We were promised Strong AI, but instead we got metadata analysis • Hacker News Board

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

#131

Post by dorgo » Mon, Apr 26, 2021, 2:48 PM UTC

cblconfederate wrote: post #37

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…

> websites ranking other websites

I can imagine how competitors are going to rank each other.

Everything is about reputation and trust. If reputation is solved then many other problems become easy.

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

#132

Post by solosoyokaze » Mon, Apr 26, 2021, 2:50 PM UTC

cletus wrote: post #126

soarfourmore wrote: post #49

The following quotes are fairly interesting and ironic: > Larry Page and Sergey Brin were originally pretty negative about search engines that sold ads. Appendix A in their original paper says: >> "we expect that advertising-funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers" > and that >> "we believe the issue of advertising causes enough mixed incentives…

At this point, the snide dismissal of all things advertising is nothing short of boring. To be sure there are many forms of advertising annoyance: auto-playing sound/video, remarketing (or what I like to call advertising a product I've already bought), interstitials, popups (to be fair, there are many non-advertising forms of these eg "sign up to our newsletter" dialogs) and so on. But what made Google a money-printi…

I don't see targeted search advertising as user friendly. A search engine should return the most valid results. As soon as you have sponsored results, there's a conflict of interest. What if the competitor to the neck pillow ad people actually have a better pillow? They should be the first result, but won't be since Google's interest is in helping advertisers, not the users of their search engine.

I don't think there's an argument where advertising is pro-user, since a service that focused on the user would return the best results for a search, not who paid for placement.

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

#133

Post by indeedmug » Mon, Apr 26, 2021, 2:50 PM UTC

0xffff2 wrote: post #117

mxcrossb wrote: post #57

Earlier quoted context omitted.

I only go to Allrecipes now for exactly this reason. It’s funny, people used to mock that site for the stupid reviews, but what matters can quickly change...

Allrecipes is a worthless time sync. The vast majority of recipes on that site are written by amateurs and just not good. It's not worth sifting through recipes on there to find a good one because the odds are so low.

Lately I been watching YouTube videos and finding recipes through that. I find that it's important to follow people with a track record of good cooking. It's no different than following authors who write good books. When people condense down their recipes into steps, you don't know how much experimenting they did or if they did at all. So you have to rely on their credibility.

I found Josh Weissman to have good recipes. You could also look for recipes by experienced chefs like by Munchies.

Also if you have more experience. You can look at a recipe and easily figure out if something is very wrong. Unfortunately, most things on cooking don't tell you how to "debug" recipes. You just cook often and hope you can figure things out.

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

#134

Post by dheera » Mon, Apr 26, 2021, 2:51 PM UTC

sseagull wrote: post #109

soarfourmore wrote: post #49

The following quotes are fairly interesting and ironic: > Larry Page and Sergey Brin were originally pretty negative about search engines that sold ads. Appendix A in their original paper says: >> "we expect that advertising-funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers" > and that >> "we believe the issue of advertising causes enough mixed incentives…

Wasn't this also the story of a dating site (whose name escapes me. OKCupid? PoF?)? Original owner wrote an article about how paying for a dating site is a bad idea. Money is offered, article disappears. Searching is failing me at the moment edit: Was OKCupid: https://www.themarysue.com/okcupid-pulls-why-you-should-neve...

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-month would make sense for dating, specifically.

I mean, if you are a dating app, you should be striving to get users to delete your app as fast as possible (for the right reason), not hang onto an annual subscription.

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

#135

Post by twodave » Mon, Apr 26, 2021, 2:53 PM UTC

marcosdumay wrote: post #90

tjr225 wrote: post #68

Earlier quoted context omitted.

> Voice recognition on the other hand keeps getting better because there's no tangible benefit for someone to game it. Knock on wood! It is creepy to imagine a world where computers have the upper hand on vocal inputs but I already sometimes feel this way with text and autocorrect...

> but I already sometimes feel this way with text and autocorrect... Just disable autocorrect. The amount of times where people communicate and a typo is a critical problem is aproximateley 0, you can manually correct them at those times.

I hate auto-correct because it doesn't stay in its lane. It tends to expect me to use the most common 10,000 or so words in English and will actually auto-bork totally valid words because it thinks it knows better than me.

I get whenever I'm using shorthand or weird acronyms or technical jargon that it might get confused, but when a word I type is both a) a real word and b) something I would use in every-day speech just leave it alone!

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

#136

Post by guerrilla » Mon, Apr 26, 2021, 2:54 PM UTC

cletus wrote: post #126

soarfourmore wrote: post #49

The following quotes are fairly interesting and ironic: > Larry Page and Sergey Brin were originally pretty negative about search engines that sold ads. Appendix A in their original paper says: >> "we expect that advertising-funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers" > and that >> "we believe the issue of advertising causes enough mixed incentives…

At this point, the snide dismissal of all things advertising is nothing short of boring. To be sure there are many forms of advertising annoyance: auto-playing sound/video, remarketing (or what I like to call advertising a product I've already bought), interstitials, popups (to be fair, there are many non-advertising forms of these eg "sign up to our newsletter" dialogs) and so on. But what made Google a money-printi…

> But what made Google a money-printing machine is that search advertising is actually largely aligned with the interests of the user. That is, just by searching for something the user has shown an intent that other advertising doesn't have (where generally it's just attention thievery). Imagine I search for "how do I sleep on an airplane". Isn't a neck travel pillow an appropriate result here?

This doesn't hold water. It's just being shown because someone paid for it to be, not because it's the best thing to be shown which is what algorithms would be tuned for if they were in the users interest.

> I get that it's popular to just hate on all advertising but that's just shallow.

That's pretty dismissive of all the thought that has gone into criticism of advertising and it's effects on products and services, without even giving a hint of an argument as to why you feel it's shallow.

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

#137

Post by jetbooster » Mon, Apr 26, 2021, 2:58 PM UTC

dheera wrote: post #134

sseagull wrote: post #109

Earlier quoted context omitted.

Wasn't this also the story of a dating site (whose name escapes me. OKCupid? PoF?)? Original owner wrote an article about how paying for a dating site is a bad idea. Money is offered, article disappears. Searching is failing me at the moment edit: Was OKCupid: https://www.themarysue.com/okcupid-pulls-why-you-should-neve...

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…

The problem is, for the _business_ the incentive is the opposite. You want the suckers who are willing to pay for your dating app to keep paying, so from a purely callous point of view you want to provide the absolute minimum benefit over the non-paying users that is required in order for them to not leave and try somewhere else. There is almost no incentive for them to _actually_ match you with someone, just string you along just enough to keep you coming back.

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

#138

Post by jodrellblank » Mon, Apr 26, 2021, 2:58 PM UTC

cletus wrote: post #126

soarfourmore wrote: post #49

The following quotes are fairly interesting and ironic: > Larry Page and Sergey Brin were originally pretty negative about search engines that sold ads. Appendix A in their original paper says: >> "we expect that advertising-funded search engines will be inherently biased towards the advertisers and away from the needs of the consumers" > and that >> "we believe the issue of advertising causes enough mixed incentives…

At this point, the snide dismissal of all things advertising is nothing short of boring. To be sure there are many forms of advertising annoyance: auto-playing sound/video, remarketing (or what I like to call advertising a product I've already bought), interstitials, popups (to be fair, there are many non-advertising forms of these eg "sign up to our newsletter" dialogs) and so on. But what made Google a money-printi…

> "Imagine I search for "how do I sleep on an airplane". Isn't a neck travel pillow an appropriate result here?"

No?

I mean, seen through the lens of extractive capitalism where "how do I X" is the same as "what product do I buy to do X", and "someone asking about X" is the same as "which product to shove in their face to make them stop asking and extract the most money out of them", maybe yes. Doesn't "information technology" suggest some alternatives? Like, information about sleeping in planes - noise reduction, positions people have found comfortable, stress reduction, light pollution, circadian rhythms, stretches that can be done in a small space or sitting down, etc?

> "As someone who has search for a lot of home furnishing stuff recently I can tell you a big problem is affiliate link blogspam. There'll be some real-sounding domain like mattressreviews.com but it becomes pretty clear it's just mass-produced "content" to justify affiliate links."

This seems to fly in the face of your previous paragraphs: you searched for home furnishing stuff, isn't some generic advertising of a mattress an appropriate result here? You want something better than that for yourself, but think other people don't deserve better and are shallow for complaining?

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

#139

Post by h2odragon » Mon, Apr 26, 2021, 2:58 PM UTC

zpeti wrote: post #88

cblconfederate wrote: post #37

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…

Sometimes I have this scary thought that paid search results are actually inherently better than organic ones, because when you spend money you do need to be relevant to the query and serve up a decent result. When it comes to organic, everything is free, so you try whatever hacks the algorithms, from keyword stuffed content to link spam etc. So as time goes on organic search results will actually get worse and worse…

https://en.wikipedia.org/wiki/Gresham%27s_law

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

#140

Post by dexen » Mon, Apr 26, 2021, 3:00 PM UTC

dheera wrote: post #134

sseagull wrote: post #109

Earlier quoted context omitted.

Wasn't this also the story of a dating site (whose name escapes me. OKCupid? PoF?)? Original owner wrote an article about how paying for a dating site is a bad idea. Money is offered, article disappears. Searching is failing me at the moment edit: Was OKCupid: https://www.themarysue.com/okcupid-pulls-why-you-should-neve...

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…

Counter-intuitively this might be about hedging the incentives for the service provider - to avoid the moral hazard of pushing for indefinitely extending the subscription.

Just as you mention, successful finding a partner means as few "attempts" (apologies) as feasible, which in turn means two "lost customers" to the platform. That introduces a perverse incentive for the platform to "spoil" the dating to keep the customers. By making one long-spanning plan, the perverse incentive is lessened.