explain xkcd - User contributions [en]

Talk:2014: JWST Delays

2021-09-09T05:55:41Z

162.158.92.245: official launch date announced

Haha - I made this same graph 2 weeks ago! [[User:Cosmogoblin|Cosmogoblin]] ([[User talk:Cosmogoblin|talk]]) 17:39, 2 July 2018 (UTC)

Suggest the last sentence be made more general: "The title text refers to a fundamental question of the Big Bang Theory; will the universe expand forever, or will is collapse back on itself? The likely answer to this question has changed over the decades as new measurements have been made, and new theories such as dark matter and dark energy developed to explain the new measurements. Apparently, and for an analogous reason, between 2018 and 2020 the likely answer to the fundamental JWST question will change." [[User:GODZILLA|GODZILLA]] ([[User talk:GODZILLA|talk]]) 17:58, 2 July 2018 (UTC)
:I agree to the current sentence saying "and compares the universe’s accelerating expansion to the apparently ever-delaying schedule" but were the hell comes the conclusion that "the JWST will have enough delays to fill a universe"? This does not make any sense. [[User:Elektrizikekswerk|Elektrizikekswerk]] ([[User talk:Elektrizikekswerk|talk]]) 07:59, 3 July 2018 (UTC)

Does today's prediction of 2026 count? If that is included in the data set, it would then skew the best-fit line to be steeper. If a new prediction is made using that new best-fit line, that would further skew the line, and so on, causing the acceleration the title text anticipates between 2018 and 2020.[[Special:Contributions/162.158.63.88|162.158.63.88]] 20:10, 2 July 2018 (UTC)

> Until the slope of the line becomes more than one and the prediction goes to the past, right? [[Special:Contributions/108.162.216.16|108.162.216.16]] 21:55, 2 July 2018 (UTC)

:No, it doesn't count, because it's just '''prediction''', while the data set is of (official) '''planned launch dates'''. -- [[User:Hkmaly|Hkmaly]] ([[User talk:Hkmaly|talk]]) 22:06, 2 July 2018 (UTC)

:The Wikipedia data (taking the midpoint for ranges) fits a linear function with slope 0.660618 and intercept 687.739. This implies convergence at 2026.45, which is why Randall is predicting late 2026 for the actual launch. [[Special:Contributions/172.69.22.56|172.69.22.56]] 15:04, 10 July 2018 (UTC)

[[wikipedia:Heinz von Foerster#Doomsday equation|Von Foersters's doomsday]] is Friday 13th of November 2026. (cue Twilight Zone intro) [[Special:Contributions/162.158.89.175|162.158.89.175]] 21:20, 2 July 2018 (UTC)

Why does he keep saying it's 2021? Is he trying to skip Trump's term or what? --[[Special:Contributions/172.68.211.10|172.68.211.10]] 00:30, 3 July 2018 (UTC)
Why do you think that Trump will get only 1 term?[[Special:Contributions/141.101.76.70|141.101.76.70]] 17:10, 3 July 2018 (UTC)

This is the same chart for the new airport in Berlin. Sadly its slope is not less than one, it is indeed accelerating... 
2006     2011 
2010     2012 
2012     2013 
2013     2014 
2014     2016 
2015     2018 
2016     2018 
2017     2022 
[[User:Fabian42|Fabian42]] ([[User talk:Fabian42|talk]]) 07:57, 3 July 2018 (UTC)
:[https://de.wikipedia.org/wiki/Bau_des_Flughafens_Berlin_Brandenburg#cite_ref-136] says that the planned launch date from December 2017 is in October 2020 (not 2022). That would make the slope slightly less than 1 (unless you ignore the 2016->2011 data point, as outlier) --[[Special:Contributions/162.158.91.35|162.158.91.35]] 09:27, 4 July 2018 (UTC)
::There was a 2022 prediction earlier in 2017, I took the maximum value for each year. And honestly, 2202 sounds more reasonable than 2020 for me. [[User:Fabian42|Fabian42]] ([[User talk:Fabian42|talk]]) 14:39, 4 July 2018 (UTC)

I feel a quadratic regression would be needed to determine acceleration / deceleration [[Special:Contributions/172.68.59.24|172.68.59.24]] 13:54, 3 July 2018 (UTC)

: If you plot out the least-squares fit as it changes over time (i.e. repeat Randall's graph as each new data point was added), it fits a quadratic quite well. And converges to a 2025 date.

I wonder what this chart would look like for new york's 2nd avenue subway.
[[Special:Contributions/162.158.75.130|162.158.75.130]] 17:36, 3 July 2018 (UTC)

At least there _is_ a slope. How about Trump's wall? [[Special:Contributions/173.245.52.151|173.245.52.151]] 00:52, 4 July 2018 (UTC)

Two more lines are coming together... the year and the XKCD index. 2018 should happen next week. [[User:IonFreeman|IonFreeman]] ([[User talk:IonFreeman|talk]]) 14:22, 5 July 2018 (UTC)

The last three data points have a slope greater than one. Just sayin'. [[User:Redbelly98|Redbelly98]] ([[User talk:Redbelly98|talk]]) 19:55, 29 July 2018 (UTC)

:"NASA announced that the launch date has once again been delayed to 31 October 2021." THE RIDE NEVER ENDS [[Special:Contributions/108.162.215.188|108.162.215.188]] 05:24, 17 September 2020 (UTC)

::November to December 2021 now... amazing. [[Special:Contributions/172.69.35.65|172.69.35.65]] 20:24, 21 June 2021 (UTC)

== November 2021 ==

Is [https://www.space.com/james-webb-space-telescope-launch-delay-november-2021] on track for the extrapolation shown? [[Special:Contributions/172.69.35.115|172.69.35.115]] 18:31, 24 June 2021 (UTC)

== Update from 2021 ==
Turns out the last update in the comic seems to have been correct. As of Sept '21, it's still due to go up in Nov. '21.
https://www.jwst.nasa.gov/content/about/launch.html
https://www.nasa.gov/press-release/nasa-announces-new-james-webb-space-telescope-target-launch-date

[https://i.imgur.com/BssUkbG.png '''Updated Image''']. I've added the blue dots to represent the updated launch dates. (Still in 2021) [[Special:Contributions/172.69.48.147|172.69.48.147]] 20:43, 31 August 2021 (UTC)

On September 8, ESA has announced December 18, 2021, as the planned launch date.[https://www.esa.int/Science_Exploration/Space_Science/Webb/Targeted_launch_date_for_Webb_18_December_2021] 2021 remains possible, but if there are further delays (and launches like this are delayed frequently), we may indeed slip into 2022. [[Special:Contributions/162.158.92.245|162.158.92.245]] 05:55, 9 September 2021 (UTC)

702: Snow Tracking

2021-09-05T19:47:39Z

162.158.92.245: /* Explanation */ redundant verbiage

{{comic
| number = 702
| date = February 15, 2010
| title = Snow Tracking
| image = snow_tracking.png
| titletext = I suppose that's more accurately a hare dryer.
}}

==Explanation==
This comic is a guide to recognizing various animals by their footprints. However, the comic typically detours into strange, ridiculous or pop-culture-referencing footprints. In order:

*The first panel is nothing special. Just a regular cat.
*"Moose and squirrel" is a reference to the cartoon ''{{w|Rocky and Bullwinkle}}''. Rocky and Bullwinkle were a flying squirrel and a moose, respectively, and were frequently referred to as "moose and squirrel" by the show's antagonist Boris Badenov.
*[http://knowyourmeme.com/memes/longcat Longcat] is an internet {{w|meme}} from pictures of cats all stretched out that make them look very tall (or long).
*"Mouse riding Bicycle" is a reference to ''{{w|Ralph S. Mouse}}'', a series of novels by {{w|Beverly Cleary}}.
*The hair dryer has melted an irregular region around the rabbit. The title text is a pun on the Rabbit with a hair dryer frame, possibly an homage to {{w|Looney Tunes}}, where shows with {{w|Bugs Bunny}} would often contain a pun on "hare".
*{{w|Legolas}} is a reference to the character in the ''{{w|Lord of the Rings}}'' trilogy of books and movies. Legolas, as an elf, was able to walk on top of snow, while the other races in his party were forced to trudge through it.
*The "Bobcat on pogo stick" panel is a possible reference to the character Bonkers D. Bobcat from {{w|Bonkers (TV series)}}
*The "Knight" panel is a {{w|chess}} reference, as the tracks move just like the knight piece in chess.
*The "kid with..." panels are a reference to ''{{w|Calvin and Hobbes}}'', a comic strip written by Bill Watterson. In it, Calvin has a pet tiger named Hobbes, and sometimes, a cardboard box that "transmogrifies" him to something else. In this panel we see tiger prints, meaning that Calvin became a tiger like Hobbes.
*The same cardboard box is now tipped on its side instead of upside down in the last panel. Now it functions as a duplicator, making multiple copies of whatever is in it. Calvin goes into it, duplicates himself, and they walk and duplicate again, and the cycle repeats.
*{{w|Prius}} is a reference to current events in which Toyota Prius's pedals have allegedly malfunctioned causing accidents. [http://www.nytimes.com/2010/02/04/business/global/04prius.html]
*The {{w|Higgs Boson}} is an {{w|elementary particle}} which, at the time this strip was posted, had not yet been officially discovered (there had been detections at the Tevatron with 4 sigma certainty since the early 2000s). It was tentatively detected in March 2013 in the {{w|Large Hadron Collider}}. The “prints” in the snow in this case humorously resemble the tracks made by elementary particles following a collision of the kind used to search for the Higgs boson in a {{w|particle collider}}.

==Transcript==
:BACKYARD SNOW TRACKING GUIDE

:[Each panel contains an overhead view of tracks through the snow, with a caption indicating the apparent source.]

:[Standard paw prints through the snow.]
:CAT

:[Large split-toe tracks and smaller rodent tracks.]
:MOOSE AND SQUIRREL

:[Cat prints, but with more space between the pairs of prints.]
:LONGCAT

:[Two similar careening tire tracks.]
:MOUSE RIDING BICYCLE

:[Longer tracks, with a large melted ring surrounding a point in the middle of the frame.]
:RABBIT STOPPING TO USE HAIR DRYER

:[No visible tracks.]
:LEGOLAS

:[Single deep holes with cratering.]
:BOBCAT ON POGO STICK

:[Round prints that suddenly turn to the right halfway into frame.]
:KNIGHT

:[Human footprints up to a square melting pattern, turning into animal prints.]
:KID WITH TRANSMOGRIFIER

:[Human footprints up to a rectangular melted area, which are then doubled to another rectangular area, which are then doubled again up to another rectangular area, which are then doubled.]
:KID WITH DUPLICATOR

:[Right curve on a road, with tire tracks careening out of frame.]
:Out of Frame Garden Owner: MY VEGETABLE GARDEN!
:PRIUS

:[A series of spiraling and outwardly traveling lines extend from a point in the middle of the frame.]
:HIGGS BOSON

{{comic discussion}}
[[Category:Calvin and Hobbes]]
[[Category:Bobcats]]

2510: Modern Tools

2021-09-02T17:44:04Z

162.158.92.245: why the comic is there twice?

{{comic
| number = 2510
| date = September 1, 2021
| title = Modern Tools
| image = modern_tools.png
| titletext = I tried to train an AI to repair my Python environment but it kept giving up and deleting itself.
}}

==Explanation==
{{incomplete|Created by CUTTING EDGE ANCIENT TECHNOLOGY - Please change this comment when editing this page. Do NOT delete this tag too soon.}}
[[Cueball]] tells [[White Hat]] how he has trained a {{w|artificial neural network|neural net}} to generate mostly valid {{w|Make_(software)#Makefile|Makefiles}}.

This is the file type that the {{w|Make (software)|Make}} searches for. In software development, Make is a build automation tool that automatically builds executable programs and libraries from source code by reading files called Makefiles which specify how to derive the target program. (See [[2173: Trained a Neural Net]]). Make is a very old tool, having first appeared in 1976.

Then Cueball continues to tell that he next will train it to distinguish between Bash and Zsh.

{{w|Bash (Unix shell)|Bash}} and {{w|Z_shell|Zsh}} are two {{w|Command-line_interface|command line interfaces}} for {{w|Unix-like}} OSes. The way to execute commands is almost identical, making detecting a script that contains a mixed syntax nearly impossible. This was previously referenced in [[1678: Recent Searches]]. Bash and Zsh are also old tools, having come out in 1989 and 1990 respectively.

A human-designed 'random Makefile'-maker might have been written with this explicit choice amongst the earlier decisions in the generation process, but an AI might be assumed to have started (many, many generations ago) with something close to utter nonsense and painstakingly reached the stage of (mostly!) valid files along the way. Some might say that the differentiation training would have been better added at another point in the lengthy process.

On top of that, the current (mostly valid) results may even be {{w|Polyglot (computing)|polyglot}} and/or {{w|Agnostic (data)|shell-agnostic}}. Dependant upon the {{w|Fitness function|fitness tests}} in use, many other {{w|List of command-line interpreters|$SHELL}}-choices and Makefile styles may have been coevolved as valid (if rarer) subgenus of outputs, such as a ''command.com''-based makefile.

In the caption it states that Cueball is using modern tools to make ancient technology, as opposed to other people who use ancient tools and UIs ({{w|User interface}}) to develop Modern Tools.

In the title text Randall states that he tried to train an AI ({{w|Artificial intelligence}}) to repair [[1987: Python Environment|his horribly broken Python environment]]. But the AI kept giving up and deleting itself. The joke partly relates to when it or is not appropriate to personify goal-driven processes. In the study of alignment of artificial intelligence, it is common to consider AIs finding ways to meet the tasks they are given that are highly unexpected, and then developing into an {{w|Instrumental_convergence#Paperclip_maximizer|apocalypse}}. A common unexpected solution encountered in research is that the agent finds a way to disable itself as more efficient to meet its reward parameters than anything else it discovers, and then learns to repeatedly do so. The AI might be so intelligent that it had developed critical 'personal' opinions that led it to be so intellectually appalled by the task, or else just found it impossible to fix the python environment and therefore justify its own existence, that it had no other recourse but to commit a form of suicide because Cueball's code was that bad ([[:Category:Code_Quality|which is a recurring theme for Cueball]]). [[Python]] has been a recurring subject as has [[:Category:Programming|Programming]] and [[:Category:Artificial Intelligence|Artificial Intelligence]].

The main joke is that Cueball is using cutting-edge tools to develop very old technologies, which is not particularly useful. As the caption implies, it is much more common for people to use fundamental and well-established tools as the toolchain or building blocks of modern technology. A concrete example of this is writing scripts using decades-old Bash to automatically set up a significantly newer (2014) technology called {{w|Kubernetes}}.

==Transcript==
:[Cueball is sitting on an office chair at his desk typing on his laptop. White Hat is standing behind the desk looking at Cueball.]
:Cueball: Okay, I've got this neural net generating mostly valid makefiles.
:Cueball: Next I'm going to train it to distinguish between Bash and Zsh...

:[Caption below the panel:]
:People often use ancient tools and UIs to develop modern cutting-edge technology, but I do it the other way around.

{{comic discussion}}

[[Category:Comics featuring Cueball]]
[[Category:Comics featuring White Hat]]
[[Category:Programming]]
[[Category:Artificial Intelligence]]

==Transcript==
:[Cueball is sitting on an office chair at his desk typing on his laptop. White Hat is standing behind the desk looking at Cueball.]
:Cueball: Okay, I've got this neural net generating mostly valid makefiles.
:Cueball: Next I'm going to train it to distinguish between Bash and Zsh...

:[Caption below the panel:]
:People often use ancient tools and UIs to develop modern cutting-edge technology, but I do it the other way around.

{{comic discussion}}

[[Category:Comics featuring Cueball]]
[[Category:Comics featuring White Hat]]
[[Category:Programming]]
[[Category:Artificial Intelligence]]

Talk:1581: Birthday

2015-09-23T11:26:53Z

162.158.92.245:

http://www.cnbc.com/2015/09/23/happy-birthday-song-now-in-public-domain.html [[User:Mwburden|mwburden]] ([[User talk:Mwburden|talk]]) 11:09, 23 September 2015 (UTC)

https://ia601904.us.archive.org/13/items/gov.uscourts.cacd.564772/gov.uscourts.cacd.564772.docket.html [[Special:Contributions/162.158.92.245|162.158.92.245]] 11:26, 23 September 2015 (UTC)