stochastimus

Thanks for this, OP. This is why I come here.

> Go to Google announcement > Find “license” in page: no matches > Go to HN thread > Find “license” in page: 28 matches > Read a few sigh could have been exciting

As OP noticed likelihoods often do show up in a comparative context. In that context one is asking which thing or sequence is most likely to occur by chance relative to another, under an (over simplistic, sure) IID assumption. In practice, the ordering of such things is often (hand-waving, sure) robust enough that, given no other information than the marginals, it is useful. So I think OP almost answered his/her own question: they are often quite useful in a comparative context and with no additional information.

Congrats and thank you for this. If you’d asked me when I was a young h4x0r just starting out, what sorts of links I should find on a site called Hacker News, I’d have said something like this one. Looking to get lost in these issues during the holidays. You made my day!

The last paragraph may be your best option, if you’ve truly exhausted the possibility of a shared vision that can keep you all engaged. I would drill into why your co-founder is against it, but I imagine it could be that they don’t feel they can do it all over again without you. If this is the case, maybe you could help them find another cofounder. Regardless, there’s no sense in misleading others, or wasting your time or theirs, since that karma will follow you. At least, that’s my opinion, FWIW. Good luck!

> working less is good > blog by “Slack”

True or false, the jokes write themselves sometimes

Speaking anecdotally from my own experience, having multiple caregivers may just increase the probability that the child can tell that at least one of them actually wants to be there.

I got downvoted 2 points for asking this question LoL

Genuinely curious, HN folks - for those who downvoted this comment, can you share why you did so? I don’t personally mind sarcasm as long as it’s not aimed maliciously at an individual, and I read this comment as a sarcastic way of saying “how come American leadership in general doesn’t have a similar plan”. Do folks just downvote sarcasm as a rule?

I don’t think it’s necessary to invoke epistemic probability to explain why the original question is tricky. At least for me, it helps to notice that if the question were rephrased “the older child is a boy, what is the probability that the younger child is also?”, then the answer would be the intuitive 1/2. You get into this counter-intuitive situation with the question as asked because of the “at least” phrasing - the remaining possibilities in the contingency table can still apply to one, the other, or both children.

Funny I came here just to say the exact same thing.

“Well-being and how to set healthy boundaries in your life and in your work is a fundamental skill,” says Sharon Milgram

What is this “life” you speak of?

Agreed. Money buys choices, simple as. What you do with those choices, and how much meaning they bring to your life, is entirely up to you.

I love outrageous opinions like this, thanks for sharing it. It opens the mind to what’s possible, however much of it shakes out in the end. Progress comes from batting around thoughts like this.

If you replaced QM with general relativity, you would be correct. So, your broader point is valid.

Would love to see this result reproduced. Also a confounding factor might be what they talked about; would be interesting to see a follow-up that tried to break out by subject matter (someone filling time vs. conveying real knowledge vs. responding to the other person, for example).

This is actually a great idea for improving the reliability of answers from a language model. I have also had trouble with GPT3 hallucinating knowledge. One could automate asking the model for a reference and then checking if the reference exists, then changing the prompt and demanding another answer if it doesn’t, until it does. I will explore this method. Thanks again!

I’m just an 1800 rated patzer in blitz/bullet on LiChess, but I would say white’s knight sac demonstrates a certain comfort/familiarity/maturity with the game.

Pretty impressive what the Cerebras folks seem to have accomplished here, reportedly verified by Argonne National Labs. Interested in hardware folks’ take on the architecture as a realistic path forward for training larger and larger models.

Too bad the Maginot line is mentioned only in passing in the blog - the Maginot line really is a testament to the failure of central planning and bureaucracy to produce and execute on ideas that stand the tests of time and battle. A drastic (but mostly true) oversimplification is this: France, seeing Hitler’s military buildup, created an impenetrable fortress along the border and sucked the life out of the French economy to build, supply, and man it. One failure was the inability to see that German light armor could navigate the Ardennes forest and circumvent it. Another failure was that the massive gun turrets periodically embedded in the fortification were designed so that they couldn’t be rotated to point back toward France; as a result, they couldn’t be used to defend against an army which had circumvented the wall. So, in the end, the French had simply provided the Germans with a massive cache of ammunition, food, guns, and other supplies which were all stocked within the Maginot fortification. Here’s the wikipedia entry for anyone interested: https://en.m.wikipedia.org/wiki/Maginot_Line

Not sure why you got downvoted for asking a good faith question, but you’re thinking of carbonic acid, which is responsible for the fizz when the pressure lowers as you pop the top and it creates carbon dioxide.