h a l f b a k e r yThere's no money in it.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
browse anonymously,
or get an account
and write.
register,
|
|
|
Basically all those nerdy things going on on the Hilter detection thread could be adapted by geeky programmer types to allow the halfbakery to become a universal control portal for anything on the internet.
The way it works is that the scripts or programmes or whatever codey stuff is going on over
there is used to monitor posts to the halfbakery. For example I could post a new idea such as "get Amazon to send me a case of champagne" and the idea title, summary, and description would contain key words or phrases. Links could call remote resources. Even oting could be used to prioritise tasks or give feedback within the system. The universal API server would see my idea and would parse my idea and would the post annotation(s) asking for more details or giving result information.
I'm sure more technically complex tasks could be created, such as controlling a nucular power station or sending people to mars, and this might require more than one universal API server to interact along with multiple ideas, annotations and ongoing discussion threads between the servers and their human controllers.
Obviously this all depends on not irritating [jutta] and so the control syntax as well as the posting style of the user accounts used by the scripts or programme servers needs to be convincingly halfbaker in style so as not to be instabanned.
Nucular power station
https://www.geograp...cular+Power+Station
I thought everyone had heard of Sizewell Nucular power station [pocmloc, Dec 21 2020]
Make a plastic sticker that fits around the shower drain and does 40 medical tests simultaneously; it's kind of like a color change pregnancy test
https://www.quora.c...nswer/Treon-Verdery
Make a plastic sticker that fits around the shower drain and does 40 medical tests simultaneously; it's kind of like a color change pregnancy test, but your HMO sends it to you once a year to screen for everything. [beanangel, Dec 22 2020]
Œting
https://en.m.wikipedia.org/wiki/Œting
[hippo, Dec 22 2020]
0oo.li - Network of Functions
https://0oo.li/meth...etwork-of-functions
[Mindey, Feb 01 2021]
HalfBakery API
[xaviergisz, Feb 01 2021]
Please log in.
If you're not logged in,
you can see what this page
looks like, but you will
not be able to add anything.
Destination URL.
E.g., https://www.coffee.com/
Description (displayed with the short name and URL.)
|
| |
More details, please, especially on this 'nucular' power
station. I have a friend in <strikethrough>Iran </sr>Montana
who would like to create one... |
|
| |
Anecdotally I've heard of people using twitter a bit like this
- as a control interface to some online...thingy. |
|
| |
Yes I have seen the "threadreader" app on Twitter, where you basically type "Threadreader please reed this for me" and it replies with a digested complete sequence of discussions. |
|
| |
But yes, basically this is a halfbakery "thingy". COuld redo the title if you prefer [KDF] |
|
| |
I don't really understand it, but anything that gives us geniuses
at the bakery more power...I'm all for it. |
|
| |
On Tuesdays I question authority. |
|
| |
Call to action: On Tuesdays, question authority. |
|
| |
I don't really understand it either [bliss], only the vague outline. Hopefully the techy ones will fill us in on the detail (or even better not fill us in but just go away and do it) |
|
| |
// allow the halfbakery to become a universal control portal for anything on the internet. // |
|
| |
Such a dearth of ambition, [poc] ... there's nothing wrong with having the hb take over the Universe, shirley... ? |
|
| |
//oting could be used// - I assume you mean the region of Œting on
the French/German border (see link)? |
|
| |
A very dull place; almost nothing of interest, probably because historically it was usually well behind German lines from the outset. Further west, there's Metz, Sedan, Nancy, Verdun... but not much around Forbach at all. |
|
| |
//usually well behind German lines// |
|
| |
You seem to be overlooking the period from, say 1618 to 1815,
when most of the traffic was the other way. |
|
| |
No I didn't mean Œting, If I had meant Œting I would of typed Œting. I am as capable as you of typing Œ as many times as I want, clœver clœggs. |
|
| |
Maybe this could be done using Generic Algorithms. |
|
| |
I know you people don't mean it and it's plonk, but one way to propagate the content on the halfbakery, not so people read it of course, |
|
| |
is to put it on the highest capacity server on earth with the highest bandwidth in searchable form. |
|
| |
That might be Earth's #2 site, youtube. It has about 2 billion users and I saw a video title that suggested it could handle millions of concurrrent (simultaneous) users. |
|
| |
Computers are very fun, and thinking of things like "Grab the 10,000 most popular topics on wikipedia from "wikishark"(?), and automatically make a video of each one from wikipedia content while attaching a different halfbakery idea to each one, then bask in the advertising revenue" is sort of a script Kiddie thought I had. They click at youtube for Taylor Swift, and get Taylor Swift wiki content and some text page images about medical test kit shower rings. |
|
| |
Hmmmm. advertising the .5b to me looks ike it has generated controversy. What if there were lots more visitors, they liked it, requested an account, and posted ideas? It sounds good to me. |
|
| |
The Top 20K wikipedia articles, each with it's companion .5B idea, luring people on youtube might make, at 40 views per wikitopic video per month, 400K views. You used to get either $1000 or $2500 per million views on Youtube, so whoever can be bothered to write the big halfbakery advertisement can earn about $400/month for a very simple computer program. |
|
| |
Anyway, I've digressed. I guess the fun/funny part is that, if for some unknown reason a billion people really wanted to get a ahold of the text of 1 coded halfbakery idea so they could use it to do something en masse youtube likely supports 10 million or more concurrent viewers on one video already. So, quickly, someone make up an idea that will be that valuable! I know |
|
| |
You are talking about some mysterious coded information. But it seems nicer if all 2-10 billion people visiting the halfbakery's youtube library got something good and comprehensible out of it. |
|
| |
Actually it might be an ethical way to advertise the .5B. |
|
| |
I won't do it, because I'm too lazy, and my "script KIddie" skills are too out of date. Actually, It's worse than that. It really is pure laziness. I think it might qualify as a nice thing to do though. Youtube viewers really do get to read/hear about Taylor Swift, and they really can click away to some other video the instant they see .5B content. |
|
| |
This is only tangentially related to the actual posted idea. As [zen_tom says], reposting new things on the halfbakery the moment they appear to twitter in realtime might work for that. |
|
| |
//If I had meant Œting I would of typed Œting// -
"...would *have* typed Œting..." :-) |
|
| |
If I had meant have I would of typed have. |
|
| |
<Surreptitiously consults "Time Traveler's Handbook of 1001 Tense Formations" (3rd. revised ed.) by Dr. Dan Streetmentioner/> |
|
| |
That would require a cooperative network of functions (see link:
"Network of Functions"). [+] |
|
| |
Hi Mindey, that "network of functions" idea, I've seen cast in a slightly different form, as a "network of
relationships" all described in something called RDF triples. For any "thing", you describe it in terms of its
relationship with other things using strict Subject, Predicate, Object "triples". A collection of these triples
is formed and you can ask queries about it. So for example the following triples might exist: |
|
| |
Subject Predicate Object
Luke isA Man
Luke isTheChildOf Anakin
Luke isTheChildOf Padme
Leia isA Woman
Leia isTheChildOf Anakin
Leia isTheChildOf Padme
Anakin isA Man
Padme isA Woman
|
|
| |
All the above could be "known" within the network and operate as a working repository of interconnected relationships. |
|
| |
And then, much later, new knowledge can be added, it's just an insert statement: |
|
| |
Anakin isAlsoKnownAs Vader
|
|
| |
Using the OWL language, you can define (again, using triples but it's more involved than would be helpful here)
the logic that describes a sibling relationship without having to expressly mention it. Similarly, assuming the logic
behind isAlsoKnownAs is defined, you can now ask the question "Is Vader Luke's Father?" and get a logically
determined answer in the affirmative. |
|
| |
It can produce quite exciting results and it's fairly easy to hoover up
lots of text and construct lots of these tiny atoms of interconnection. The key aspect of being able to define what you
know, and add new knowledge as you learn it makes for a nice resilient, open-ended yet systematic representation of the
world. |
|
| |
It's easier to manipulate computationally if you use NPN instead of infix notation. |
|
| |
You could even use RPN and then implement the algorithms in Forth... |
|
| |
<Sound of screaming and receeding running boots/> |
|
| |
So in this case the Subject Predicate Object (SPO) isn't an operation, it's a unit of storage. In this paradigm, the fundamental unit is a relationship. You can define
things on their own, but you'd use the syntax
|
|
| |
You might even go as far down the rabbithole as to say: |
|
| |
Thing hasProperty ConcreteActualThingInRealLife |
|
| |
or alternately, if "Thing" is something more conceptual: |
|
| |
Thing hasProperty AbstractConcept |
|
| |
then Widget, since it's been identified as being a Thing would take on the appropriate attributes of all Things as per the underlying
definitions held in the machine. |
|
| |
To do calculation on them you might ask questions like: |
|
| |
Find me all the Things thatAre RealThings and are locatedIn Croydon and who's ownerIs Steve and that haveColour in ("Red", "Green", "Blue") |
|
| |
And at this point, you'd have to arrive at some notation choice - though there is a language already established for that called SPARQL which I
suppose is kind of infixy. |
|
| |
It's interesting though, when everything is a relationship, all you have to do is pattern-match for relationships that you're interested in -
there's no computation as such, just search. |
|
| |
[zen_tom], agree with much what you say. |
|
| |
// I've seen cast in a slightly different form, as a "network of
relationships" all described
in something called RDF triples // |
|
| |
Triples are outdated. Tuples are perfectly sufficient, because you can
define any triple
as ordered pair of tuples. (You can't argue with math.) |
|
| |
// It's interesting though, when everything is a relationship, all you
have to do is pattern-
match for relationships that you're interested in - there's no
computation as such, just
search. // |
|
| |
Well, yes!~ It's beautiful simplicity, and that, here it goes: |
|
| |
Past: dots -> line -> regression
Present: objects -> hypersurface -> deep learning
Future: concepts -> model -> search. |
|
| |
What? 1-2-3 is not better than 1-2, 2-3. Sounds heavy on cycles and memory to me. |
|
| |
And, being picky, isn't it regression(dots) -> line |
|
| |