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

There is no A.I.

newyorker.com

Post reply on HN
Print view View original

231–240 of 523 posts

Re: There is no A.I.

#231

Post by optimalsolver »

throwaway8503 wrote: post #133
ravenstine wrote: post #100

Earlier quoted context omitted.

Dunno if there's some trickery going on under the hood, but GPT-4 does basic arithmetic correctly.

This is what I got on ChatGPT today. I assume it used GPT4: Prompt ChatGPT Actual Match 397,356 * 930,547 369,685,207,932 369,758,433,732 FALSE 36,330 * 26,951 979,458,630 979,129,830 FALSE 8,681 * 9,330 80,911,430 80,993,730 FALSE 278 * 903 250,734 251,034 FALSE 82 * 77 6,314 6,314 TRUE Edit: # of correct digits (counting from leftmost) only exceeds 3 on the smallest pair. It drops to two, as well, on the 3x3 set.

Unless you're using Plus (black icon), you're using GPT-3.5 (green icon).

Re: There is no A.I.

#232

Post by GauntletWizard »

coldtea wrote: post #154
sclarisse wrote: post #126

Earlier quoted context omitted.

Corporations have rights because they are owned by humans and those humans have rights. Corporations are just a convenient way to do things together, like conduct a business and own property. Corporate personhood is just a legal abstraction to represent those peoples’ rights, a facade pattern that lets multi-person groups neatly fit in existing laws that might discuss individuals. Corporations like Citizens United ca…

> Corporations have rights because they are owned by humans and those humans have rights. One does not really follow from another. Washing machines are also "owned by humans" but they don't get rights. > Corporations are just a convenient way to do things together, like conduct a business and own property. Corporate personhood is just a legal abstraction to represent those peoples’ rights, a facade pattern that lets…

As the foundation, the cornerstone, the most base part of said "Corporate law". Corporations do not exist without some humans behind them. There are no fully-autonomous corporations in corporate law (despite what ethereum would have you believe) . There are only groups of people. Those people, officers of corporations, impart the personhood - The Corporation is just an embodiment of the collective action of the officers.

Re: There is no A.I.

#233

Post by GauntletWizard »

HyperSane wrote: post #139
counttheforks wrote: post #94

Earlier quoted context omitted.

I never really understood this. Can a company go to prison or get the death penalty?

Arthur Andersen was basically executed.

As was Silicon Valley Bank (though not it's parent corporation, SVB Financial).

Re: There is no A.I.

#234

Post by bell-cot »

agronomov wrote: post #221

It seems that one danger is often overlooked, which is humans believing AI has consciousness once it becomes advanced enough. It will start with dying people migrating to cloud hoping for eternal life, and on the outside it would look like they continue to live in the metaverse forever. With time the metaverse would look better than real world in every possible way and people would start to voluntarily move to metave…

I don't think that scenario actually works. Reason: the superficial, "sounds like a person" AI's don't actually have what it takes to sustain a huge, high-tech society.

OTOH, letting the idle rich "move" to the matrix could be a good thing...

Re: There is no A.I.

#235

Post by visarga »

vimax wrote: post #146
jstummbillig wrote: post #120

Earlier quoted context omitted.

Agreed. One of the more interesting things about AI is how it forces humanity on a trip to consider what it means to be human. There is no soul, creativity or inspiration. There's only (very complex) agency. And that's completely fine.

To me it shows how fundamental language is to intelligence. Not just that LLMs are very capable by themselves, but significantly improve speech and image models. We’ll need to break out of the Chomsky hierarchies and develop some new theories of language. Vocal communication isn’t unique to humans, a new theory should be more broadly applicable to non human and non vocal languages. I’m excited to see what big questio…

> To me it shows how fundamental language is to intelligence.

I believe most human intelligence is based on language. We are smart because language gives us the keys to solving all sorts of situations based on the experience of prior generations. We are wondering if GPT architecture can generate real understanding, but forgetting the training corpus. The language corpus is the real magic here. The source and accumulation point of intelligence.

Take a baby and give it access to language, you get a modern human. Take a random init neural net, use language corpus for training, you get GPT4. It doesn't matter what the model is - brain or transformer, it doesn't matter if it's a transformer or just a RNN (RWKV). What matters is the training signal.

We should not get so hung up on the model architecture and substrate when thinking about LLM capabilities. Language is a separate evolutionary system of ideas and concepts. It is a self replicator, like DNA. Language just got a new pathway for self replication with LLMs. Every new token visits the whole model and by extension the distillation of our culture before being emitted.

Re: There is no A.I.

#236

Post by agronomov »

bell-cot wrote: post #234
agronomov wrote: post #221

It seems that one danger is often overlooked, which is humans believing AI has consciousness once it becomes advanced enough. It will start with dying people migrating to cloud hoping for eternal life, and on the outside it would look like they continue to live in the metaverse forever. With time the metaverse would look better than real world in every possible way and people would start to voluntarily move to metave…

I don't think that scenario actually works. Reason: the superficial, "sounds like a person" AI's don't actually have what it takes to sustain a huge, high-tech society. OTOH, letting the idle rich "move" to the matrix could be a good thing...

Well, imagine advanced AI beyond imagination. It is so advanced that it mimicks all of the human activity in exactly the same way. By definition it would be able to do whatever we are doing (and probably better)

Unless you don't believe that such an AI is possible, in which case I wonder if you have tried GPT-4 and whether you remember what chat bots were like 5 years ago!

This of course would be AGI. It would still not be conscious though, but a lot of people would fall victim to it and believe that it is. And that's what will end us. We will not pass the mirror test (yep, this one https://en.wikipedia.org/wiki/Mirror_test)

Re: There is no A.I.

#237

Post by Lio »

Joeri wrote: post #205
nanidin wrote: post #180

Earlier quoted context omitted.

This is the infuriating part of dealing with Amazon customer service these days. Anything that doesn’t fit into the box of what the chatbot will do is met with “I understand your frustration, …” and it’s like, no, there is no understanding here! Just like call center trees have escapes to real service by pressing 0 ad nauseum or swearing loudly, these AI service agents will have ways to get to real people and they’ll…

How many times have you read about someone locked out of their account and having no recourse except taking it to HN and hoping a human on the inside reaches out? Imagine if that was every corporation, and some of them had zero humans on the inside.

Whilst it's interesting to imagine every company being run by the most aggressively rent seeking MBA-like AI, I don't think that will ever be the entire market.

I would guess it all comes down to risk. The future won't be uniform, it'll be poorly defused, just like then present.

People will make a value judgement on the types of corporations they want to deal with. The sort of company that no has employees will by necessity have to be very cheap compared to those that provide interactive human service.

Dealing with AI firms will be like dealing with an anonymous Chinese eBay seller. You'll effectively have no legal recourse as such and they won't discuss things but complaining might get you a replacement item or a full refund.

If it doesn't the product will have to be cheap enough to write off as a bad buy.

Let's hope cheap AI powered corps don't suck all the profit out of the market and we loose all firms capable of providing real services.

Or we end up paying for human service that's been quietly outsourced to AI much like we do for many big name manufactured goods.

Re: There is no A.I.

#238

Post by 4ndrewl »

Animats wrote: post #68

The elephant in the room in these discussions of AI is the concept of the corporation. Most of the things people are worried about AI doing are things corporations already do. Deceptive marketing, exploiting customer information, conspiring to keep prices high, creating distractions from things companies are doing, setting up monopolies, lying to customers - the usual. With AI, this can be automated. Most of this is…

Well yes - these LLMS are biased towards existing structures of power, and employed within them and will only reinforce them at scale.

Re: There is no A.I.

#240

Post by visarga »

vimax wrote: post #192
coldtea wrote: post #163

Earlier quoted context omitted.

> We’ll need to break out of the Chomsky hierarchies and develop some new theories of language. We might need to develop some "new theories of language", covering other areas, but why exactly do you think we'd need to "break out of the Chomsky hierarchies"? Nothing about LLMs challenges Chomsky hierarchies. They do challenge Chomsky's ideas about language being an innate human quality, but that's about it. Besides th…

The Chomsky hierarchy will remain unchanged and very useful for constructed languages. I didn't mean break out, as in create a higher level. LLMs are not going to magically become more than a turing machine or anything. I meant we need to stop thinking about spoken language in terms of the hierarchy and try to find another approach. You can't tell me with a straight face that actual spoken language fits well in that…

> 100B parameter ... not well suited for describing spoken languages if it takes such a large definition.

Humans by contrast use about 700T synapses to support our own understanding of language.

Post reply on HN