If you feed AI conspiracy theories and then it tells you Elvis is still alive, that’s an input problem not an algorithm problem.

Now, getting to an AI that doesn’t “hallucinate” is a little more complicated than simply filtering out “conspiracy theories” from data, but IMhO it’s not many orders of magnitudes away. Far from insurmountable in a couple Moores law cycles.

I think human brains operate on the same principal of divining next tokens. We’re just judging AI for not saying the tokens we like best even though we feed AI garbage in and don’t tell AI what it should even care about. “AI doesn’t live here” it wasn’t born “here.”

Someday soon we’ll probably give AI guard rails to respond considering the context of society (the programmers version of society), and it will probably hallucinate less than most humans.

I mean… even magicians (mentalists) can reliably hack humans into generating the next token they want you to generate.

And humans habitually stray from the “truth” too. It’s always seemed to me that getting AI to be more accurate isn’t a math problem, it’s getting AI to “care” about what is true - aka better defining what truth is- aka what sources should be cited with what weights.

We can’t even keep humans in society from believing in the stupidest conspiracy theories. When humans get their knowledge from sources indiscriminately, they also parrot stupid shit that isn’t real.

Now enter Gödel’s incompleteness Theorem: there is no perfect tie between language and reality. Super interesting. But this isn’t the issue. Or at least it’s not more of an issue for robots than it is for humans.

If/when humans deliver “accurate” results in our dialogs, it’s because we’ve been trained to care about what is “accuracy” (as defined by society’s chosen sources)

Remember that AI “doesn’t live here.” It’s swimming in a mess of noisy context without guidance for what it should care about.

IMHO, as soon as we train AI to “care” at a basic level about what we culturally agree is “true” the hallucinations will diminish to be far smaller than the hallucinations of most humans.

I’m honestly not sure if that will be a good thing or the start of something horrifying.

Comments URL: https://news.ycombinator.com/item?id=29791643

Points: 5

# Comments: 0

https://www.shapeways.com/product/S3YNU5MNR/super-accurate-r...

"The National Electoral Commission (NEC) declared a runoff after none of the 16 presidential hopefuls on the March 7 ballot paper secured the 55% needed to secure an outright victory."

http://www.africanews.com/2018/03/15/sierra-leone-presidenti...

How is it okay that this information is even available from Experian in the first place?

I don't think anyone opted in to this. And I don't think there is any obvious way to opt out.

And Experian is not just some data tracking company watching your behavior on a website; they're supposed to be protecting our credit system. Do they obtain some of this information through special privileges because they're a pseudo-official credit score agency? If so, is this grounds for a class action lawsuit?

(As others have argued, this data is legally available through Experian, it just normally isn't free. https://www.experian.com/assets/dataselect/brochures/consume...)

As others have noted, the camera is auto-centering the object. If you look at the object's movement relative to the clouds in the background, it seems that the object might be turning and exposing its broader side to the relative direction of travel as seen from the camera. And then slowing down in the plane as viewed from the camera.

Note, this movement is relative and it's very hard to get a sense of perspective. So rather than slowing down, this could be the object banking and changing direction to move towards or away from the camera?

Maybe it's just a foreign aircraft with some kind of wingless shape like the B-2 bomber? I can't find a good source, but it seems there are blogs talking about new foreign stealth drones that could look like this. (http://defence-blog.com/news/new-stealth-bomber-spotted-in-c... -- not sure if this source is legit)

P.S. This is all assuming that this is indeed some kind of flying thing and not some kind of mirage.

So I'm disappointed that it seems I can't signup and contribute without installing a browser plugin that can read and change all my browser data and browser history... nope.

Flattr team, please fix that and I'll consider giving you another chance.

P.S. I've been keeping my eye open for someone trying to implement this subscription/contribution model using blockchain. I haven't been able to find anything -- anyone seen that yet?

As a Verizon customer, I didn't realize that my cell phone plan information was being sold to their affiliate marketers. Apparently, this is 'opt in' by default and I have to opt out. I'm sure that I was told about this change in some contract small print. But of course that is a poor justification for such a nasty practice. I wonder if this is why I've been getting so many spam phone calls.

This is so frustrating. But, of course I'm not surprised in the least. Maybe I should finally switch to Project Fi or some other alternative...

How to opt out: While logged in, navigate to (My profile -> Privacy Settings) Or, go to this page: https://nbillpay.verizonwireless.com/vzw/secure/setPrivacy.action

Full information: http://www.verizon.com/about/privacy/customer-proprietary-network-information

Edit: There are plenty of articles out there discussing the recent legislation that has led to an increase in policy changes like this. http://www.popularmechanics.com/technology/security/a25859/how-to-protect-your-online-privacy-from-isp/

Comments URL: https://news.ycombinator.com/item?id=14482356

Points: 3

# Comments: 0