The utility is a fairly robust signal about the probability of certain events occurring. It's pretty helpful for grounding your beliefs about the world (in certain domains).
For example, I had a lot of people ask me about the Hantavirus and if it was worth worrying about. I didn't have to do any research other than look at the various prediction markets and see them all at 5-10% to determine it was probably a nothingburger. Much quicker than having to parse a bunch of fearmongering news reports and such.
Even for that function, I think their benefits are overrated. Someone could easily manipulate that number if it fits their purposes (ie a health tech startup trying to peddle some cure). The numbers given to you are an illusion of understanding that does not replace actual research.
Smaller markets are easily swayed by not so much money. Arbitrage is difficult because there is an inherent threat that you're betting against insider information.
In this specific case it's clear that the market doesn't represent our collective knowledge about hantavirus at all. Twenty days ago the odds of a hantavirus pandemic being declared by the WHO were ~10% and now it's 5%. This is just a reflection of the news cycle, it would be nuts to say that this is a reliable estimate for the underlying scientific facts.
That is not surprising. If your point is that 10% is not 5% and the odds change over time then you have unreasonable expectations. It's not an oracle. Where did I say it was? I said it's a quick and robust signal. Of course there's plenty of ways to get alpha, but those require both effort and domain expertise.
Is this a standard you hold to any other single signal? Name it. What signal (a) predicted <3% over a month ago and (b) you believed prior to be a reliable signal?
There is no substitute for effort and domain expertise, that's my point. Polymarket would like you to believe that their odds represent some sort of collective knowledge, but they don't, it's not like the world's top experts in epidemiology are trading on this issue. Worse, as I explained someone might be trying to manipulate the odds to their benefit, which for the particular case you cited would be quite cheap to do.
This is an insane response to someone having their carefully written work casually bastardized by an LLM that rewrote the entire design spec without even being informed. The amount of institutional noise generated by such carelessness far exceeds whatever improvement in readability you could possibly imagine. Any criticism you could aim at the original text that you don't even have on hand (i.e. are completely speculating wrt its readability) you could direct 100x over at the manager's horrible communication skills.
You're assuming malign motivations, I'm assuming misplaced ones. It seems more likely to me the manager tried to read it and struggled, then generated something of equivalent size or larger. I'm taking it the generated document passed around was actually at least as large as the one-pager, and hence entirely pointless to rephrase even with the malign motivations you're assuming.
Since the poster here wears his personality and writing motivations on his sleeve, it is very obvious to me that he writes at cross purposes with those who read. he says very clearly: he writes for precision, expended a vast cognitive effort per word.
Even if, in this instance, my analysis is wrong -- its a comment for the poster here worth considering. Because people don't like to read writing which has taken such effort to produce, because it then requires a great effort to read.
> It seems more likely to me the manager tried to read it and struggled, then generated something of equivalent size or larger.
Either way, it's poor management to interpose oneself between employees. As a manager you should be connecting groups of people to talk to each other directly, not injecting oneself as a go between. If they have issues understanding the material they're much better off asking the OP directly than asking the manager who doesn't understand it either. And they'll be in a much better place to do that if they have read the material OP actually wrote.
> it's poor management to interpose oneself between employees.
I didn't interpret mjburgess as defending the manager or even condoning the action. In fact, I read their comments as recognizing that action as a failure.
The difference is that mj was trying to give advice to donatj, and donatj can't control what their manager does. So the advice is crafted such that it gives actionable suggestions to donatj.
Yet, that might not be the correct interpretation. I don't know, I'm some third party, like you. Personally I agree that this is poor management but I don't think just blaming the problem on the manager solves anything, it just leaves the problem broken. So the things to do are either fix the problem or figure out how to work with the broken thing.
> So the things to do are either fix the problem or figure out how to work with the broken thing.
So, I will say that if you did not seen or read a text in question, there is no way for you to accurately diagnose issues with the text and give out advice on how to make it better. Such advice from someone who simply assumes the way the text was wrong based on some manager rewriting it with ai is less then useless.
It is, frankly, ridiculous to think one can give meaningful advice about text you have never seen. And then double down in comments.
The feedback is to the author of the post complaining (understandably) about their manager using AI and destroying the carefully written document.
That post alone is plenty to give feedback on absolutism and the nuances of existing in the world with mostly neurotypical people. [My interpretation of the feedback]
We dont care about the manager, they don't matter. This is not "defending" or "justifying" the manager, in case you see it that way.
> It is, frankly, ridiculous to think one can give meaningful advice about text you have never seen. And then double down in comments.
I'm not the manager, there's no doubling down.
As for answering the question, you're right and wrong. You're right that I don't have enough information to give specific and nuanced suggestions. Like you said, I didn't see the original doc nor the AI rewritten one. I only have OP's comment. But you're wrong because I do have enough information to write the advice that I did. The advice is still broad because that's all that I can give with the information I have, but that doesn't mean it is useless.
You accuse me of not reading, but I'd encourage you to read my comment again. I explicitly stated that I am not advocating for what the manager did. I explicitly stated that I think it was wrong to do. But we're trying to solve problems, right? I took OP's comment as a legitimate request for help. Maybe they were just complaining and wanted someone to validate them. But hey, they posted in public and so are going to get a wide range of interpretations. That can be helpful, but may not be what was intended.
So either we try to respond to each other in good faith, acting as if everyone here is not acting maliciously, and trying to do their best, or we fight. I'm sorry, but that's what the situation is. If you just want to poke problems with things, then you're not helping. There are issues and limitations, and you're more than welcome to point them out but don't treat the setting like we're enemies. We're not. If you can't figure out how to critique while being collaborative then you aren't going to work well with others. And who knows how effective what I've said is. All I have to respond is your one comment that feels out of place to me. IDK if it's an off day for you, you didn't read right, I wrote wrong, or one of a million other things. But I'm trying to work with others (including you), are you trying to work with us?
Well, OP can learn from the experience or turn it into a hill to die on. Learning doesn't imply you were ever wrong, only that something you did produced an unintended result -- people are themselves problems to navigate around, not people whose actions you have to read as judgements.
The op is probably not the world's most reliable narrator. Based on his very, very specific preferences around writing I'm guessing he can be a bit prickly when it comes to feedback. The manager might have had a bad day and dreaded having the conversation and ended up with this. Still very stupid, but OP is not quite so clearly the martyred hero in this version.
When I read documentation, I'm not there to enjoy the experience. I'm there to find out how the documented thing works and how to use it. It's not a novel. I'm not there for entertainment.
Chasing readability without maintaining accuracy is a failure in the context of documentation no matter the motivations involved.
I'm not saying that readability can't be a consideration when making documentation. I am saying that if you discard accuracy in the process, you've fucked up quite badly.
This anecdote would likely be very different if the AI-modified version had been passed back to engineering for a review before sending it out.
> I'm not saying that readability can't be a consideration when making documentation. I am saying that if you discard accuracy in the process, you've fucked up quite badly.
You're right to elevate accuracy to a high level of importance, but that is NOT ENOUGH if the thing is has poor readability. The audience has to be able to understand the document if the document is to be useable.
There's only a certain amount of effort anyone can deliver in producing a document. But if the author can't deliver readability, they need to follow up the document with a lot of support and/or get some help to make it useable.
I've struggled through some absolutely awful documentation over the years. I'll put up with incredibly broken English and other problems as long as the accuracy is there. Just last week I encountered a pinout diagram that used emojis to indicate which pins related to which data channel. Not a choice I would have made, and I found it made the diagram harder to read. But it was accurate - I wired it up per the diagram and everything worked as intended.
Documentation lacking accuracy is useless. It can be the most readable thing ever produced, but if it describes a different thing than what was intended to be documented, it's trash. Documentation that is hard to read but is accurate still has value.
Regarding "follow up the document with a lot of support" - did you catch the part of the anecdote where the author is having to deal with support requests because of the inaccuracies?
I would assume the manager's a terrible reader and writer.
Most people are. Most managers are.
It's one of those upsetting things I've learned about the world as an adult, that's sharply contrary to what I believed as a child. I kept being surprised by all kinds of things until I began really to appreciate that, simply, most folks aren't especially literate. Even ones who attend and attain degrees from universities—surely at least nearly-all of those people can read and understand "college level" texts with some fluency? But no, that's, somehow, not even close to true.
> It seems more likely to me the manager tried to read it and struggled
Well obviously the manager struggled to read and understand it and also struggled to read the AI results. The clear root cause is utter incompetence on the managers part. Any feedback they wanted to give could be given, but you NEVER pass along modified documents as though they where what another author had created AI or not. If you don’t like it you can provide feedback or ask to coauthor another document, you never just append TFTFY with changes and send it along with the original author standing to take the consequences of whatever you cooked up. It’s just complete incompetence on the part of the manager.
I see this behavior in many people, usually conflict averse. In a poor attempt to mediate, spread incompetence like butter on a slice of bread. Ranges from tiring to infuriating.
I thought the reply was generally helpful. Something to consider about in my equally exacting wording as I share the same frustration as the original comment and this give me a framework to view possible issues with my own writing. I.E. You can't change what others will do, you can only change what you yourself do. In this case: Carefully crafted exacting documentation is being ignored = frustrating to me = can't change if others don't want to read it =;;; sorry I have a more elegant way to do this: My meaning is thus: While it is sometimes easier and apt to blame others for their actions, blaming others doesn't actually contribute to any meaningful growth or change. If you take on the blame yourself, even if 100% of the blame falls on the somebody else, then it leads to open ended questions on how that process can be better. Given that you have no control over other people, blaming yourself shifts the issue back onto you for a solution. This can reveal a treasure trove of oppurtunities not before explored. It can be as simple as understanding that there are different levels of technical documentation: How-tos, vs explanitory, vs laymen, etc. Or it could lead to a different exploration as to: How did I end up in this situation, what is the mistake that *I* made? Which could be an easy fix or it can be a philosophical or temporal fix. I made the mistake of:
+ Assuming people cared about this as much as I do
+ Allow another person to control then narrative: (I could have sent it out to stake holders myself; and bare whatever consequences from my hiearchy)
+ Not written any documentation and given the endpoints to an AI to communicate to laymens (because I may or may not have communication skills)
+ Take a course in communication
The list goes on and on, but the beauty is that sometimes it's truely and deeply philosophical such as, because I trusted somebody who wasn't to be trusted; because I'm in the wrong place and *know* I know I should be here.
Shifting the blame to the self is less about accepting blame and more about introspection and it is the most valuable lesson I learned from my wife when we first started dating. (It help me identify that as a person I tend to blame others first before blaming myself, and to spend 10 years practicing the muscle to reverse that order)
TLDR: You have willpower, use it by taking ownership over yourself. This is a learned skill and is not enate and requires breaking preconceptions and stepping out of yourself to find.
Based on other comments in the thread but not any direct reply to mine. I would also express that I was surprised when a coworker of mine complained that nobody read message boxes we put up to help the user. It was my first corporate job and I had already learned and ingrained from my experience at a small office that nobody reads technical instructions either. That alas also requires training. Usually by having the documentation open and doing exactly what is written with them watching, or with them doing (better). (Helps reveal gaps in documentation such as, Oh most users don't know how to traverse a file system, let alone what one is... how?, It's an analogy to office filings which they did everyday? why??? I never understood but, alas I've never been able to teach somebody who doesn't understand the file system, the file system.... my weakness)
Absolutely. I consider that to be primarily Biden's fault for not announcing in advance that he would not seek a second term. After that point, I think each decision made was the best that could be done at the time to minimize the damage.
There have absolutely been cases of VPs becoming President without ever winning their own primary though, and I doubt most would describe those cases as DEI despite demographics often playing a large part in VP picks.
Though Harris was an unserious VP pick in the first place (2020). Given that Biden was 183 years old at the time, he should have picked a VP that Americans or at least Democratic voters had demonstrated at least moderate acceptance as a President in the primary, instead of picking essentially the least popular Democrat in the race (just to pander? Why else?). I guess the DEI dogma told him that it's better to have a Black woman on the ticket even if she was the worst choice by any measure: ability to get votes, relatability, or political experience. The funniest part is that Harris was most unpopular in the primary with the 'wokest' Democratic voters -- they hated her for being a decent D.A. and charging criminals with crimes, even ones who were 'disadvantaged minorities.' DEI forced her selection anyway because she checked two identity boxes.
Indeed. The biggest election win she had outside of San Francisco prior to her coronation as the nominee in 2024 was a Senate special election where she drew 40% of voters. 3 million Californians voted for her out of 7.5 million voters. California has 39 million residents, but about 5 million are non-citizens.
Actually more Californians voted for the Republican against her in the 2014 election for attorney general, than voted for Harris when she later ran for Senate in the special election.
Obama by contrast had won 3.6 million votes, in a smaller state, for a decisive 70% win in his Senate race.
Harris was a joke of a candidate who was obviously unelectable outside of a deep blue state, but she was forced on us so the DNC could virtue signal. It was a slap in the face to every qualified Democrat, many of whom would have had a chance to defeat Trump (a low bar if there ever was one).
They do. They understand you cited ADL in lieu of an argument. If it could stand on its own you wouldn't need the citation and guilt-by-association. They understand the culmination of all the surrounding context reduces to a schmittian friend-enemy distinction where you are placing yourself as enemy. Everything else is sophistry.
I read what you wrote and read it as sophistry. You reply by adding more.
"It's why you believe [...]" But you don't know what I believe.
"Scott Adams claimed [...] because of a response to a survey question [...]" But his statement is, if one applies some very basic "media literacy" (as you like to call it), clearly rhetorical, with the underlying message that there seems to be a lot of racial hatred from blacks towards whites in the United States in 2023, and that this racial hatred seems to be institutionally supported, and that as a white person of means he'll use his means to avoid this racial hatred and suggests others do the same. The cited survey is merely one data point he presents to support this belief. Arguing as if he arrived at this conclusion purely off of that alone is total sophistry.
I don't live in the US, so perhaps that will give you some reprieve. Scott Adams might well have been wrong. I don't claim to know here if he was, just that you haven't actually contended with his position at all despite writing a lot of angry words, and that this excess of sophistry justifies a dismissive response.
Wait, there are hundreds of cases of American citizens being deported? I've only heard of the one guy (whose name I have in a text file somewhere). Where's a good list of the others?
This demonstrates where a lot of the mismatch in impressions of this tech arise. The thousandth amateur Wonderwall rendition is not at all interesting as a piece of recorded music, but for the performer (and those listening around them) it can be a fun and playful experience. The same could be said for AI generated music: it could be a fun and playful experience in the present moment, even if the resulting product is totally worthless to the market. This would still be a valuable thing for the human experience.
Arguably this is a return to a more traditional way of experiencing music from before the invention of recorded music. Before this, music was an entirely transient and often communal experience. Once the musician stops playing, the music is over. Songs from these times have largely unknown authors, and likely don't even have any single author or for that to even be a coherent concept. They were simply part of the shared culture that many had contributed to. Now music is owned by specific people and you can play back their performance as much as you like (for an increasingly insignificant price).
This tech may be a negative thing for the market of recorded music, but it needs to be argued that recorded music is the only authentic way to experience music, and that this is why that's how most people experience music currently, rather than that being an historical anomaly due to the technology available. Once you step away from treating music like it's only valid when it's a product for a market, the problems of AI music seem a lot less catastrophic.
Not that this affects the political calculus (where perception may as well be reality), but the cost burden specific to universal healthcare is actually opposite this intuition.
Things like obesity, smoking, and alcoholism all kill you before you can get too old. Healthy citizens end up using far more of the far more expensive end-of-life care, to the point where it outweighs the extra healthcare the unhealthy citizens use in their youth.
This (French) study [0] published in 2023 on data from 2019 calculates that the costs from legal drugs such as tobacco and alcohol, including higher helthcare spend during the life of smokers/drinkers, are still higher than revenue from unspent money on pensions and taxes, and cost of healthy person living years.
> It took me a weekend to write the initial Perl script that made this site. It took me another weekend to do the Rust rewrite (although porting all the content took two weeks). These are not complicated programs.
My last Hugo site took 30 minutes to deploy, not a whole weekend. Picked a theme, pasted in content.
> You want free web hosting? Hugo might be the right option.
An extremely good reason to pick Hugo especially if you don’t have the know-how to build your own SSG. You don’t need to know a programming language at all to use it.
Again, I have to throw criticism toward this idea that everyone who wants a static site generator already has the skills required to make one.
And I’m not saying it covers every use case like the kind of person who is willing to pay $100+ per year on a full blown solution like Shopify and Squarespace. It fits a niche: someone who wants their content online without coding with no hosting cost and doesn’t want to rely on third party platforms like Substack.
I have practiced "street photography" for years, where I purposefully take pictures of people on the street. Sometimes people ask what I'm doing, I tell them, and they say "cool can I see the pictures"? Sometimes I send them a file or whatever. No one's gotten all out of sorts over it.
For example, I had a lot of people ask me about the Hantavirus and if it was worth worrying about. I didn't have to do any research other than look at the various prediction markets and see them all at 5-10% to determine it was probably a nothingburger. Much quicker than having to parse a bunch of fearmongering news reports and such.
reply