davidmanescu

I have no special knowledge about IBM Vs Apple historically, but: a quarter billion in CAPEX when you've earned a billion in revenue in a single year is extremely different to what we're seeing now. These companies are spending all of their free cash flow, then taking on debt, to the tune of percentage points of world GDP, and multiples of any revenue they've seen so far. That kind of oversupply is a sure fire way to kill any ROI.

Archived: https://archive.md/fqc2u

Silver's forecasts do predictably change, more precisely in that they swing deterministically in favour of one candidate whenever the most recent poll to come in favours that candidate heavily.

Each time the probability drops below 40% (in your example) it is likely because the fair odds swung in the latest polls. To attempt to take advantage of this is to ignore the recent evidence (conditional upon that evidence being quite different to the current concensus).

It's not at all unreasonable to do so, but isn't arbitrage. What if all polls started being increasingly conclusive leading up to election? 40, 39, ..., 0. No surprises and no arbitrage.

Perhaps another way to express this critique of Silver's methodology is that it lends too much weight on recent outlier observations? Sure if you knew this you could profit off these swings, but making the arbitrage argument doesn't add much to the discussion to my mind.

Adaptive Brightness: "solving a problem that's been solved numerous times, but this time using machine learning"

In the article they've trained the AI to predict what frames the game would provide given a certain input, but to demonstrate its effectiveness they show the original game given a certain input, side-by-side with the AI's prediction given the same input. I sure hope that wasn't part of the training set. Teaching an AI to recognise something it's already been taught and mirror it relatively closely isn't an achievement...

Equalities that involve units aren't straightforward. Not a physicist of any sort but don't all the 1's still fundamentally change the equality since they change the units? I would guess they therefore implicitly change the intuition behind these equations?

Very true - neither the person who wants BTC (but has no opinions on BCH) nor the BCH buyer wants to take the risk that the other coin will drop before he/she can sell it. In other words the risk premium isn't zero.

However a rational (impartial?) actor should be willing to take that risk if the premium is big enough. That could mean buying pre-fork or selling pre-fork if there is an obvious opportunity. Which means in an efficient market the risk premium should be zero "on average". It's hard to reason about averages with one sample but that sample was distinctly not zero!

The interested parties on either side would have been better off buying pre-fork and selling the one they didn't want, rather than buying after the fork, even though the end result is the same. That is irrational.

PE ratios correlate highly to dividends for shareholders. Does holding bitcoin have a way of compensating you when the transaction count increases?

Indeed. I guess volatility like this isn't a dealbreaker for other asset types so it shouldn't be here. Just seems odd that we call it a "currency" when these things happen routinely.

To me an asset that moves 15% in a day, and which no-one knows what it's truly worth, isn't great as a store of value. I'm not going to speculate about where bitcoin will be trading when it becomes a stable, useful asset, but in order to justify a market cap anywhere near this high it will eventually need to generate actual utility for its users. Can someone explain where that will come from? Cheap currency conversion and easy money transfer?

That manifold is a torus - you can walk in circles by walking West-East or North-South, each of which repeats every 1 distance unit. If you walk along the diagonal you get a repeating path every 1.4 distance units. All "great circle" paths (straight lines on a manifold) on a sphere repeat every 1 unit of distance.

I would expect these are broken down by enzymes in the mouth/gut long before they have a shot at entering the bloodstream to bind with any antibodies.

Payment systems like visa are relatively cheap to operate when being used honestly. There may be overhead when people commit fraud, but to me that seems more prudent than giving honest actors an intentionally inefficient system just to make abuse more difficult.

Since I personally don't worry about centralisation and "trust", I like the idea that a system like the blockchain could be used to revolutionise payments without the need for tin foil hats and proof-of-work.

My suggestion would be to place the car by distance travelled since last leaving the home straight. Use steering + throttle inputs to determine when you are on the home straight, make a best guess at where the start/finish line is, then guess where you are based on distance travelled since last passing it.

Did a back of the envelope trigonometry calculation as an approximation, seems to check out :)

Follow-up question - after extending the rope by 0.12mm and pulling it up to allow the cat to pass under it, how far away are each of the two points where the rope lifts off the ground?

I'm assuming the GIF showing a phone taking a Motion Still was itself made using Motion Stills. Fun.

Ballpark numbers here; by similar reasoning, since the industrial revolution humans have put out no more than 26,000 years worth of all volcanoes' emissions (say 200 years worth of 130x emissions). Therefore (obviously hypothetically) if we were to cease emitting any CO2 at all, today, the effects of all the CO2 ever produced by humans would become largely irrelevant in just tens of millennia, which is a mere instant in geological timescales!

I think in the scheme of things, no matter how bad things get the planet won't even notice us. However the question remains what life will be left to see it when it does finally return to normal.

I hope I don't come across as too obtuse, but it might still make economical sense for the PSF to backpedal on their decision: even if Python3 is a good solution for the future, it's caused years of arguments and still hasn't been adopted by the majority[1].

Ridiculous as it might sound, that implies it would actually take less work to devise a path for Python2 that's backwards compatible but still has a future, and invest the time to backport all 3-only code, than it would to proceed with killing off 2 and porting everything over. It's also nothing more than everyone whose code is 2-only is being asked to do sometime in the future (they chose "the wrong competing standard" and have to pay the cost). Either way someone has to foot the bill for a unified Python but couldn't it just be that the early adopters of 3 could pay that cost if it saves updating (say) twice as much legacy code in favour of backporting the 3-only code?

Of course, the future version of Python should be the best form of the language with the right features, and that's why it was decided to kill off 2, but if after this many years the initiative hasn't completely succeeded, there's always the option to reverse course.

Heck, in the time until Python 2 is officially gone, maybe this new fork will evolve into a better language than 3 and still be backwards-compatible with 2!

For now I'm sticking with 2.7 for as long as I can but will just accept it when the time comes.

[1]: https://mobile.twitter.com/vlasovskikh/status/80172061331236... for example (I don't think I've made a controversial claim here but could be wrong)

https://xkcd.com/927/ comes to mind when I read this.

My favourite part of this (aside from the message) is that it links back to this exact discussion page.

Interestingly, it appears as though the total salary of those same computer programmers follows a similar trend, up until around 2005, and then reverses. http://www.wolframalpha.com/input/?i=%28computer+programmers...