robochat

I think they mean that the model should have sleep period where they update themselves with what they learnt that day.

I remember being amazed when I learnt about summing arithmetric series.

They took a multiplier of 5x (4 indirect deaths for every direct death) and stated that this was conservative given studies of previous conflicts.

"Valuable Humans in Transit and Other Stories" by Qntm has some good (harrowing) stories about human uploads too.

We actually use the twitter detections to launch analyses of the seismic data in order to get confirmed results for events that aren’t reported yet [1] but there are some statistics for the twitter detections in the supplementary material of that article [2]. Basically, in 2016-2017 (wow, so long ago), we detected 893 earthquakes via twitter, with a median delay of 67s and a median separation from the published epicentre of 94km. Note that estimating earthquake epicentres is nontrivial anyway and so, for comparison, 10km accuracy would often be considered ok. So the twitter, I mean X, method isn’t optimal but it gets you down to the right region. Partly it’s because geocoding the tweets is inaccurate and partly it’s because people live clumped together in cities rather than smoothly spread over the surface of the earth.

[1] https://www.science.org/doi/10.1126/sciadv.aau9824 [2] https://www.science.org/action/downloadSupplement?doi=10.112...

There is also the Earthquake Network (EQN) app that works on very similar principals to Google’s system - phones monitor their MEMs accelerometers, when they are left charging with their screen off, and when enough neighbouring phones detect vibrations simultaneously an earthquake is detected and apps nearby are alerted. It’s been running since 2012.

[1] https://sismo.app

The USGS created a system to do exactly this about 15 years ago. I’m not sure whether they’re still running it but at the EMSC, we've been running a similar system for many years to highlight earthquakes important to the public and improve our messaging. Twitter doesn’t give access to geotags anymore but we do manage to roughly estimate an earthquake’s location by analysis of the tweets. Estimating magnitude is much more difficult. Naturally there are some false positives but it works well overall.

[1] https://www.usgs.gov/media/audio/shaking-and-tweeting-usgs-t...

I have used Thunderbird for years and my work account has to handle large numbers of emails. In general it is fast for me on both Windows and Linux. I have had some issues over the years but they normally turned out to be due to corrupt mail folders.

Can anyone explain the maths that he did because my calculation give different results. Isn't there only a 1 in 2588 chance of a student 'guessing' 19 'answers' out of 45 where there are 5 options for each answer? Whereas the article states that it is 1 in 100 ? (actually he writes 1:100 which is odds, so it is 1 in 101?) Don't we just use the Binomial distribution, so: Prob = 45!/(19!26!) * (1/5)*19 * (1 - 1/5)*26 It's really bugging me that I can't get the maths to work.

Actually, I can't seem to get the maths to work. Isn't it just the Binomial distribution? Each question has 5 options so the probability of 'success' is 1/5 and so to get 19 questions 'specifically wrong' out of the 45 planted questions by chance is just (1/5)*19 * (1-1/5)*26 * 45!/(19!26!) = 1 in 2588 but in the article it is 1:100. What am I doing wrong ?

This feels harsh because if the tests were similar but with small differences then getting the right wrong answers seems likely to happen anyway. Also, he went the extra step and outright failed the student but I'm not sure why that extra step was justified, surely the cheaters were automatically going to get very poor marks.

It was an online course and the exam was online too.

For all of my major exams in the UK, we were encouraged to study by practising on old papers and my score would have been abysmal if I hadn't. So not being allowed to look at past exams feels strange to me too.

There is a strong connection between the incandescent lightbulbs and the sun because the sun is almost a blackbody emitter too. Studying incandescent lightbulbs allowed researchers to discover the blackbody spectrum and Kirchhoff's law of thermal radiation, which was then applied to the Sun to discover its apparent temperature. The spectral lines in the Sun's spectrum (the frequencies where it deviates from the blackbody spectrum) then allowed researchers to study elements present in the Sun's photosphere.

The paper is written in scathing tones, by page 11 I was laughing at each new report of QED's 'perfect' agreement with whatever was the latest experimental result for the electron's anomalous magnetic moment although as it continued as far as page 14, it became sobering. However, the recent discrepancies are way down in the decimal points of the result now; when these calculations require 1000s of complicated Feynman diagrams and pages and pages of calculation, errors seem inevitable. My hope is that an alternative approach can be found that avoids these complicated calculations and in fact, researchers already seem to be working on this [1].

[1] Nima Arkani-Hamed - The End of Spacetime (Sep 2022) https://www.youtube.com/watch?v=GL77oOnrPzY

I agree with this sentiment but in this particular case, with the personality that Putin projects to the world, it seems like appeasement would just encouragement him to try to take more.

I like nim but I've never really gotten used to the slicing syntax. I guess that I'm just too used to python, plus you have to be a bit careful to leave a whitespace when doing backwards indexing so that the operators work correctly.

So python:

    a = [0,1,2,3,4,5]
    a[0:5] => [0,1,2,3,4] - half closed interval
    a[0:-1] => [0,1,2,3,4] - negative indexing
    a[:5] => [0,1,2,3,4] - skipping start index 
    a[3:] => [3,4,5] - skipping end index

    var a = [0,1,2,3,4,5]
    a[0..5] = [0,1,2,3,4,5] - closed interval
    a[0..^1] = [0,1,2,3,4,5] - closed interval with negative indexing
    a[..4] = [0,1,2,3,4] - skipping start index
    a[..<5] - Error - doesn't work
    a[3..] - Error - doesn't work
    a[0..<5] = [0,1,2,3,4] - half closed interval
    a[0..< ^1] = [0,1,2,3,4] - half closed interval with negative indexing 
                              (Important - the white space is required, it will crash otherwise!)

You can easily get bash to do that by adding these lines to your .inputrc file.

  "\e[A": history-search-backward
  "\e[B": history-search-forward
  "\e[C": forward-char
  "\e[D": backward-char

Yes this is why many high frequency traders have installed laser or microwave links to their nearest stock exchanges, simply to gain a few microseconds advantage in their trades. Fibre isn't as fast as laser going through the air. It's 50% slower (as the parent comment states) (unless you are using special hollow core fibres which are uncommon for now). More importantly, fibres rarely go in a straight line between 2 points, they wind their way through buildings, down into basements, through buried pipes etc and this all adds extra distance to the route and hence more delay to the signals. A line of sight link is the shortest route between 2 points.

Actually, Einstein's 1935 paper along with Boris Podolsky and Nathan Rosen basically highlighted the existence of entanglement in Quantum Mechanics. Of course Einstein was trying to show that QM was wrong via the EPR paradox but although reality has since proven him wrong, it was still a major paper written decades after his more famous work.

There's been some hypotheses of photosynthesis relying on quantum effects to improve its efficiency but it's still contentious the last that I checked [1].

[1] https://physicsworld.com/a/is-photosynthesis-quantum-ish/

Taco Town!: https://www.youtube.com/watch?v=evUWersr7pc

Jim Carrey did a pretty good Matthew McConaughey SNL skit: https://www.youtube.com/watch?v=z3eN9u5N2Q4

There is an offhand remark in the article that MRAM failed but why didn't it work out?

Ricardo Montalbán excellently sent himself up in Freakazoid.

"Because he tasks me! He tasks me! 'round the moons of Snivia, I chuckle at thee. Beyond the Corpian clouds I chuckle more at thee. Revenge is a dish best served with pinto beans and muffins! Kirk, old friend, I... Oh, sorry. [He fixes his tie.] Good-bye."

I know that there's an xkcd comic about this but I think that the best solution would be to create a new text based format with a new extension ie. "bsv" - better separated values. This new format would either be much stricter than csv or have a compulsory header that defines the file's format (like the text encoding, separator character etc).

If it is stricter, it would have one type of field separator that is not commas since some locales use them as decimal places (I'm looking at you, France) but something like '|'. It would insist that dates were iso8601. It could define how fields can be escaped and quoted - although I would prefer if quoting was kept to a minimum. The format should also allow for comments i.e # so that people can comment their datasets inside the same file.

Alternatively or in addition, it could have some header lines:

1) A header that defines the encoding, separator, decimal separator, quote character, escape character, line ending character, date format ...

2) A header that defines each column's name

3) A header that defines each column's data type and formatting

4) A header that defines each column's unit like m/s or kg - ok, this is a bit of a stretch but it would be great to have.

or some variation of the above.

Fundamentally, this bsv format would still be csv and most programs would still be able to read it with the parsers that already exist or be quickly adapted to read it. It could still be easily edited by hand but the metadata would be present.

I suspect that this is just a pipe dream because people would find hundreds of ways to break it but toml took off and that didn't exist so long ago.

I've often thought that it would be good to have a app that could tell me that my doors and windows were shut and locked for when I suddenly can't remember if I actually did it. However the security implications of having such a system would worry me, particularly if it became popular. If hacked, it could tell thieves which houses had been accidentally left unlocked in an area, which could be worse than accidentally leaving a door unlocked all day. Also it might even make the company running the system liable for any thefts caused by hacks to their system and so it might be hard for a company to run such a system.

I guess that I have a smart home albeit a cheap version. I have used sonoff smart switches to give me remote control of my flat's electric radiators. This has given me extra control over their schedules than available using their in-built interfaces. Plus I can turn them on remotely if I am unexpectedly coming home so the flat is warm when I get back.

I live in France and the radiators have a control pin that accepts various 230V AC waveforms (!) to set different functions but luckily simply on/off can be used to toggle between the radiators' high and low temperatures (that I set via each radiator's internal interface). It's crude but effective. I use an app to control the remote switches; I'm not totally happy about relying on a cloud based app but I can always bypass the remote control locally if there's an issue and in summer the radiators' are simply switched off so the risks are minimal.

Relative imports used to work much more naturally IMHO in python2 but then they broke it in python3 because Guido wanted scripts and modules to always be separate codebases. So, whereas it used to be easy to have a module that could also be run as a script inside a package, this is now very difficult to implement. To the extent that any python2 code that does this, should probably be refactored when being ported to python3.

At one point, maybe 10 or 12 years ago, there was lots of buzz about electrowetting displays that were meant to able to display video. I think that Phillips were working on it. Then I never heard of them again.

On the other hand, it's possible to buy oled televisions now so sometimes new technologies do make it to market.

This one sounds interesting although using both gold and platinum sounds expensive.