Would you feel differently if your child’s murerer was on trial?
Not trying to be gross, just trying to shine light on the idea that judges are in a difficult situation. They have to consider fairness for all parties involved.
That question is one of the pit falls we have to avoid to preserve any chance of avoiding a police surveillance state. "What would you allow if you were maximally emotionally impacted by a crime" is basically a blank check for 100% surveillance because inevitably any restriction on methods will eventually allow some crime to go unsolved.
Exactly this. At worst, you need to at least taper the emotionally charged question with the counter-point; "Would you feel differently if your child was falsely accused of murder because of genetic comtamination?"
Yep, to see how badly things can go when 'scientific' evidence goes unquestioned look at the mess that is stuff like bite pattern analysis which gets presented as scientific but has really bad actual statistics.
Justice can't just be about catching people it has to also weigh the cost of false positives and the effect that the methods have on society.
It's a constant issue with things like TOR, encryption, and alt currencies like BTC. They can do a lot of good for repressed people but they're also inevitably going to be used by people most people would agree are awful people to protect themselves from justice. If you're developing those tools you have to be ok with that just like we have to decide as a people where we're ok with the balance between preventing crime/catching criminals and personal liberty and privacy.
What if you were accused of a child's murder that you didn't commit? Perhaps it was particularly horrific and the police really want to catch the killer quickly to look good in the media. Your DNA matches the DNA the weapon, but match isn't the correct word - a better word is that it aligns / doesn't conflict.
The situation gets worse once bigotry is added to the mix. I stand by my opinion.
> Would you feel differently if your child’s murerer was on trial?
Why would someone be less concerned with the reliability of the criminal justice system in that case? A low standard means it is easier for the government to let the guilty go free while making a show of punishing someone convenient.
There is a trade off between improving the probability of convicting the correct perpetrator and decreasing the probability of falsely convicting someone who is innocent.
This tradeoff is essentially managed by the threshold required for a conviction.
The legal system is generally not optimized purely for the first part of this trade off.
It makes sense that people who suffered from a crime would emotionally prefer more focus on a correct conviction at the cost of more false positives.
Surely the parent of a murdered child is massively compromised in judgement. Considering what they would think to be fair is like asking a madman what's fair, or rolling a die.
Isn't this Would you feel differently argument goes both ways?
Would you feel differently if you killed someone and the police found you only thanks to such dubious from privacy standpoint practice.
Totally an interesting article though, even if the content didn’t completely live up to the title.
I learned things about the precise shape of the payoff curve I should be optimizing for (how investors evaluate assets), which was even an “unknown unknown” for me before I read it.
If memory serves me well, in one of the letters he sent a number of years back (10-ish?), he mentioned that they had someone in their staff managing a portfolio of stock picks and doing a great job at it, and gave a heads up to readers that they might find tech stocks come in and out of their portfolio every so often as a result. Presumably he liked the returns he was seeing and jumped in too.
Also, Apple displays a number of characteristics that he reportedly likes to invest in: the business that Apple is in is straightforward to understand, the market is big and they've a strong or growing market position, they're profitable and have a big cash position, they've a strong brand and difficult to reproduce expertise (i.e. a moat).
What he says around 00:27 about Snickers applies very much to Apple as a brand. This is the only Buffet-ish explanation of why Berkshire would invest in Apple that I could think of.
And apart from this, the decision to buy a lot of Apple stock a while back was made by other asset managers, not Warren himself, and I'm sure was the result of thorough analysis, not based on some offhand comment in an interview.
A couple of possible points on that. Buffett is always learning and maybe he feel's he understands Apple enough now.
Also to quote Buffett "If you look at Apple, I think it earns almost twice as much as the second most profitable company in the United States". Which I guess attracts him - he's always been into consistent profits on modest capital which Apple is good at.
He also said somewhere that a person's "circle of competence" can grow as you learn.
That being said, lot of Berkshire's bets are now coming from Berkshire's next generation of managers - Ajit Jain, Greg Abel, Todd Combs and Ted Welscher. It is speculated that the decision to buy Oracle has come from either Todd Combs or Ted Welscher.
Todd Combs is also said to be driving Berkshire's investment into Paytm, an Indian e-wallet company which again is a technology company.
It does exist: https://github.com/llvm-dcpu16/llvm-dcpu16. The only problem is that there's not going to exist a compiler in-game. People will have to import the compiled assembly code, which isn't nearly as fun as staying in the game during the whole process.
Which is a pretty good argument for building a self-hosted Forth. You get a compiler and REPL suitable for commanding the computer in a package that will fit easily in the limited memory resources available. If you use threaded code with an "inner interpreter", it's possible to generate very compact code and fit a lot in a small space, or you could emit machine code directly ("subroutine threaded code") for speed.
While certainly very cool, this is unfortunately using an unofficial sprite spec[1].
There's plenty of awesome things being done with the official specifications though, like this minesweeper clone: http://0x10co.de/lqnit and this simple raycaster: http://0x10co.de/o3xss
I think it'd actually be pretty cool if the actual game had multiple DCPU versions in the universe. It would give it even more of 1980's computing feel.
Explore the planet, fight off hordes of zombies to obtain... a DCPU16 - REV C!
It's not certain if there's going to be different DCPUs with different specs. Notch has said[1] that there's definitely going to be versions with different clock rates floating around in space or on abandoned spaceships.
Notch has said that when he "upgrades" the DCPU spec, the old CPUs will continue to work as they have, but new CPUs will become available. I don't have the reference handy though.
While it might be using an unofficial sprite spec, everything you linked at 0x10co.de is using an unofficial input and graphics spec… That's what the community does: fill in the missing pieces!
