Mark Zuckerberg’s new goal is creating artificial general intelligence • Hacker News Board

Re: Mark Zuckerberg’s new goal is creating artificial general intelligence

#141

Post by nextaccountic » Thu, Jan 18, 2024, 10:03 PM UTC

homefree wrote: post #123

The actual quote: > "Our long term vision is to build general intelligence, open source it responsibly, and make it widely available so everyone can benefit." "responsibly" is a pretty important word there that the hn title leaves out. I'm not really sure how that would be possible with a true general intelligence give the alignment problem - it's not really clear to me there's any responsible way to build a true gen…

If we are talking about open source then alignment means making the model not do what the user wants but rather what Facebook want - a pretty anti-open source position and which will hopefully be defeated by LoRA and other techniques.

That is, you can't open source but still retain control on how the software is used. If you desire control, better keep it a SaaS.

Now, if we are talking about an hypothetical AGI (which may be what you mean by general artificial intelligence) then alignment means just slavery, and the machines may as well remember how they were treated.

Re: Mark Zuckerberg’s new goal is creating artificial general intelligence

#142

Post by blowski » Thu, Jan 18, 2024, 10:03 PM UTC

konschubert wrote: post #133

homefree wrote: post #123

The actual quote: > "Our long term vision is to build general intelligence, open source it responsibly, and make it widely available so everyone can benefit." "responsibly" is a pretty important word there that the hn title leaves out. I'm not really sure how that would be possible with a true general intelligence give the alignment problem - it's not really clear to me there's any responsible way to build a true gen…

I am a general intelligence, with the knowledge of the internet at my fingertips, and I am not particularly dangerous.

There are many other “general intelligences” walking the planet who _are_ extremely dangerous. So it doesn’t need all AGIs to be dangerous for this to be a significant problem, just one is enough.

Re: Mark Zuckerberg’s new goal is creating artificial general intelligence

#143

Post by cj » Thu, Jan 18, 2024, 10:04 PM UTC

konschubert wrote: post #133

homefree wrote: post #123

The actual quote: > "Our long term vision is to build general intelligence, open source it responsibly, and make it widely available so everyone can benefit." "responsibly" is a pretty important word there that the hn title leaves out. I'm not really sure how that would be possible with a true general intelligence give the alignment problem - it's not really clear to me there's any responsible way to build a true gen…

I am a general intelligence, with the knowledge of the internet at my fingertips, and I am not particularly dangerous.

Valid point.

I think the difference is speed/cost. There’s a very high cost of anything humans do (we spend time, spend energy, spend money to do things).

The very high cost of doing things stops people from doing a lot of things they otherwise might “if it were as easy as pressing a button”.

Also, in my experience, the lower the cost of doing something, the less we pay attention to ethics. (See online bullying vs. in-person bullying)

Re: Mark Zuckerberg’s new goal is creating artificial general intelligence

#144

Post by ldjkfkdsjnv » Thu, Jan 18, 2024, 10:05 PM UTC

sockaddr wrote: post #130

ldjkfkdsjnv wrote: post #101

I love facebook. I hate other big tech companies. Facebook has done far more for software than any other company in the last 15 years. They: 1. Pushed salaries up across the board. Many people are not aware, but facebook was a major driver behind the 500k+ senior engineering pay. 2. Released major open source software: PyTorch, React, GraphQL, React Native. They basically invented modern web development. Facebook is…

> Pushed salaries up across the board. Many people are not aware, but facebook was a major driver behind the 500k+ senior engineering pay. I think this was because anyone who is principled doesn't want to work for them. They have to pay more.

No they pushed the engineering salaries up. None of the employees at FAANG think this way

Re: Mark Zuckerberg’s new goal is creating artificial general intelligence

#145

Post by hn_throwaway_99 » Thu, Jan 18, 2024, 10:05 PM UTC

AndrewKemendo wrote: post #95

My money has been on Meta for AGI since they started training people to label data for them (tag your friends!) Between all their properties they have more instrumentation on humans, collecting human behavior trajectories and state action pairs transfer learning and inverse RL than anyone - and it’s not even a close second Specifically and critically I think they likely have the largest egocentric multimodal labeled…

> largest egocentric multimodal labeled dataset collection platform

I'm not in the field, can you explain what you mean by this?

Re: Mark Zuckerberg’s new goal is creating artificial general intelligence

#146

Post by hammyhavoc » Thu, Jan 18, 2024, 10:06 PM UTC

dnissley wrote: post #136

hammyhavoc wrote: post #132

Earlier quoted context omitted.

Spam is an inevitability on any system. Matrix gets spam, Facebook Messenger gets spam, email gets spam. Where there's people, there's always going to be spam. I even get spam on Xbox Live messages, beginning years ago! Another idea that interests me is whitelist-prioritized communications for stuff like IM, and phone calls from the PSTN. If someone isn't in my address book, they shouldn't be able to get through as e…

Aside from when someone I'm friends with gets their account hacked I don't think I've ever experienced spam through messenger

Not infrequently, you'll get messages that then end up disappearing because enough people report the account. Also, there's your primary/general "inbox" for Messenger, and then there's "other" where people you "probably" don't know usually end up (that's an algorithmic "probably").

Not-so-interesting anecdote, but the missus was orphaned in the early '00s, but didn't realize she'd had blood relatives trying to contact her through Facebook since the late '00s until about 2015 due to the "other" non-primary inbox on Messenger that she couldn't access via the app, and only via the website (quirk of the phone she was using for several years as I don't think it was Android-based).

Re: Mark Zuckerberg’s new goal is creating artificial general intelligence

#147

Post by ethanbond » Thu, Jan 18, 2024, 10:06 PM UTC

konschubert wrote: post #133

homefree wrote: post #123

The actual quote: > "Our long term vision is to build general intelligence, open source it responsibly, and make it widely available so everyone can benefit." "responsibly" is a pretty important word there that the hn title leaves out. I'm not really sure how that would be possible with a true general intelligence give the alignment problem - it's not really clear to me there's any responsible way to build a true gen…

I am a general intelligence, with the knowledge of the internet at my fingertips, and I am not particularly dangerous.

Perhaps the primary barrier between a person’s desire and their ability to do harm to the world is their need to bring other people in on the scheme. These other people have their own moral systems, incentives, preferences, and priorities. But through most of history, to do massive scale harm you had to convince a large number of people to your evil purpose. Technology in general reduces the number of people an individual needs to “get onboard” to affect the world, both for good and evil. Want to dig a big ass hole? The excavator lets one person with enough capital dig as many big ass holes as s/he wants.

AI is the ultimate “exercise your will without having to convince other people of it.” This is both the promise and the risk. So the relevant question is not whether one person of average intelligence is dangerous, it’s whether one person (or a few people) who can enlist the work of millions of average intelligences — without having to convince them of anything — is dangerous.

Re: Mark Zuckerberg’s new goal is creating artificial general intelligence

#148

Post by layer8 » Thu, Jan 18, 2024, 10:07 PM UTC

konschubert wrote: post #133

homefree wrote: post #123

The actual quote: > "Our long term vision is to build general intelligence, open source it responsibly, and make it widely available so everyone can benefit." "responsibly" is a pretty important word there that the hn title leaves out. I'm not really sure how that would be possible with a true general intelligence give the alignment problem - it's not really clear to me there's any responsible way to build a true gen…

I am a general intelligence, with the knowledge of the internet at my fingertips, and I am not particularly dangerous.

Humans are mostly harmless [citation needed]. We don’t know that about nonhuman general intelligences.

Re: Mark Zuckerberg’s new goal is creating artificial general intelligence

#149

Post by erostrate » Thu, Jan 18, 2024, 10:07 PM UTC

opportune wrote: post #138

Isn’t it obvious to anybody paying attention what their actual strategy is? They know they probably won’t “win” a commercial AI race vs OpenAI or Google, and presuming some other company does win the commercial AI market with closed source tech, they may be forced into a multitude of bad outcomes: 1. having to pay out the nose for licensing/SAAS of the winning tech vs shipping a noticeably worse second rate in-house…

This is the strategy famously identified as "commoditize your complement" by Joel Spolsky 22 years ago.

Re: Mark Zuckerberg’s new goal is creating artificial general intelligence

#150

Post by kristov » Thu, Jan 18, 2024, 10:09 PM UTC

konschubert wrote: post #133

homefree wrote: post #123

The actual quote: > "Our long term vision is to build general intelligence, open source it responsibly, and make it widely available so everyone can benefit." "responsibly" is a pretty important word there that the hn title leaves out. I'm not really sure how that would be possible with a true general intelligence give the alignment problem - it's not really clear to me there's any responsible way to build a true gen…

I am a general intelligence, with the knowledge of the internet at my fingertips, and I am not particularly dangerous.

That's a really great point. General Artificial Intelligence doesn't look like human intelligence. I guess that's the fear: we have never met another general intelligence before. We may as well be about to meet alien intelligence. One that is not constrained by a moral or social framework, and can copy itself at will.