ACow_Adonis

Because I don't have a few billion dollars in my back pocket and even if I did, planned obsolescence and dark patterns are infinitely more profitable thus regulation is needed to achieve consumer positive outcomes?

You mean the internet you pay to access and which was around before the ads were even on it? That internet?

I'm not trying to be mean I'm just trying to historically parse your sentence/belief.

Because for me this is a simplified analogy of what happened on the internet:

a) we opened a club house called the internet in the early 1990s, just after the time of BBSs

b) a few years later a new guy called commercial business turned up and started using our club house and fucking around with our stuff

c) commercial business started going around our club house rearranging the furniture and putting graffiti everywhere saying the internet is here and free because of it. We're pretty sure it might have even pissed in the hallway rather than use the toilet and the whole place is smelling awful.

d) the rest of us started breaking out the scrubbing brushes and mops (ad blockers, extensions, VPNs, etc) trying to clean up after it

e) some of its friends turned up and started repeating something about social contracts and how business and ads built this internet place

f) the rest of us keep crying into our hands just trying to meet up, break out the slop buckets to clean up the vomit in the kitchen and some of us now have to wear gloves and condoms just to share things with our friends and stop the whole place collapsing

Even theoretical AI still has the other mind problem from economics.

Communicating and predicting desires, preferences, thoughts, feelings from one mind to another is difficult.

Fundamentally the easiest way of getting what you want is to be able to do it yourself.

Introduce an agent, and now you get the same utility issues of trying to guess what gifts to buy someone for their birthday. Sure every now and then you get the marketers "surprise and delight", but the main experience is relatively middling, often frustrating and confusing, and if you have any skill or knowledge in The area or ability to do it yourself, ultimately frustrating.

Surely the correct conclusion is to question the value/veracity of those degree issuing institutions and rituals?

And if you previously were unaware of the insanity and irrationality passing under the surface of such human activity, I guess it can come as a bit of a shock :)

The issue is that your standard is borderline setting up an impossible strawman, and when we actually do textual (and/or biblical) analysis and historical science, we never really have a "this is THE version of Q" nor do we have a "this is the FINAL/REAL version of the book" or "this is the final/absolute version of the hypothesis".

The idea that there is "one authoritative version and it was the version that was copied into this one authoritative derivative and we found the derivative so now we have to find that exact original or else its all bumpkin" simply isn't the way 2000 year old books or texts were written, copied or used. You will never find it because that's not what happened.

But we can lay out the texts side by side, arrange the narratives and see how they differ chronologiaclly from book to book, notice where particular linguistic quirks take place, notice where words are copied word for word in a particular order, where embellishments or insertions or changes are made, then just like taking several witness accounts, we build up a probablistic version of events that happened.

So there isn't "one Q", just as there isn't one authoritative version of mark, luke, john, mathew, etc. But there's patterns in the texts which strongly suggest that there was some kind of shared knowledge and a common source in the authors of the later gospels. We hypothesised that this common source seemed to be shared amongst the other gospels was a "sayings gospel", because the common ground that seemed to be repeated in the other books were primarily sayings and the other bits seemed to come from mark, which at the time met the problem that people didn't accept that such a book or source would actually exist because we'd never seen one before.

Then, after this hypothesis was formed and that objection raised, with the discovery of the nag hammadi library and the gospel of thomas, we found an actual historical sayings gospel. A confirmation that this type of literature did exist and was written in early christian communities. It was not Q, but it confirmed the hypothesised genre and existance of early christian literature.

If you're waiting for the discovery of two literal peices of text, whereby the carbon copy of the first is deduced from the discovery of multiple other historical books to the letter that followed, well then you're setting up an impossible standard. Even literal transcription probably wouldn't meet that standard.

What on earth is it about intellectual property that breaks someone's mind so much that genuinely, when presented with a translation of a 2000 year old text that itself is based on another authors translation and who's translator is now dead, they go onto a website to proclaim "it's not in the public domain!".

In practice, I find it depends on your work scale, topic and cadence.

I started on the $20 plans for a bit of an experiment, needing to see about this whole AI thing. And for the first month or two that was enough to get the flavor. It let me see how to work. I was still copy/pasting mostly, thinking about what to do.

As i got more confident i moved to the agents and the integrated editors. Then i realised i could open more than one editor or agent at a time while each AI instance was doing its work.

I discovered that when I'm getting the AI agents to summarise, write reports, investigate issues, make plans, implement changes, run builds, organise git, etc, now I can alt-tab and drive anywhere between 2-6 projects at once, and I don't have to do any of the boring boiler plate or administrivia, because the AI does that, it's what its great for.

What used to be unthinkable and annoying context switching now lets me focus in on different parts of the project that actually matter, firing off instructions, providing instructions to the next agent, ushering them out the door and then checking on the next intern in the queue. Give them feedback on their work, usher them on, next intern. The main task now is kind of managing the scope and context-window of each AI, and how to structure big projects to take advantage of that. Honestly though, i don't view it as too much more than functional decomposition. You've still got a big problem, now how do you break it down.

At this rate I can sustain the $100 claude plan, but honestly I don't need to go further than that, and that's basically me working full time in parallel streams, although i might be using it at relatively cheap times, so it or the $200 plan seems about right for full time work.

I can see how theoretically you could go even above that, going into full auto-pilot mode, but I feel i'm already at a place of diminishing marginal returns, i don't usually go over the $100 claude code plan, and the AIs can't do the complex work reliably enough to be left alone anyway. So at the moment if you're going full time i feel they're the sweet spot.

The $20 plans are fine for getting a flavor for the first month or two, but once you come up to speed you'll breeze past their limitations quickly.

Ah yes, but that results in two problems.

Firstly a fish without legs objectively does not have legs, but we do not necessarily call it disabled, even though it clearly lacks a facility.

Secondly, the autism spectrum disorders are, as I previously mentioned, not obviously just about deficits of behaviours or functions but also can take in extended and exceptional abilities in some areas and greater sensitivities rather than deficits or lack of an ability, so it is not clear that the entire diagnosis can be defined by deficits or lacking things. The high functioning and Asperger's type diagnosis is not about a universal deficit diagnosis and we do not generally call neuro-typical humans disabled because they lack prodigious activity or interest in math, language, or other subjects, even though that can also objectively be measured and called a deficit.

But going by the strict notion of DSM-V criteria of providing a hindrance, we hit the somewhat problematic definition whereby a person can have autism at one point in their life (when it hinders them in a context), moves into another point or context in their life (where it does not) and therefore they do not or would not meet the criteria for having autism if they sought a diagnosis at that point in time, and then move back into another point or context in their life where it hinders them and so now they meet the criteria and presumably have autism again.

Now, needless to say, this is not how anyone actually thinks about psychiatric or psychological issues in practice, especially with conditions such as autism, and just highlights the relative absurdity of some of the diagnostic metrics, practices and definitions.

What we tend to do is tie the diagnosis of autism to the individual identity and assume that it is a consistent category and applicative diagnosis that stays with a person over time because it is biological. We know, of course, that this is despite not having any working biological test for it, and diagnosing it via environmental and behavioural contexts. And don't even get me started on tying in diagnosis of aspergers/autistic individuals with broadly differing abilities and performance metrics on a range of metrics under the one condition such that the non-verbals and low-functioning side of neurotypicals get lumped in with the high iq and hyper-verbal high-functioning aspergers as having the same related condition even though neurotypicals are closer to the non-verbals and low-iqs on the same metrics and scores.

The entire field and classification system, along with the popular way of thinking about the condition is, if i might editorialise, an absolute mess.

I think it depends on how one interacts with it. As far as I know it doesn't have a personalised feed and I'm seeing the same front page as everyone else. So I mainly use it to scan once or twice a day to pick out if there's anything going on in the world I need to know about.

Then for one or two threads I'll perouse the comments to see what our particular class of HN-esque people think about a topic. About once a month or a fortnight I might even post a comment. But it all has to be taken in context. Half of the time I'll close out the comments section immediately because it's clear the whole thing has gone down a tangent in not interested in hearing about. Another risk is when talking about topics that the HN crowd knows nothing about, which in my case is primarily economics where some of the takes are borderline delusional/ignorant and backed by a kind of tech worker/startup ideology.

The anti-politics thing is both a blessing and a curse. On the one hand it's one of the last sites on the internet where there is comparatively little vitriol and thankfully, comparatively little populism. On the other hand, it means defacto support for a dominant ideology and compressive censorship of anything that threatens that ideology, and obviously that ideology is the one that supports tech workers, startups and venture capitalists.

I think taking all those things into account you can still get value out of it but know what you're engaging with. But like the other forms of social media since the death of forums, it's not made for serious engagement or deep thinking on a subject, and discussion can't really be anything more than temporally ephemeral.

At the very least it's borderline whereas the other forms of social media can basically be judged to be explicit write offs in my opinion.

I find it much more intuitive to think of LLMs as fuzzy-indexed frequency based searches combined with grammatically correct probabilistic word generators.

They have no concept of truth or validity, but the frequency of inputs into their training data provides a kind of psuedo check and natural approximation to truth as long as frequency and relationships in the training data also has some relationship to truth.

For a lot of textbook coding type stuff that actually holds: frameworks, shell commands, regexes, common queries and patterns. There's lots of it out there and generally the more common form is spreading some measure of validity.

My experience though is that on niche topics, sparse areas, topics that humans are likely to be emotionally or politically engaged with (and therefore not approximate truth), or things that are recent and therefore haven't had time to generate sufficient frequency, they can get thrown off. And of course it also has no concept of whether what it is finding or reporting is true or not.

This also explains why they have trouble with genuine new programming and not just reimplementing frameworks or common applications because they lack the frequency based or probabilistic grounding to truth and because the new combinations of libraries and code leads to place of relative sparsity in it's weights that leave them unable to function.

The literature/marketing has taken to calling this hallucination, but it's just as easy to think of it as errors produced by probabilistic generation and/or sparsity.

As someone who has an android personal phone and an iPhone for work for several years, I literally do not know what the hell people mean by "polish", beyond just the informal emotional utterance that can be translated back to "what I'm used to". Half of the stuff in the iPhone is equally arbitrary and mindboggling as the android.

As a comparatively politically aware Australian, I had absolutely no idea who he is/was, but then I don't have any Twitter or general social media presence or consumption.

I don't know where you live, but it quite clearly is where I'm from.

Oooh, to be sure they don't call them IQ tests explicitly, but the psychometric capabilities and performance tests they've gotten me to do (mathematical, logical, verbal, reasoning etc) are pretty obviously IQ proxies.

Honestly, imo clinically in aggregate the actually score itself provides very little information beyond what a 5 minute conversation would achieve, and the result could be better thought as bordering on 5-6 level categorical variable rather than a gradient due to their biases and inherent individual patient variance on performance and test taking context.

The sub-sections of things like the WAIS can be of some value for identifying specific abnormalities or deficiencies, but as you said, is probably of more value clinically to split them out into separate tests/activities rather than to group them all together into an aggregate score. It's a bit like judging athletic ability and skill by BMI and fat percentage rather than just playing an opponent in tennis to find out if they're a good tennis player.

Well yes, but that's not actually empathy. Empathy has to be felt by an actual person. Indeed its literally the contrary/opposite case. They have to emphasise it specifically because they are reacting to the observation that they, as a giant congregate artificial profit-seeking legally-defined entity as opposed to a real one, are incapable of feeling such.

Do you also think that family values are ever present at startups that say we're like a family? It's specifically a psychological and social conditioning response to try to compensate for the things they're recognised as lacking...

Presumably for the same reason Google doesn't let you block or filter shit sites.

If you genuinely let user's preferences be taken into account, it's incredibly hard to make money from ads if the user's true preferences are not to be shown them.

The entire point of ads is to manipulate and change user preferences and behaviours.

So any preferences or customisation has to be minimal enough that their use can only partially implement user preferences. White listing is a step too far against the purpose of YouTube.

Thus Google will always be biased to not letting you implement full customisability and user control.

Except its a bit like that PERL quote.

You have a problem. You want to figure out a way to get people to pay for things like news, investigative reporting, art, community and positive externalities.

You think, I know, i'll use ads!

Now you have two problems.

So i guess overnight we have a new floor for unempirical projections of future project profit :D

A lot of made up numbers now have a min(50.0) equivalent somewhere :P

See this article: Why do econ journalists keep making this basic mistake.

https://www.noahpinion.blog/p/why-do-econ-journalists-keep-m...

Source: am economist, and the writer of the blog is 100% correct.

Reporting and commentary on GDP and economics stats is just generally bad.

Well, I think surely the entirely fair thing to do is to just admit we don't know rather than make any attribution or imply any possession of an answer to those questions?

Your countrymen literally voted in Trump, and they did it with broad support across many parts of the country. You can argue that a significant proportion didn't, which is true, but while democrats and the like show increasing distaste which can be explained along general partisan lines, on the other side is a significant number of Trump supporters who show themselves strongly supportive of and behind his general policies and proclamations. To pretend otherwise or call that fearmongering is to deny all reality.

Yes, this is product differentiation, features, demand elasticity, network/ecosystem lock-in, loyalty tax etc.

Just as software and internet also has a mixture of free to use, once off, subscription, usage metering and fixed cost revenue, so too does banking and credit services vary their fee revenue structure and sources for its various financial services and products.

It's very hard to make universal statements to how these are structured, but about the only thing I think I can say is that if you aren't paying fees, the commercial entity must be aiming to make revenue off of you being a customer in some way. Some savvy customers can get services with a net positive value for free, but the business at large scale is going to try to extract what value it can.

It's hard to give a universal statement on price structuring, but generally fees can help squeeze revenue out of sticky customers, customers who need your services, customers who do not provide enough net capital or volume-based revenue to cover your costs.

You can also use fees to explicitly target or select for particular users. For instance, in some money and credit markets or markets which offer various costly rewards you can sometimes split your customer base into good revenue sources and costly freeloaders. A fee structure can even help signal or self-select these costly customers out of a product or identify them for your other product lines.

I'm having a hard time wrapping my head around someone someone in business not understanding charging fees.

You charge fees to get money.

Banks are there to make profits.

You say that banks don't make much money off of it, but maybe your idea of "much money" and mine are different.

In my country, it is reported that bank fees make up approximately 5% of bank revenue, and I'm reasonably sure we have less fees than others. I don't know what business or country you're in, but there's a pretty good guess your country takes in more money in fees, and earning a marginal 5% on anything is not chump change.

May I humbly suggest that your premise is wrong. They ARE significant sources of revenue and profit and banks charge them because they are in the business of generating profits.

I attended a private school for a couple of years, and I have to say it was worse than the public ones.

Now obviously this is going to be neighbourhood, country, and community specific, but the problem I had with observing private schools was that now the school had an additional incentive not to expel students, rich and influential parents had extra influence over whether their child could be disciplined and how the school should do things, and half of the time the problematic behaved people were... the rich people and their children. Having and paying money isn't exactly a free ticket to well-adjusted children especially if the children are mimicking the culture they see at home and the society awards bullying and various behaviours with more money... Which most of ours do.

And this was on top of the downsides of private schools: being image obsessed over academics and intellectual investigation, surrounded by non egalitarian private school twats, and bunches of arbitrary private school rules. Now obviously this is not all private schools, but in the same way it's not all public schools either.

I think in this system it's a roll of the die. In my country, neighbourhood and in my life, my kid is currently going to public school and, touch wood... thriving for now. The other private schools around here have too much woo like Waldorf and Steiner and they steer away from evidence based methods in literacy and numeracy.

But I don't know if that's going to hold off into the older ages, and I can't promise, much to my wife's chagrin, not to consider homeschooling considering my own experience of high school also approaching that of a dysfunctional prison and a poor educational environment.

PS: there was plenty of interpersonal violence at the private schools when I grew up.

I genuinely don't think the tech medium is the problem. There's good and educational software available on tablets just as there is on anything else.

While I've always been anti-tablet myself, and anti-phone for similar reasons (though why a kid would need a phone is beyond me at this stage), there's programs for literacy and learning available on the tablets that work just fine, as well as art, creativity and painting. The "secret" is just no access to the toxic: we have one tablet, my 5/6 year old loves reading-eggs and Minecraft but we have no social media and no YouTube and no purchases, plus the entire house is behind an ad blocking firewall. My kid is allowed to watch videos and access apps like ABC iView/kids, but from the open web they have to be ripped and placed on our local media share drive before they're viewable. No YouTube or YouTube kids obviously because it's a toxic cesspool.

My experience is that without the trash\addictive nonsense, while the kids are drawn to the iPad, they'll also self regulate a bit and move between other activities of they're available and get bored of spending all their time on the pad.

In his room I've just built him a Linux computer, and that's restricted both in time (can only log on to his account at certain hours), has no native web browser, and requires me to install programs and updates. He can access the games we approve but like everything else in the house it's basically a social media and almost ad free zone.

I've introduced him to modded Minecraft because it's basically Lego... And though it's frustrating that Microsoft took it over, we basically just ignore the Microsoft side of things and just operate on the free community modded side of things. I showed him how we could change his sword and I've got a whole modded world where we can set ourselves up and play on LAN together. The other game he loves at the moment is beyond all reason, a community made RTS game I've been involved with, and he just won his first match against the simple AI (though I did have to talk to the other Devs to allow us an account from the same IP so we could play together). The goal of all this is to give the message that technology is all about being a creative tool you choose to use to make what you want, not a medium about being a passive recipient for advertising or marketing or endless video consumption, feeds, recommendations or outrage cycles. No one needs any subscriptions or recommendations or likes or feeds. But there's nothing inherently toxic about a screen or a tablet, those are choices we make (or more accurately are pushed upon and chosen for us if we're not careful).

I think there's a non-zero chance that I'll be setting up a local server for the neighbourhood kids over the next year or two so they can all play together without worrying about the wider internet and I'll probably have to work to admin that in some way.

Wait wait wait... You're either misinterpreting the article or reading extra implications into it.

The observations of the pacific cargo cults are an example of religious and cultural syncretism which took on many varied and unique forms most commonly with the sudden increased presence of colonial and military forces in the Pacific along the very varied cultures and groups in that area especially during and after WW2 (and their subsequent sudden departure).

But it is similarly a step too far to imply this phenomenon or the resulting cults that developed was deliberately cultivated by colonialists or that military presence.

The "cargo" elicited in the term cargo cults is directly tied to this phenomenon: WW2 saw suddenly huge amounts of goods and logistics suddenly appear and then disappear from the Pacific regions as the war was fought and then subsequently finished.

Now there are also examples of religious syncretism that forms from missionaries trying to introduce Christianity to places all over the world (see South American and African interpretations of Christianity blending with local traditions), but those are not the cargo cults referred to that explicitly capture the primarily WW2 Pacific phenomenon even though there are other examples of syncretism of Christianity and Pacific religion that aren't cargo cults and aren't deliberately cultivated. Indeed, many times I'm guessing some of these practices are explicitly meet with resistance and annoyance from the likes of colonial missionaries and authorities, so as with most things its not that simple.

Yes, that's fair. You can generally either go for the "1 time bet the house" bugs, or you can go for the salami slicing so small across many transactions and there's minimal grey areas in between...

But as a general rule there's a trade off between signal generated by your winnings vs the frequency of your trading all balanced against your power and ability to manipulate and capture the market.

If you bet big enough and pull off the implementation successfully once such that you can get away in that one time (and ensure the system doesn't respond to cancel your winnings as it will often try to due to the cost benefit of paying you out), congrats. And you can go the opposite strategy and have lots of small winnings and fly under the radar and make profit just below the cost of entry or dealing with you, but you won't make squillions.

If anyone repeatedly makes big money, what is almost certainly going on is some other expression of market power, because if you don't have that power, the real market power will make the calculations on shutting you down once you're detected.

I don't have any immediately at hand, so whether you believe some random guy on the internet from here on out on is up to you.

I actually did a little bit of private investigation myself into betting markets of a handful of sports a number of years ago out of both curiosity, and potentially profit. And I'm currently working on perceptions of skill and merit along various systems just for my own intellectual project.

I find it interesting you've discovered the systematic bias strategy of choosing your opponents and events (these are done both consciously by players in various systems, but often formally and informally supported or enabled by various governing bodies and competitive structures).

Aside from both the boredom (arbitraging and scraping stats can be meticulously dull especially in fields you have no interest in or if you lack a hustler or profit maximising mentality or significant bankroll), opportunity cost (how much you can earn in other markets with such skills given equal investment of time and safety), the moral question (you're taking money from other ignorant bettors and I personally consider a lot of gambling and exploitation of people's ignorance relatively immoral, the house/market maker in many contexts still profits just through sheer trading frequency or other profit making actions), these days you have many market makers targeting little sharps presumably because they want to keep their ignorant base happy and gambling and don't want the perception of playing agaist certain losses or against people capable of seeing through the systems to reveal internal operations.

That's the only reason I can come up with for why they target and close accounts of profitable gamblers while they're still taking the house edge on every bet as profit.

What I (of course) also informally discovered in some markets I looked into was evidence of various fixing activities and systemic biases.

Now if you were going to do this for large scale profit, one of the things you are going to have to do is stop the illusion you're modelling skill and start modelling corruption.

It's all well and good for these articles to come out and pretend skilled prediction will win you millions, but you can be the best statistician in the house and if you place bets in a game based on models of skill, but the game outcome is rigged, you can kiss goodbye to your bankroll.

Criminals win because their outcomes aren't statistical or based on skill.

And now you're suddenly in the business of modelling and fighting people who profit from rigging markets rather than morally neutral statistical models in games of skill.

You can of course model such and go down that rabbit hole, but at some point you either buy into such a world and get involved (either by being hired by or interacting with criminal and questionable enterprise) or you are effectively trying to do statistical surveillance of criminal activities... Which will then show up in the statistical patterns of your own betting activities and potentially alert said enterprises to your existence.

And that's about where I bowed out of my investigations and turned back to more theoretical work.

That's not particularly a world I want to be involved in, but I am relatively confident popular sports and markets have illicit match fixing and other activities going on in them at significant levels, but I'm not interested in fighting with or building implicit knowledge bases of people who have million dollar anonymous backers from Vegas, billion dollar contracts, and whom die from accidental sleeping pill overdoses... If you know what I mean ...

I'm a very cynical man who has also had the honour of working for the national statistics organisation, so take what i say with a grain of salt.

While there are lots of "real and theoretical gains" to be made in gambling, a far more cynical take is that the "advanced analytics" is the frontbusiness for the real money being made by criminal enterprises, insider-trading, front-running, special access advantages, arbitrage, regulatory and competetive inefficiencies and insider-jobs.

"No no, we're all really smart see, look at this math that you don't understand, that's how we make millions while being bankrolled by mysterious money and business partners from vegas, who also by definition can't understand the math and hence can't validate what we're doing but who are totally willing to just front us a bankroll :)"

Its possible that there are certain gains left on the table, especially in old and early markets, but be wary of anyone being bankrolled by external forces and earning millions/billions and needing to explain that wealth with "smarts".

Often i wonder whether these articles that appear aren't a form of marketing or reputation laundering, but i have absolute no proof and i'm not particularly interested in looking into it much further. So you know, i'm probably way off :)