Hacker News Reader

Hacker News

Comment by dr_dshiv | Hacker News Reader

Full thread

dr_dshiv

Propositions:

1. GPT3 is AGI

2. In combination with people, GPT3 is superintelligence

Strong claims. Anyone care to argue for or against?

View on HN

> Anyone care to argue for or against?

You go first.

I tried in the thread to give clear definitions of intelligence and AGI in order to make an argument for.

I didn’t give references for the definitions, but for intelligence, I think Legg and Hutter (2007) & Russell and Norvig (2010) work. And Wikipedia has some tests for AGI.

I asked GPT3 “if you were going to go into a person’s house and make a cup of coffee, detail what steps you’d take.” GPT3 responded with text that is indistinguishable from what I’d expect from a high school graduate, at least.

1. Knock on the door or ring the doorbell.

2. Wait for the person to answer the door.

3. If the person invites you in, step into the house.

4. Find the kitchen and locate the coffee pot and coffee beans.

5. Measure out the desired amount of coffee beans and grind them.

6. Place the ground coffee in the coffee pot.

7. Fill the coffee pot with water and turn it on.

8. Wait for the coffee to finish brewing.

9. Pour the coffee into a mug and enjoy.

Intelligence is the capacity to survive and thrive by taking actions that are likely to succeed. Artificial intelligence is defined by the ability to take actions that maximize measures of success. Artificial general intelligence is defined by the ability to operate intelligently regardless of context; ie, not merely trained in a particular context but capable of abstracting and generally succeeding in a range of contexts roughly as broad as a human.

By those reasonable and common definitions, GPT3 seems like a form of AGI. I mean, it is used for protein synthesis sequences, robotic sequences, design sequences— and it is a fabulous writer in a vast range of domains.

It doesn’t need to be god to be AGI v1.

I work with it every day and I recognize my contributions to it. But it is so generally intelligent. That is, it produces successful outcomes (useful text) in a vast range of contexts.

There is zero evidence for the AGI claim or even a weaker intelligence claim. GPT 3 is just a search engine on steroids. I'm regularly suprised by the results but it's honestly just an illusion as far as 'intelligence' goes. This has nothing to do with intelligence, and anyone who thinks otherwise is succumbing to the Eliza effect.

Why can't the same thing be said of human intelligence? If we knew more about neuroscience it'd probably seem just as trivial. Wetware trained on examples we've seen. Experience encoding neurons. Neurons running on the laws of physics. Data goes in, endogenous processing happens, then outputs. I don't see why it's categorically different. ANNs are just demystified because we made them and know how they work.

You've answered your own question.

"why can't the same thing be said of human intelligence"

"if we knew more about neuroscience..."

So that's why we can't say the same thing.

Fwiw, my take is that whatever the brain does, its a machine of some sort but possibly not even a turing machine. Many disagree but noone can prove me wrong. Yet at least.

So are we intelligence or just a pattern recognition system on steroids?

The human mind in its totality, ie including consciousness, is a fundamentally different kind of machine to a computer. It's a machine, there is nothing supernatural here, and I am not invoking a 'god of the gaps' argument. However, like Roger Penrose, I think that our consciousness is not a computation for example. That kind of thing seems pretty key. If that is true, and what we call 'understanding' is a function of consciousness, then that's going to have implications.

It doesn't matter by the way, if Turing machines can't replicate human minds, why would we want that anyway? "Intelligent" systems that provide outputs and solve problems for us is what we want right? I think we'll have something that is, for all intents and purposes, "AGI", using classical computers. A 'philosophical zombie' in other words.

GPT3 works only on text, both input and output. A system thus constrained, solely by virtue of that fact, is by definition not an AGI.

Why not? You can control robots with a sequence of text instructions.

You can represent anything as text.

That's either under some pretty narrow definition of "anything" or some pretty broad definition of "text".

You can start proving your point by representing the smell of incense as "text" and see how far that takes you.

Do the constraints our brains work under disquallify us from having general intelligence?

No.

Then your defintion of intelligence is kind of arbitrary

Reply on news.ycombinator.com