naasking

60 ms·

961.
by naasking 2y ago
I don't see a statement of his opinion on fluoridated toothpaste, which is the real test.
962.
by naasking 2y ago
> The IQ link is very heavily lacking in evidence. Not really: https://jamanetwork.com/journals/jamapediatrics/fullarticle/...
963.
by naasking 2y ago
Why is that missing context? I don't think anybody who is against fluoridated tap water rejects the benefits of fluoride, they just think the harms of adding it to tap water outweigh the benefits.
964.
by naasking 2y ago
> Except it's not—fluoride in the drinking water concentrations is proven safe and it doesn't affect brains. You mean like this: Fluoride Exposure and Children’s IQ Scores A Systematic Review and Meta-Analysis, https:/&#x
965.
by naasking 2y ago
I didn't say they were equivalent, I said they were both sound in that context. By "doesn't really matter", I meant in the sense that I had just described, eg. that it "cannot go wrong".
966.
by naasking 2y ago
> The language can just have an implicit conversion rule for convenience By "it is ok pass a String where we expect a String?", I think he means that no such program can go wrong, therefore this is sound. Whether you do this vi
967.
by naasking 2y ago
Yes, but they were describing an implicit conversion going the other way around, eg. String <: String?.
968.
by naasking 2y ago
> And people just sit around, unimpressed, and complain that ... what ... it isn't a perfect superintelligence that understands everything perfectly Hossenfelder is a scientist. There's a certain level of rigour that she needs
969.
by naasking 2y ago
> And no users which are facing a LLM today have been trained on next token prediction when they were babies. That's conjecture actually, see predictive coding. Note that "tokens" don't have to be language tokens.
970.
by naasking 2y ago
> It's a worldview that necessitates only teaching what leads to a job. It is anti beauty and anti human. The question is always who is footing the bill while you find yourself? Education that leads to productive vocations can argua
971.
by naasking 2y ago
The plain English reading of "He used [X/Twitter] to help get Donald Trump elected" is quite clearly "X/Twitter was an integral part of getting Trump elected". There is zero evidence that this is the case. >
972.
by naasking 2y ago
> It is not well established at this point. I get the impression you're reading a textbook on economics, and not how it works in the real world. You're writing this from the comfort of your home where you're enjoying the f
973.
by naasking 2y ago
1. Even if that were true, why is your "set of people" representative of such a large fraction of the US population that it would sway an election? 2. Even if it were somehow representative, is it more plausible that all of these
974.
by naasking 2y ago
> He used it to help get Donald Trump elected The idea that X had any influence on the election is laughable.
975.
by naasking 2y ago
I found Qwen never completed answering my standard coding task that I ask to check a model. Claude did great, DeepSeek R1 did well.
976.
by naasking 2y ago
Private goods and club goods are provided by markets because incentives are already aligned. Public goods and common goods need artificial markets created by regulatory frameworks.
977.
by naasking 2y ago
> If suddenly an endeavour is not profitable any more, or is simply less profitable than an alternative, what happens? They pivot or they slowly die.
978.
by naasking 2y ago
And companies have to compete for customers on cost, which yields an equilibrium that works out well for customers on most goods and services. This is well established at this point.
979.
by naasking 2y ago
Maybe not that irrelevant, since another desktop and mobile operating system used around the world is also based on Mach.
980.
by naasking 2y ago
> If solar/wind producers had to guarantee power 24x7 at fixed rates the cost would vastly exceed natural gas costs. Possibly, but neither usage nor generation is 24x7 at fixed cost, so what would be the point of that requirement?
981.
by naasking 2y ago
There is nothing wrong with markets solving basic human needs as long as the incentives are properly aligned.
982.
by naasking 2y ago
> When seL4’s proof was published, people regarded seL4 to be flawless software No, that's not true. The framing was always that it surpassed an incredibly high safety bar that no other kernel had surmounted, far beyond even securit
983.
by naasking 2y ago
> but you would not have made anything more secure. This is does not follow. Why do you think a hypervisor with no verification to a spec is necessarily more secure than a hypervisor with a verification to a spec? Can you somehow quant
984.
by naasking 2y ago
"Safe enough" means "actually unsafe".
985.
by naasking 2y ago
seL4 is used in a lot of cellular phone firmwares I believe.
986.
by naasking 2y ago
What? You'd put more faith in a kernel that has no formal verification of its conformance to a secure spec and has orders of magnitude more code in it that can go wrong? What is the basis for this confidence?
987.
by naasking 2y ago
> If you build a bridge to a specification and the bridge collapses, does saying that the specification was wrong in any way lessen the reality that the bridge failed? You framed this as a loss of confidence in the techniques that prove
988.
by naasking 2y ago
> If I have to re-architect the entire program every Monday, that's not going to make leadership very happy. Good thing FP doesn't require that then.
989.
by naasking 2y ago
That's arguably one of the only downsides of zero-based, and can be handled easily with negative indexing. Basically all indexing arithmetic is easier with zero-based.
990.
by naasking 2y ago
Index is ambiguous. I think "Offset" and "Itemized".

More ›