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

#301

Post by dkarl » Mon, Apr 26, 2021, 9:28 PM UTC

dheera wrote: post #269

dkarl wrote: post #266

Earlier quoted context omitted.

Assuming you're looking for one lifelong partner, which isn't true of everybody, is it normal to find somebody "compatible" that quickly? Without apps, I think it's common for people to go for years between serious relationships. I don't know why the timeline needs to be so compressed. For me as a fairly awkward and introverted person, who didn't naturally generate a high volume of new social contacts, one of the thi…

Maybe. But I would think that that also introduces a paradox of choice where you are constantly doubting the person you are currently dating, thinking that maybe there is someone that is a better fit for you. The problem is I don't really think "fit" is an absolute thing. I think the reality is that there is a large set of people can be your best fit if you can grow together with them to be that best fit. A healthy r…

For me, doubt in my ability to know who I could be happy with rose dramatically with a little bit of experience and then fell as I accumulated more and more. Meeting more people made me more and more comfortable with my own judgment about other people and my understanding of what made me happy. I think people who find partners very early in life are very lucky in some ways, though. It's a trade-off, like so many other things. You can have X more years of experience with relationships and with yourself when you choose your partner, or you can have X more years of shared history with your partner.

I do think any doubts you can put to rest in six months or a year, the time is worth it. Couples who divorce take years to do it, and I think they're unhappy for at least half that time.

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

#302

Post by imtringued » Mon, Apr 26, 2021, 9:28 PM UTC

behnamoh wrote: post #219

derefr wrote: post #177

Earlier quoted context omitted.

Month-to-month is just as bad. The ideal business model for a dating site, from the users' perspective, is a one-time advance payment. This puts the business into the situation where they have an incentive to get you satisfied as quickly as possible, so that they can spend as little time/money on you as possible, so that your value to them doesn't go negative from allowing you to spend too much of their time/money. T…

IMO the ideal business model from users' perspective could be pay-as-you-go, where you pay for each individual you want to send message to (e.g. $1.99).

If you wanted to go down this route you would pay per date, otherwise what are you paying for? Sure, the algorithm may jinx it by sending you on more bad dates than you wanted, but it would get you further than just a message.

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

#303

Post by TeMPOraL » Mon, Apr 26, 2021, 9:43 PM UTC

dheera wrote: post #217

GCA10 wrote: post #212

Earlier quoted context omitted.

But wait! Yes, aspiring monogamists will fit your bill of people who "want to be off the site in 12 months" or sooner. That's one segment of your users, but it really isn't everyone by a long shot. Plenty of users are signing up for the chance to meet ("get to know") a steady stream of people. We don't stigmatize people who subscribe to Netflix for many years so that they can keep watching different movies and shows.…

> There's some segment of the dating-site world that has more of a Netflix model in mind Although I'm sure those users exist, I'm sure they aren't the majority of the world, who would rather just be happily married and get on with life? And even if not, these users who have different expectations should not be matching with the former.

> And even if not, these users who have different expectations should not be matching with the former.

Actually, given the extreme social stigma worldwide (even in the most progressive western countries) against casual hookups and low-commitment dating, people looking for "more of a Netflix model" will still gravitate towards the same sites ostensibly servicing those "who would rather just be happily married and get on with life"[0], because these services offer the widest choice of possible partners, while giving everyone plausible deniability.

--

[0] - I think that, given aforementioned stigma, it's even hard to estimate how many people in a given age bracket want this, and how many just say they want this, because it's the only accepted thing to say out loud.

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

#304

Post by p_j_w » Mon, Apr 26, 2021, 9:44 PM UTC

kingsuper20 wrote: post #232

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…

My main problem with advertising and the technologies surround it, aside from the obvious privacy issues and their misuse, is that it seems to suck all the air out of the room. The birth and dominance of the online advertising business model looks to be the greatest misallocation of engineering talent in the history of humanity.

This is a bit how I feel about advertising in general. Human beings' time is being taken and mouths are being fed not to increase overall output, and thus lifting the overall well being of members of society. Instead, Company A hires advertisers to convince the public to buy their product instead of a competing product to Company B. Value is created for Company A, but entirely at the expense of Company B. At no time in the economic... chain?... of events that is advertising is anything actually created, yet vast sums of money, and thus allocation of resources, is put here. It seems INSANELY wasteful.

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

#305

Post by ipython » Mon, Apr 26, 2021, 9:45 PM UTC

zinok wrote: post #220

ipython wrote: post #44

Earlier quoted context omitted.

I have an Audi that has both speed limit info and stop light info in the dash. I know for a fact that it uses the camera for the speed limit signs as it will interpret school zone areas. Sometimes it misses the “end school zone” sign or not recognize the yellow border around the school zone begin sign. You can see the display become out of sync with the road conditions until the next speed limit sign appears. As far…

Are you sure it's not using static data generated by a car which has driven the same roads and then has had speed limit sign detection run on the captured video?

Yes, it does the recognition on the car through a front facing camera.

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

#306

Post by bayesianbot » Mon, Apr 26, 2021, 9:47 PM UTC

behnamoh wrote: post #219

derefr wrote: post #177

Earlier quoted context omitted.

Month-to-month is just as bad. The ideal business model for a dating site, from the users' perspective, is a one-time advance payment. This puts the business into the situation where they have an incentive to get you satisfied as quickly as possible, so that they can spend as little time/money on you as possible, so that your value to them doesn't go negative from allowing you to spend too much of their time/money. T…

IMO the ideal business model from users' perspective could be pay-as-you-go, where you pay for each individual you want to send message to (e.g. $1.99).

There are a lot of dating sites at least in Scandinavia/central europe, where men pay per message (or usually buy message packs, it ends up being around 1€ a message IIRC).

The (mostly) men answering these messages, pretending to be women, get paid around 0.15€ per reply. And obviously writing messages where they try to prolong the conversation and turn down real life meetings or changing to other (free) messaging system "for now"

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

#307

Post by imtringued » Mon, Apr 26, 2021, 9:55 PM UTC

sudosysgen wrote: post #249

edgyquant wrote: post #204

Earlier quoted context omitted.

Let me know when we reach post scarcity and no longer need a rationing device such as money.

People thought of this issue since 1870 and the solution they came up with is simply non-transferable rationing devices. In traditional Marxist terms this was labour vouchers but nowadays there are much better solutions. Marx himself wrote about this, in terms of primitive accumulation in socialism. For sure there are a lot of issues, but this isn't one of them. You could in theory have bribery in material terms, but…

I don't really get it. How does this prevent people from holding onto foreign currency or gold?

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

#308

Post by rkagerer » Mon, Apr 26, 2021, 10:03 PM UTC

johndags wrote: post #214

zitterbewegung wrote: post #174

Earlier quoted context omitted.

The AI that makes Google work is not the algorithms they use to index the web its AdWords which uses a general purpose algorithm to auction ads. https://static.googleusercontent.com/media/research.google.c... Sure, google search was the core prototype that used algorithms to give great search results but without Adwords google wouldn't be google.

I think this is largely misunderstood about Google. It seems like ads placed in search results only accounts for a small portion of their revenue. Not sure how much the targeting of ads by analyzing your search history actually contributes either. I think Google just figured out how to scale online ad sales really well.

It seems like ads placed in search results only accounts for a small portion of their revenue.

Do you have a source for that? Last time I checked (which granted was some years ago) my understanding was that over 90% of their revenue came from advertising, and I think most of that was driven by search results.

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

#309

Post by sudosysgen » Mon, Apr 26, 2021, 10:03 PM UTC

edgyquant wrote: post #282

sudosysgen wrote: post #249

Earlier quoted context omitted.

People thought of this issue since 1870 and the solution they came up with is simply non-transferable rationing devices. In traditional Marxist terms this was labour vouchers but nowadays there are much better solutions. Marx himself wrote about this, in terms of primitive accumulation in socialism. For sure there are a lot of issues, but this isn't one of them. You could in theory have bribery in material terms, but…

My argument was with the idea that money will be gone as a rationing device. You can come up with alternatives but it’s the human nature that is the problem not the technology we use to ration resources. Let me know when this bug is fixed.

Well, that's not the problem we were talking about, is it? We were talking about the issue of bribery, which is made possible by the fungible nature of money.

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

#310

Post by imtringued » Mon, Apr 26, 2021, 10:03 PM UTC

p_j_w wrote: post #304

kingsuper20 wrote: post #232

Earlier quoted context omitted.

My main problem with advertising and the technologies surround it, aside from the obvious privacy issues and their misuse, is that it seems to suck all the air out of the room. The birth and dominance of the online advertising business model looks to be the greatest misallocation of engineering talent in the history of humanity.

This is a bit how I feel about advertising in general. Human beings' time is being taken and mouths are being fed not to increase overall output, and thus lifting the overall well being of members of society. Instead, Company A hires advertisers to convince the public to buy their product instead of a competing product to Company B. Value is created for Company A, but entirely at the expense of Company B. At no time…

>At no time in the economic... chain?... of events that is advertising is anything actually created, yet vast sums of money, and thus allocation of resources, is put here. It seems INSANELY wasteful.

You're forgetting something, companies start off completely unknown. How did they reach the point where the market has been fully saturated and the only real way to gain more customers is to take them from someone else? Oh right, it's because advertising increased the grow rate of your company to the point where there is barely any growth left.

Let's manufacture a completely artificial scenario to illustrate my point:

Person A: So, you're telling me you spent $5 billion on advertising and all you have to show for it is a 5% higher market share than your biggest competitor?

Founder: Yes, we used the advertising budget to grow our market share from less than 1% to 40%. Our next biggest competitor has a 35% market share.