Live data from Hacker News

We were promised Strong AI, but instead we got metadata analysis

calpaterson.com

381–390 of 427 posts

Re: We were promised Strong AI, but instead we got metadata analysis

#381
post #242

Earlier quoted context omitted.

Anything you measure gets destroyed at scale.

Anything your measure for the purposes of providing a benefit...

What is measured that is never used is effectively never measured at all...

Re: We were promised Strong AI, but instead we got metadata analysis

#382

The problem is that google created a negative feedback loop with the web. People have a tangible interest to game google, which worsens the quality of their AI datasets, which makes their AI suggestions terrible. There was the expectation that google's AI suggestions and info boxes would improve over time, but i 've noticed them getting worse and consider them a permanently broken gimmick now. They probably have the…

> There was the expectation that google's AI suggestions and info boxes would improve over time, but i 've noticed them getting worse and consider them a permanently broken gimmick now.

A funny example I saw recently is searching for "can men get pregnant".

Re: We were promised Strong AI, but instead we got metadata analysis

#383

Earlier quoted context omitted.

I literally append "site:reddit.com" like in the article whenever I'm looking for reviews and comparisons. Google is nearly useless for finding content among the sea of crap and autogenerated near-crap (like Slant). I might click on the links going to sites that I half-remember by name, but for open-ended searches it's a lost war.

> I literally append "site:reddit.com" Stop saying that! The more people repeat that on HN, the more we risk advertisers realizing it and doing reddit-targeted SEO and reddit will be useless too! (I'm only half joking) Also, it's funny, I was just reading the thread about malaria eradication and DDT-resistant mosquitoes; the problem of SEO is eerily similar (any countermeasure is eventually defeated by evolution).

Evolutionary arms races are extremely common, especially between parasites and their hosts.

https://en.wikipedia.org/wiki/Evolutionary_arms_race

The same thing happens in our societies. It's a constant never-ending struggle.

Re: We were promised Strong AI, but instead we got metadata analysis

#384
post #251

The misunderstanding that leads to belief in strong AI is that meaning is somehow embedded in the symbols used to communicate it. Meaning is a natural process that occurs inside each of us, speech is just a symbol of that meaning, text is a symbol of that speech. Further, meaning is an ever-evolving, ever-mutating process much like the universe. Training on symbols cannot arrive at meaning, since the meaning isn’t co…

> The misunderstanding that leads to belief in strong AI is that meaning is somehow embedded in the symbols used to communicate it

yet you have Roget's thesaurus, where words are grouped into sets of words that have a similar meaning; it might make a difference when dealing in terms of these categories, instead of dealing with word instances; i mean if you have a dependency graph of a sentence, and the nodes are the corresponding sets in the thesaurus, then you might get a similar, but more general interpretation of that sentence.

(i once had a project for parsing/representing Roget's thesaurus in python: https://github.com/MoserMichael/roget-thesaurus-parser There are probably other projects like that around)

Re: We were promised Strong AI, but instead we got metadata analysis

#385

Earlier quoted context omitted.

For one, that Ad logo spent about two years being yellow-on-white, so people without well-tuned contrast monitors couldn't see it at all. Last year they made a change that made them so indistinguishable to the untrained eye that even Google rolled them back: https://techcrunch.com/2020/01/23/squint-and-youll-click-it/ https://www.cnbc.com/2020/01/24/google-will-iterate-the-desi... But no, generally, they don't notice…

it normally takes me 1 sentence max to realise i've clicked on an ad by mistake. I think even if there was no ad shin i would still manage to work out what is promoted content. If ads ever stops being generic, irrelevant clickbait, we might have a problem on our hands. But that particular danger doesn't seem to be very likely at all.

> it normally takes me 1 sentence max to realise i've clicked on an ad by mistake.

Good thing that has no cost, then. Especially with how poorly optimized all the javascript on these pages is, slowing loading to a perceptible crawl (unless, perhaps again, if you’re in the bubble where every machine you use is latest flagship android or a workstation, blind to the experience of the less ‘sophisticated’ users).

Re: We were promised Strong AI, but instead we got metadata analysis

#386
post #134

Earlier quoted context omitted.

From the article > 12-moth plan > 6-month plan Why would a dating site have a 12-month plan, and why would a user of a dating site want a 12-month plan? Not only would you hopefully want to be off the site within 12 months, as soon as you found someone compatible, you would hopefully delete the app, but you've unnecessarily paid for months you will (hopefully) never use. I don't understand why anything but month-to-m…

I'm building a dating app, SwanLove ( https://swan.love ). It's still in MVP mode and centralized mode. I'm thinking of pivoting into this kind of business model: B2B. So I make my dating app something like GitLab or WordPress. You can install it and host it yourself. You pay me every month if your users exceeds 100. Say, you are a priest or a gym owner. You have a community. You want your people in the community (ch…

This sounds perfectly ethical to me but it's not clear what value it provides. People who are part of the same church or gym already have the opportunity to get to know each other by being in the same physical community in the first place and don't need a website where they can interact with the same people but online. If they do want that, they're likely to just set up a Facebook group. While the obvious disadvantage there is now Facebook owns their data, but it's also free and most people are going to choose free.

Ironically, online venue for a real-world network limited to verified members of that network was what Facebook itself originally was, until they realized opening up to everyone was the difference between a novelty for college students and a multi-trillion dollar world eater.

Re: We were promised Strong AI, but instead we got metadata analysis

#387
post #339
post #221

Earlier quoted context omitted.

> This puts the business into the situation where they have an incentive If the payment is a one-time advance payment, I would imagine this disincentives the business to truly do their best, since they already have your money. I would think, idealistically, maybe the best model would be an advance payment but with a money-back guarantee of say half the payment if you don't find a match through them. Legally establish…

You don't need a refund. Just presume an efficient market of such companies, where the matchmakers who actually make matches have better reputations, and the ones who don't quickly go out of business. That's how most single-shot service-provider businesses work. (At least, the ones with clear success criteria. Psychics and the like never lose reputation, because there's no standard to measure their claims against.) T…

> Just presume an efficient market of such companies

And those companies are in the business of matching spherical cows in a vacuum.

Efficient markets are useful as a simple model, but you don’t get to wish away real-world problems by pretending the world conforms to that model.

Re: We were promised Strong AI, but instead we got metadata analysis

#388

Earlier quoted context omitted.

Holocaust is not due to capitalism. It's actually due to socialism. "The National Socialists German Workers Party"

That is the same thing as saying North Korea's failures are actually due to democracy. "The Democratic People's Republic of Korea" Obviously the name of an organization doesn't mean jack compared to their actually implemented and enforced policies. The first things the Nazi's did when they gained power was kill all the socialists, communists, and unionists. They are anything but socialist.

This is a common rewriting of history in the last few years. Hitlers speeches are rife with rants against capitalism:

"We are socialists. We are the enemies of today’s capitalist system of exploitation … and we are determined to destroy this system under all conditions.” ~ Hitler

Re: We were promised Strong AI, but instead we got metadata analysis

#389
post #387
post #339

Earlier quoted context omitted.

You don't need a refund. Just presume an efficient market of such companies, where the matchmakers who actually make matches have better reputations, and the ones who don't quickly go out of business. That's how most single-shot service-provider businesses work. (At least, the ones with clear success criteria. Psychics and the like never lose reputation, because there's no standard to measure their claims against.) T…

> Just presume an efficient market of such companies And those companies are in the business of matching spherical cows in a vacuum. Efficient markets are useful as a simple model, but you don’t get to wish away real-world problems by pretending the world conforms to that model.

I’m talking about a game-theoretic dynamic that’s easily observed in real world professions (e.g. plumbers, cleaners, art conservators, etc.), and even in exactly the same industry (human professional matchmakers.) The “presumption” here isn’t really much of a stretch.

It just-so-happens that dating sites don’t currently follow this model, because an external force (Match Group) came in and explicitly chose to consolidate the market into a cartel, where 95% of “competing” dating sites are actually in collusion due to shared ownership. But there’s no reason to expect that situation to last forever, any more than there’s reason to expect the dominance of the currently-dominant social network (MySpace/Facebook/etc.) to last forever.

Re: We were promised Strong AI, but instead we got metadata analysis

#390
post #339
post #221

Earlier quoted context omitted.

> This puts the business into the situation where they have an incentive If the payment is a one-time advance payment, I would imagine this disincentives the business to truly do their best, since they already have your money. I would think, idealistically, maybe the best model would be an advance payment but with a money-back guarantee of say half the payment if you don't find a match through them. Legally establish…

You don't need a refund. Just presume an efficient market of such companies, where the matchmakers who actually make matches have better reputations, and the ones who don't quickly go out of business. That's how most single-shot service-provider businesses work. (At least, the ones with clear success criteria. Psychics and the like never lose reputation, because there's no standard to measure their claims against.) T…

The efficient market hypothesis is wrong.

Here’s why: it assumes that the only form of power or leverage that exists is supply and demand. However there are all kinds of forms of leverage in the real world. There is legal power. Voting. Guns. Unions. Price fixing. Cultural norms. Marketing. Blackmail. All of these are forms of leverage and they are not special cases; rather, supply and demand is one special case which comprises a fraction of the total pressure on wages and prices and success or failure at any moment.

Post reply on HN