Someone I know speaks in a reverse tree of sorts which actually does resemble a "weave", they start with various statements about the topic at hand without ever mentioning the topic, and eventually arrive at stating the topic near the end (hopefully). Sometimes I have no idea what they are talking about because they forgot to mention it until the very end when they have merged all their branches.
Yeah but a callback is just a reference to a previous joke, not a tying together of unrelated points. Does the latter have a name? It seems so hard that I can't remember seeing it much.
quuxplusone · 3h ago
It's the third beat (final stage) of the Harold, for what that's worth.
I had a professor in a math heavy subject that just started throwing out maths on the blackboard to derrive formula after formula.
It could take 20 minutes until we reach the conclusion, at which point he finally explains what the purpose of the final formula is and why we want it.
I got the habit of reading his book in reverse before the lectures; reading section by section in reverse order from the end. This way the mathematical calculations had a clear goal and were faar easier to follow.
... brilliant maths, but he was fully and utterly incompetent at teaching it. And he had a bit of an ego about how many students fail each year because "they're lazy".
Me and a few friends did deep recaps to de-tangle the explanations using his book, rephrasing it in a easier and shorter format; and he accused us of cheating because our scores deviated from the normal distribution.
All this to say; sometimes clever doesn't correlate well to great with words.... though dont take that as a endorsement of trump.
mebassett · 4h ago
"As with many books, this one is best read piecewise backwards. In describing the contents, I accordingly begin with Part III."
Intro to one of the maths books I had to reference to do my masters thesis. :)
xtiansimon · 2h ago
> “…at which point he finally explains what the purpose of the final formula is and why we want it.”
I can’t learn this way. Great description of some of my instructors. I gravitate towards design and engineering. Goals are first. This sounds like play: “…and this happened, and this happened…” nope. Not me.
smusamashah · 5h ago
This reminds a lot of a post here on HN. Someone described a very different way some people talk to each other. If someone else was else was listening in, no sentence and it's response from the other person would seem related at all. But to them that communication is very coherent.
Trying to search that post.
Edit: in the discussion there was a link to a do a YouTube video where to movie characters were playing word badminton with each other.
My mother-in-law is also a master of 'the weave'. When my wife and I visit she uses it to inform us about every possible happening of gossip and 'important events' . She just needs a key word and then she uses the chance to get all the information out in one stream of thought. (One problem is that she speaks in dialect and also assumes some deep knowledge about my wife's family relations). I often have thought of trying to map the branching. I often assumed that my wife could follow but I found out that everyone seems happy, without any deeper exchange of information.
locallost · 3h ago
I talk like this and not sure why. I thought about it a lot over the years and figured that I for some reason need to have a, well, reason to tell a story. And so I tell it not as just a context free snapshot, but a bit more detail to it. And sometimes that resonates with people well, especially when it's something funny, but a lot of times it's too much to ask, especially these days when everyone has a short attention span. Which is fair, I'm the one doing the talking, listening is definitely more challenging.
shrx · 7h ago
I think the diagram would be more comprehensible if the branch (topic) name would be shown next to the "New Topic" label, not only at merges. I had to read it from the bottom up the first time to understand what's going on.
heresie-dabord · 5h ago
It's hard to graph the semantics of a set of arbitrary human statements. It's even harder to work with literal, even intentional non sense.
Humans can indeed make sense. Not to be too Swiftian, but in some countries, children even go to school for it.
For semantic analysis, however, git is just not the right tool. It's a chronological graph that affords diffs. For code.
We need Python NLP and spacy here. But even the best tooling won't get far.
A compiler would abort nonsequential logic and unsatisfied contracts and grammar.
An important business presentation would have structure and facts. Inside the theoretical classroom, a public speech is different from casual, random remarks. Unless the speech is intended for entertainment (e.g. comedy, theatre) or some dark usage, such as propaganda.
From TFA:
> the cyclical pattern of his speeches, little snippets of “the best words” and talking points assembled like a ransom note cut from a magazine
That's gold!
keepamovin · 9h ago
That's funny. I'd like to see you do this more, on more of them. More diagrams! A good source is the introduction he gives at cabinet meetings. Or the way he answers questions with the press. You could compare with his campaign speeches and improvisations.
AceJohnny2 · 8h ago
> Four minutes of the weave was about all I could handle
I wouldn't want to torture the author by force-exposing him more to Trump's inanities. That's already 5x longer than I can stand before the urge to tear my ears off becomes unmanageable.
MPSimmons · 8h ago
I hate to be like, "Let's use AI for this" on everything, but this actually does seem kind of like a good thing for AI to do, rather than subject a human to it.
teleforce · 6h ago
> this actually does seem kind of like a good thing for AI to do
Actually this type of applications should be the killer application of AI.
The extreme analogy is that robot instead of human should explore the nuclear incident of Chernobyl and Fukushima.
Now we're trying to build automated analysis of ECG interpretation using AI. In order to interpret standard 12-lead ECG waveforms signal of 5 minutes for Afib detection that's equivalent of 60 minutes duration. For long duration multi-lead Holter ECG it's at least 24-hour, but I think you get the idea.
You can hire cardiologist that interpret the ECG but you probably need sub-specialist cardiologists specializing in ECG interpretation for the best results but only handful of them exist in the world with ratio of 1:100K or 1:1M for experts:populations, depending on where you're living. These expert cardiologist would rather spent the time doing more meaningful exercises for them like pace maker & IPC surgeries, teaching future cardiologists, and having holidays once in a blue moon.Even they can mistake due to human errors and other limitations. Thus this exercise is tailor made for AI.
AdeptusAquinas · 8h ago
Seems really unfair to the AI
robaato · 7h ago
Douglas Adams: video recorders watched tedious television for you, thus saving you the bother of looking at it yourself;
Change over time. Get this done for 10y or more back non teleprompted lower edit recordings, chart and compare for some metric
iamflimflam1 · 8h ago
This would be very useful for understanding business people who answer questions with endless word salad answers.
grues-dinner · 8h ago
Reminds me of a flash game (branch) or tool, art or general "thing" perhaps (branch) I suppose it depends how you see it(return)(return) that I cannot not find (branch)such is the state of the modern web(return) (branch) something in the flavour of a Jared Tarbell piece (return) where you could type and the text would be displayed in a spiral in 3D space (branch) there was a way to make a branch a thought (branch) like this (return) and then return to the parent level. but even that wouldn't handle the weave (branch) not sure if that's a good thing or not (return)
aDyslecticCrow · 5h ago
Could we do this to Jordan Patterson? I feel like his branching really needs a git tree view to parse sometimes.
N_Lens · 4h ago
Can git handle recursion?
codegladiator · 3h ago
that depends on what you mean by "can" and "git" and "handle" and "recursion"
pointlessone · 6h ago
I wish people did more testing or at least acknowledged the limitations. I'd expected at list some level of diligence from Dave. This is half-broken in Firefox and looks like a regular nested list in Safari.
I understand it's made for personal use but if it's posted on the public web at least a disclaimer would've been nice.
GeoAtreides · 6h ago
works perfectly on my firefox
Liquid_Fire · 5h ago
I thought it worked fine on Firefox, but then I saw the screenshot of what it actually looks like in Chrome (elsewhere in the comments).
qwertox · 3h ago
An AI could generate them automatically and visualize the contradictions.
sorokod · 5h ago
Perhaps Sankey diagrams is a better way to visualise
ivape · 9h ago
We act like small models are inconsistent and incoherent, but we rarely point out that it actually matches certain mental states and capacity accurately. We may need to actually see, how would a 0.5B model handle the presidency, because … it could be accurate. Having a super large model simulate these things would not be authentic.
For example, it could be truly true that a developer is roughly as good as a 1.5b model. It could really be true, in which case we’re not valuing these models for their true simulation power (yet). Might be the best interview test, you must generate hand written code that’s better than a small model (or show better judgement).
For the presidency, the current benchmark to beat is GPT2 it seems.
AceJohnny2 · 8h ago
> For the presidency, the current benchmark to beat is GPT2 it seems.
You're generous. I would've gone with Eliza, but on second thought that's an insult to Joseph Weizenbaum.
grues-dinner · 6h ago
To be representative, the models need a stronger bias towards what the last person they talked to said was the truth rather than what was in their training data.
brador · 3h ago
I talk like this. It’s an ADHD thing. All RAM no HDD. Just swap file thrashing like it’s 1999.
neuroelectron · 4h ago
*made famous by Donald J. Trump
huflungdung · 9h ago
Silly “hit piece” for lack of a better term on Trump. Do this with many conversations and you will have multiple branches of topic.
bootsmann · 7h ago
This isn’t a conversation, he is holding a monologue.
DonHopkins · 4h ago
Moronologue!
conorcleary · 4h ago
holding it underwater
atombender · 7h ago
This wasn't a conversation between ordinary people like you and me, but a statement delivered to reporters by a sitting US president. So the fair comparison would be to other such statement made by other sitting US presidents.
fortyseven · 3h ago
Won't someone think of the orange felon?!
georgemcbay · 7h ago
Too much of a coward to post this on your main account, huh?
DonHopkins · 6h ago
Oh but it's ok for him as a public figure to do thousands of weaving slanderous hit pieces on everyone else from his bully pulpit, yet you still gleefully leap to his defense, because you hate the same people he does, since you're both thin skinned sociopathic bullies and cowards. Why are you so ashamed to post that under your own name?
https://wiki.improvresourcecenter.com/index.php?title=Harold
It could take 20 minutes until we reach the conclusion, at which point he finally explains what the purpose of the final formula is and why we want it.
I got the habit of reading his book in reverse before the lectures; reading section by section in reverse order from the end. This way the mathematical calculations had a clear goal and were faar easier to follow.
... brilliant maths, but he was fully and utterly incompetent at teaching it. And he had a bit of an ego about how many students fail each year because "they're lazy".
Me and a few friends did deep recaps to de-tangle the explanations using his book, rephrasing it in a easier and shorter format; and he accused us of cheating because our scores deviated from the normal distribution.
All this to say; sometimes clever doesn't correlate well to great with words.... though dont take that as a endorsement of trump.
Intro to one of the maths books I had to reference to do my masters thesis. :)
I can’t learn this way. Great description of some of my instructors. I gravitate towards design and engineering. Goals are first. This sounds like play: “…and this happened, and this happened…” nope. Not me.
Trying to search that post.
Edit: in the discussion there was a link to a do a YouTube video where to movie characters were playing word badminton with each other.
Edit: this clip https://youtu.be/swqfFHLck1o
Humans can indeed make sense. Not to be too Swiftian, but in some countries, children even go to school for it.
For semantic analysis, however, git is just not the right tool. It's a chronological graph that affords diffs. For code.
We need Python NLP and spacy here. But even the best tooling won't get far. A compiler would abort nonsequential logic and unsatisfied contracts and grammar.
An important business presentation would have structure and facts. Inside the theoretical classroom, a public speech is different from casual, random remarks. Unless the speech is intended for entertainment (e.g. comedy, theatre) or some dark usage, such as propaganda.
From TFA:
> the cyclical pattern of his speeches, little snippets of “the best words” and talking points assembled like a ransom note cut from a magazine
That's gold!
I wouldn't want to torture the author by force-exposing him more to Trump's inanities. That's already 5x longer than I can stand before the urge to tear my ears off becomes unmanageable.
Actually this type of applications should be the killer application of AI.
The extreme analogy is that robot instead of human should explore the nuclear incident of Chernobyl and Fukushima.
Now we're trying to build automated analysis of ECG interpretation using AI. In order to interpret standard 12-lead ECG waveforms signal of 5 minutes for Afib detection that's equivalent of 60 minutes duration. For long duration multi-lead Holter ECG it's at least 24-hour, but I think you get the idea.
You can hire cardiologist that interpret the ECG but you probably need sub-specialist cardiologists specializing in ECG interpretation for the best results but only handful of them exist in the world with ratio of 1:100K or 1:1M for experts:populations, depending on where you're living. These expert cardiologist would rather spent the time doing more meaningful exercises for them like pace maker & IPC surgeries, teaching future cardiologists, and having holidays once in a blue moon.Even they can mistake due to human errors and other limitations. Thus this exercise is tailor made for AI.
The difference between posting an image url without and with a random string in the url.
Blue line is successful requests (people viewing the image I posted), green are unsuccessful requests (people trying to find other files).
Second blue bump is the screenshot with a randomized „hard“ to guess url. First bump the default iOS screenshot name in the url.
I understand it's made for personal use but if it's posted on the public web at least a disclaimer would've been nice.
For example, it could be truly true that a developer is roughly as good as a 1.5b model. It could really be true, in which case we’re not valuing these models for their true simulation power (yet). Might be the best interview test, you must generate hand written code that’s better than a small model (or show better judgement).
For the presidency, the current benchmark to beat is GPT2 it seems.
You're generous. I would've gone with Eliza, but on second thought that's an insult to Joseph Weizenbaum.