Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

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

calpaterson.com

Post reply on HN
Print view View original

271–280 of 427 posts

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

#271

Post by disqard »

ajani wrote: 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…

Thank you for expressing this nuanced idea so well.

It ties in with deBord's "Society of the Spectacle" [0], a theory that societies evolve from Being, to Having, and ultimately devolve into merely the Appearance of Having.

Your point about the symbols being tools for communication (as opposed to the ineffable ideas being communicated) is also echoed in Lockhart's Lament [1].

[0] https://en.wikipedia.org/wiki/The_Society_of_the_Spectacle

[1] https://www.maa.org/external_archive/devlin/LockhartsLament....

Edited to add that this was extra thought-provoking:

> "It may even lie in between us, rather than in us."

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

#272

Post by semi-extrinsic »

sobellian wrote: post #151
cletus wrote: post #126

Earlier quoted context omitted.

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…

We know that paid advertising works, so it should not surprise us to discover that paid advertising on the Internet also works. But Google and other search engines seek to organize the world's information, not the world's commercial products and services. For a multi-multi-billion dollar company's core product, I'm somewhat surprised they cannot do a better job killing the blogspam. Given the resources at their dispo…

"We know that paid advertising works" [citation needed]

I mean, the ad business of course likes to throw various metrics around. But as far as I'm aware there are no proper randomized controlled trials that show statistically significant positive ROI of online advertising versus no online advertising.

I mean, it would be really simple to do, right? To provide conclusive proof of the efficiency of ads? Pick a populous state in the US where people enjoy Soft Drink X. Randomly divide the households in the state into two groups. For the next full year, run normal amount of targeted online ads for Soft Drink X in Group 1, no targeted online ads whatsoever in Group 2. Did the sales in Group 2 decrease by more than what the cost of advertising to that group would be, yes or no?

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

#273

Post by daenz »

uoaei wrote: post #240
daenz wrote: post #164

Earlier quoted context omitted.

>your product In a hypothetical world where I don't benefit from a competitive advantage, in what sense do I own the product / company?

Consider firms which are wholly owned by either 1) every single employee (worker-owned firm), 2) every employee and every customer who opts in (consumer co-op), or 3) literally every citizen of a government (publicly owned). You own it in partnership with the other people who own it. This is, in the strictest definition, the socialism that people are so scared of.

How much control do you have over the military of your government? Or even the DMV? Are these organizations behaving directly as a response to the majority will of the people? If not, how do you solve that problem before you add more organizations?

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

#274

Post by disabled »

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…

This is a thoughtful post. However, the issue is that these advertisements play into your hopes and fears to maximize the likelihood of getting a click from you. The fact that Google (especially) and other adtech companies are playing into your hopes and fears, by microtargeting and hoarding the most private and intimate details about your life is abusive.

I have to say that I am lucky that I have a print-related disability, because I almost never need to go websites with ads.

Services I get access to (no-ads):

* 975,000+ books for $50/year (Bookshare.org)

* 60,000+ professionally narrated audio books for free (US National Library Service)

* 80,000+ volunteer narrated audio books for $135/year (LearningAlly.org)

* Hundreds of Newspapers and Magazines for free (NFB Newsline)

* 99% of the books posted on OpenLibrary.org for free (even books currently "borrowed")

* Virtually all libraries for print-related disabilities around the world (sometimes free, sometimes paid) (I can get books in foreign languages easily)

Additionally, I use the paid audio apps Blinkist, Audm, and Curio, which everyone has access to. I find them to be super helpful. Blinkist in particular is almost 100% of the time a YouTube and TED talk replacement for me. I also use The Economist app, which has the entire weekly edition professionally narrated, along with the vast majority of the rest of its material.

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

#275

Post by purerandomness »

spijdar wrote: post #261
purerandomness wrote: post #223

Earlier quoted context omitted.

You're assuming that everyone uses online dating platforms to find one (1) long-term partner with whom they'll be monogamous, which is not the case. There are couples looking for other couples or thirds, there is the BDSM scene with people looking for casual play partners, and so on.

This is true, but I believe that the majority of users on "normal" dating sites are looking for single, long-term partners. As I understand, within the BDSM scene there are several websites including social networking sites and dedicated match making sites catering to the specifics of BDSM. I find it unlikely you'd use a "normal" dating site when you likely have pretty specific interests that likely (?) need specific…

Dating sites that make you answer questionnaires and match based on answers are a really good way to get to know people with similar kinks and interests.

While there are specific sites for BDSM dating with more nuanced optoins, the ads for generic dating sites are all very "tame" and try to not deviate from the perceived norm too much (= "find a partner, have a happy family" type messaging)

The reason is that if you do, it's virtually impossible to get included in Ad networks and App Stores. So you naturally see only dating ads catering to the very conservative viewer.

Example: A BDSM dating site got banned from Googles Play Store after including a background image of a simple leather whip. [1]

[1] https://twitter.com/devianceapp/status/1384015666185834501

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

#276

Post by auntienomen »

semi-extrinsic wrote: post #272
sobellian wrote: post #151

Earlier quoted context omitted.

We know that paid advertising works, so it should not surprise us to discover that paid advertising on the Internet also works. But Google and other search engines seek to organize the world's information, not the world's commercial products and services. For a multi-multi-billion dollar company's core product, I'm somewhat surprised they cannot do a better job killing the blogspam. Given the resources at their dispo…

"We know that paid advertising works" [citation needed] I mean, the ad business of course likes to throw various metrics around. But as far as I'm aware there are no proper randomized controlled trials that show statistically significant positive ROI of online advertising versus no online advertising. I mean, it would be really simple to do, right? To provide conclusive proof of the efficiency of ads? Pick a populous…

There are two significant complications:

1) Google is running many parallel ad campaigns, which may target the same individuals. This in some ways gives opportunities, because one can run 'natural experiments' on the effeciveness of advertising for X by simply selecting the people who never saw the ad for X. But there is also probably some legal peril; Google has to be careful about what promises it makes to people purchasing ads.

2) Google has very little incentive to release the results of any such studies, because -- whether or not advertising works -- they don't need their customers to have accurate side-info about the value of advertising.

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

#277

Post by 8note »

dhdc wrote: post #145
omgwtfbyobbq wrote: post #128

Earlier quoted context omitted.

So a negative positive feedback loop?

Just positive feedback loop. The term "positive" or "negative" refers to the effects of the feedback path on the overall system, a negative feedback loop attempts to negate any changes in the outputs, while a positive feedback amplifies any changes.

Just feedback loop*

Adding positive/negative on it is unnecessary specificity

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

#278

Post by pjmorris »

simonh wrote: post #270
nowherebeen wrote: post #205

Earlier quoted context omitted.

> The idea that companies should only be beholden to shareholders This concept was popularized by Milton Friedman with his Shareholder Theory. OP is not wrong. https://en.wikipedia.org/wiki/Friedman_doctrine

Maybe, but it’s repackaging a point Smith made 194 years earlier.

A shareholder is more like the butcher or baker's brother than the butcher or baker. The incentives are different.

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

#279

Post by ahepp »

simonh wrote: post #270
nowherebeen wrote: post #205

Earlier quoted context omitted.

> The idea that companies should only be beholden to shareholders This concept was popularized by Milton Friedman with his Shareholder Theory. OP is not wrong. https://en.wikipedia.org/wiki/Friedman_doctrine

Maybe, but it’s repackaging a point Smith made 194 years earlier.

I think Friedman's ideas are substantially different.

The quote from Smith is discussing tradesmen running a business in their own self interest.

In some ways, Friedman's point is the opposite. That the laborers perform in the self interest of the owner.

I don't know the full context of the Smith quote. I did a bit of digging for Smith's views on publicly traded companies, and came across this quote[0]:

>The directors of such [joint-stock] companies, however, being the managers rather of other people's money than of their own, it cannot well be expected, that they should watch over it with the same anxious vigilance with which the partners in a private copartnery frequently watch over their own.... Negligence and profusion, therefore, must always prevail, more or less, in the management of the affairs of such a company.

[0] https://en.wikipedia.org/wiki/Criticisms_of_corporations

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

#280

Post by uoaei »

daenz wrote: post #273
uoaei wrote: post #240

Earlier quoted context omitted.

Consider firms which are wholly owned by either 1) every single employee (worker-owned firm), 2) every employee and every customer who opts in (consumer co-op), or 3) literally every citizen of a government (publicly owned). You own it in partnership with the other people who own it. This is, in the strictest definition, the socialism that people are so scared of.

How much control do you have over the military of your government? Or even the DMV? Are these organizations behaving directly as a response to the majority will of the people? If not, how do you solve that problem before you add more organizations?

The best answer we have today, IMO, in terms of ethics of freedom and agency, is "representative democracy," which may be extended to "liquid democracy" to bridge the gap between small, direct-democracy-capable organizations and large, unwieldy ones.

Unfortunately, no solution will be perfect, but that doesn't mean some aren't better than others. The problem is in essence unsolvable. Politics and civilization is an exercise in minimizing harm rather than eliminating it, maximizing utility rather than spiking it.

Post reply on HN