doing some pretty serious PCB layout. We already benefit from MCAD that runs in the browser-- KiCad running in the browser would be a major boon to reduce setup requirements for student computers.
Thinking about how this could play with revision control and KiCad documents (footprint/symbol library, schematic, board, production gerbers) in Git repositories would be really helpful for my use case, too.
I think it's also fascinating that you can be so judgy, and felt the need to reply to yourself to announce this to the room instead of to me. Really, what's the point of that? Does it help you feel superior?
I've found the information density and accuracy of tags to be poor. I tend to ignore navigation infrastructure like this.
Note that I did not say it was not possible to determine it was satire promptly; as I said, I had ascribed it a high probability of it being satire early on. I also didn't need to announce that I had been uncertain. Telling the truth like this comes with psychological safety.
And psychological safety, in turn, depends upon people not coming out of the woodwork to congratulate themselves on being smarter than you because you were not quite so quick on the draw as them. I feel like this entire subthread just exists for this purpose, starting with Xandrius's "I guess it's time to rethink how sharp the HN folks truly are." That's what I sought to counter by admitting my uncertainty.
I don't understand how you refer to this as "Cerebras-style". Cerebras is wafer-scale and unique. Jalapeno is an inference-optimized conventional chip.
Yah well, I don't read all front-matter like that. Most of the time it's noise. Count it in the stuff that becomes cognitively invisible, like banner ads.
I'm not saying that food banks never pay full price for food-- I'm sure it happens some. But for the most part food banks pay way below market rate.
But, for example, if you make a >$100 donation to Second Harvest Food Bank (i.e. so that transactional costs are small), each 50 cents becomes a distributed meal. Note that they collect a few additional cents from the partner charity that is distributing the meal.
OTOH, the school I work at does a Thanksgiving meal drive where students buy food at retail and bring it in. It is definitely less efficient than giving funds somewhere like SHFB, but I think it's an important tangible experience especially for younger kids to give something they recognize as food to the less fortunate.
A really big part of why I did this is because I wanted to help. I make basically nothing. There are many more personally lucrative things that I could do that help society and people less.
But there's millions of ways that I could help. I didn't maximize my impact, I don't think. I did one that was a confluence between altruism, feeling good to me, conferring other advantages, etc. In other words, altruism was not the sole factor in my decision -- just a very large one.
The high school kid who volunteers at a homeless shelter and hopes it will help their college app is likely doing it both out of altruism and self-interest.
(Actually, the person who helps people because it feels good is also acting out of self-interest).
Given many ways to be altruistic, people will usually pick the ones that coincide more with their self interest. And in turn, self interest can warp a lot of the outcomes, even if people are trying to help.
It's probably a few milliseconds difference in timing between when the (short) exposure is taken, resulting in different brightness of lights powered by 60Hz AC.
> That is not in any way related to Fable (visibly) being switched to a less strong model if you’re trying to discuss certain topics.
We're talking about it being invisibly moved to a weaker model if it looks like you're distilling (which is best detected through something that is at least partially a reputational / account metric).
Now, Anthropic stepped away from this, but it highlights one more kind of systemic risk you're exposed to when you're not running the model yourself.
NAT was not designed as a security boundary. Sure, it may block some kinds of incoming traffic accidentally and as a side-effect disrupt some attacks.
But why would you rather have an always-broken network that might block attackers instead of a deliberate "deny incoming" rule that does exactly what you want -- and that you can punch holes in if desired?
Instead we have apps circumventing this accidental barrier with STUN, uPNP, etc with little/no oversight and we also regularly encounter brokenness.
We're finally getting there in the US, though. Top ASNs are >75% IPv6 capable.
It's Optimum Communications and Frontier (my provider) that are really holding the numbers down at ~15% each. The latter is improving very slowly, but not a lot of evidence of change in the former.
And, again, the economics. The hypothetical Polish project cited in the paper showed a breakeven in 6.1 years... if one ignores time value of money and assumes it is delivered for the estimated cost. The annuity formula tells me that at a reasonable 10% discount rate, payback would be in 10 years. Modest overruns in cost from this naive estimation push that to never.
And again this is a pretty idealized case-- very nearby housing, optimistic estimated costs.
Other projects analyzed in your paper had "raw" paybacks of 15 years and 17 years-- AKA never in actuality.
When we are talking about protecting the environment, efficiency and return on capital is important. You'd be better dumping those projects into generating more green energy instead of trying to reclaim the DC energy. This is why:
> > Even much hotter low grade heat usually goes unused. You usually need a perfect confluence of a warmer source, very close need for building heat, and a willingness to pay more for environmental friendliness for it all to work out.
You gotta include 50k+ in power plus other expenses. Still looks like an ok return on capital, but you are betting heavily that cost of computing doesn’t fall much.
If token prices fall a bunch then it may not even be worth leaving on, depending on your facility’s relative power, cooling costs.
If we push far into oversupply eventually a bunch of firms building this infrastructure are going to lose out.
I am not making an argument about limits. I just expect some degree of diminishing returns.
A related argument is speed of intelligence vs capability at that speed. You can think of a three way trade off between latency, cost, and capability that is unlikely to be linear in any dimension and that changes in steps as technology or biology evolves.
Ultimately relating to the properties of the computing substrate and almost certainly bounded by some kind of thermodynamic limits that present systems do not approach.
Do the math on pumping 40C water any distance in a loop. It is hard to make it worth it. The water cools off and the pumping takes significant energy, plus the pump and all the piping cost something.
Then using a heat pump on it makes it even worse.
Low grade heat source is not a term I made up- it refers to heat sources under 100C.
Even much hotter low grade heat usually goes unused. You usually need a perfect confluence of a warmer source, very close need for building heat, and a willingness to pay more for environmental friendliness for it all to work out.
https://paulbourke.net/fractals/