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

281–290 of 427 posts

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

#281

Post by thescriptkiddie »

marcusverus wrote: post #263
thescriptkiddie wrote: post #201

Earlier quoted context omitted.

> This article has multiple issues

I assume that this criticism is offered in good faith, and that you're in need of good, solid Wiki articles about the tens of millions of victims of Communist regimes--well, I'm happy to get you started! Here's an article about the Soviet terror-famine (known as the Holodomor) which killed 4 million Ukranians. No worrisome notifications on this article, so I assume it meets your rigorous standards: https://en.wikiped…

I'm not denying or excusing that historical atrocities occurred under socialist governments, but for perspective one should also look at the myriad atrocities that were and continue to be committed under capitalist governments. That doesn't excuse such actions, but neither side is innocent. I suggest reading The Wretched of the Earth* Chapter 1, "On Violence".

Here are a few examples off the top of my head:

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

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

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

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

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

https://en.wikipedia.org/wiki/1985_MOVE_bombing

https://en.wikipedia.org/wiki/Bengal_famine_of_1943

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

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

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

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

* Though I will object that it is unfair to attribute the actions of the Khmer Rouge to socialism. Like the Nazis they were socialist in name only, and in fact were supported by the United States in their war against the Socialist Republic of Vietnam.

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

#282

Post by edgyquant »

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…

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.

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

#284

Post by bglazer »

xyzzy21 wrote: post #239

Well, yeah. Because "machine AI" has NOTHING much in common with how our biological brains work. And we aren't smart enough to know what intelligence really is when we can't even define it for ourselves or in animal models. And 99.999% of everyone working on machine AI has never taken a biology class let alone a class related to anatomy, neurology or experimental psychology so it's nothing more than "flinging shit on…

You're incorrect that neuroscientists and computational researchers don't collaborate. In fact, "computational neuroscience" departments have existed at major universities for a number of years. DeepMind was a spin-off of the Gatsby Computational Neuroscience Unit at UCL, and they continue to be influential in both neuroscience and (more famously) in artificial intelligence research. Here's an article from 2020 by authors at DeepMind and Geoff Hinton about potential biologically plausible mechanisms for backpropagation in the brain [0]. So, you can see that the fields actually influence each other in both directions. Computational researchers often propose ideas that neuroscientists then attempt to understand in biological systems. This same principle is true at all levels of computational and biological abstraction, from simulating individual neurons to machine learning and common sense. For the latter, research by Josh Tenenbaum might interest you.

The reason this works is that there seem to be fundamental principles underlying information processing. Brains and CPU's are both systems that manipulate and store information, albeit in very different ways. Hell, even single cells and slime molds are capable of rudimentary decision making.

So, the point is that we don't need to copy the brain. Instead, we just need to understand the principles of information well enough to build machines that can efficiently manipulate information and intelligence will arise out of that. Information theory and (by extension) statistics are the fields that deal most closely with this question, which is why they're used heavily by both neuroscientists and ML researchers.

A rough analogy is that we don't build airplanes that flap their wings to fly. Instead, we understand aerodynamics well enough to generate lift and thrust through other mechanisms that evolution can't find. Like jet engines.

Also, "slavishly" copying nature is insanely difficult. Biological neurons and brain tissue are extremely complex and poorly understood. Much of this complexity is likely incidental to information processing and would only hamper our efforts to build intelligent machines. Like, do we want our computer brain to get multiple sclerosis if the simulated neurons demyelinate?

[0] https://www.nature.com/articles/s41583-020-0277-3

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

#285

Post by heterodoxxed »

SkyBelow wrote: post #227
Zelphyr wrote: post #129

Earlier quoted context omitted.

Or, maybe, "Everybody is altruistic until they have shareholders." The idea that companies should only be beholden to shareholders that has taken firm hold over the past 50(+/-) years doesn't look to be a good one, in hindsight.

Psychology studies in the past showed people were altruistic. Then psychologist accounted for social capital/good will and worked to remove it from their altruism tests. People stopped being overly altruistic when it stopped benefitting them, likely meaning that what we see as altruism is really a failure to account for all the benefits a person expects to gain and all the negatives they expect to avoid when choosing…

The concept of altruism does not require it to be "pure" or entirely selfless. Altruism can have benefits but those benefits can exist outside of economy and into the realm of the personal, spiritual, social, etc.

In other words, that doesn't disprove altruism so much as it proves that economic self-interest is not our only motivator.

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

#286

Post by heterodoxxed »

throwaway3699 wrote: post #191
erikpukinskis wrote: post #124

Earlier quoted context omitted.

That doesn’t really explain anything, given there are millions of deaths under capitalism too.

Big difference.

Not to the bodies in the graves there isn't.

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

#287

Post by toomuchredbull »

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…

This guy doesn't understand modern dating...

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

#288

Post by heterodoxxed »

marcusverus wrote: post #263
thescriptkiddie wrote: post #201

Earlier quoted context omitted.

> This article has multiple issues

I assume that this criticism is offered in good faith, and that you're in need of good, solid Wiki articles about the tens of millions of victims of Communist regimes--well, I'm happy to get you started! Here's an article about the Soviet terror-famine (known as the Holodomor) which killed 4 million Ukranians. No worrisome notifications on this article, so I assume it meets your rigorous standards: https://en.wikiped…

I don't know if you want to open the can of worms that is "accidental deaths from mismanaged resource distribution" under capitalism.

Global deaths from hunger result in one great leap forward every 5 years.

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

#289

Post by asdff »

jeffbee wrote: post #110
hoppyhoppy2 wrote: post #91

Earlier quoted context omitted.

>I'm sure it's not using the internet to do it. How sure are you? >Daniel Dunn was about to sign a lease for a Honda Fit last year when a detail buried in the lengthy agreement caught his eye. Honda wanted to track the location of his vehicle, the contract stated, according to Dunn — a stipulation that struck the 69-year-old Temecula, Calif., retiree as a bit odd. [...] >There are 78 million cars on the road with an…

>> I'm sure it's not using the internet to do it. > How sure are you? Pretty sure. The Honda Sensing display of the local speed limit changes on the dashboard at the instant you pass the sign, it's too exact to be using GPS. Nobody has geocoded the location of every sign in America, that would be too much work. Also, it's often wrong but in ways that you would be wrong if you were just reading the signs. For example,…

My garmin does this instant speed change, and it is pulling GPS.

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

#290

Post by mcguire »

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 roughly that same time, doubleclick.net was the most hated company on the Internet.

Then Google bought them.

Post reply on HN